SAAR Oe ye ene eS ~The NT eee Seafa SeD ent eee eT cee Rope tates tnm rode sr ee SS i dots od PARES me oe Hand AOE i Ne has at POP ESS 5 a ; y iA Thee # fubf nAt. ct Ab eis PBI eis: ‘ee ae a =, AE , <i: _) ‘ 4 i = a - 7. . @ : 7 iy - = = aan | oe > | a i _ = a - od 8 ® l«@ Vw SIP YD ei ALSIUU WuldU LUGUGHLINY Z ( Fe TL IRATE eye Us song Reigate eam PROCEEDINGS OF THE CALIFORNIA AUADEMY Ut SOLENGES. SECOND SERIES. VOLUME 1, 1SSS. Committee of Publication: H. W. HARKNESS, GEORGE HEWSTON, H. H. BEHR, CARLOS TROYER, MARY K. CURRAN. EDITOR : - Mary K. CURRAN. “ Sait Francisco, 1880. Officers of the Academy, President. HARVEY W. HARKNESS. First Vice-President. H. HERMAN BEHR. Second Vice-President. GEORGE HEWSTON. Recording Secretary. J. R. SCUPHAM. Corresponding Secretary. FREDERICK GUTZKOW. Tabrarian. CARLOS TROYER. Treasurer. I. EK. THAYER. Director of Museum. J. G. COOPER. Trustees. CHARLES F. CROCKER, GEORGE. C. PERKINS, D. E. HAYES, K. J. MOLERA, S. W. HOLLADAY, IRVING M. SCOTT, S a — ae TABLE OF CONTENTS. PAGE Behr, H. H, Changes in the Fauna and Flora of California .......... 94 Brandegee, Townshend H. Flora of the Santa Barbara Islands .. . 201 Bryant, Walter E. Unustial Nesting Sites. IT................0.0. 7 Birds and Eggs from { ‘e Farallon Islands.................-+. 25 Description of a Subs 2cies of Song Sparrow from Lower Cal- UORRSIET 1 draft Pee ae Te REP Mey eis eh acer, « atlas ae eis 197 Cooper, J.G. West Coast Pulmonata, Fossil and Living............ 11 Curran, Mary i ebotanicaleNotess a0. a9. emsoek's dos eh are Ne oi 227 Eigenmann, Carl H. and Rosa§, American Gobiide and Calliony- Maa eos MRA RA IY cin ASG A Hy Cahn yc, Be See ck 51 Preliminary Notes on South American Nemotognathi......... 119 Description of a New Species of Cyprinodon......... ........ 270 Gutzkow, Frederick. Determination of Bromine in Sea Water by Bractional UMtraionigen eee er) Settles acck chert occeere 81 Lindgren, Waldemar. Geology of Baja California ................. 173 Rivers, J. J, New Genus and Species of N. A. Scarabwidew.......... 100 Walmornian Tnepid@ptera.cccsss auc scue ccs ce sess tlvcea tenes 103 Trelease, William, Synoptical List of North American Species of (US) AVon A ee ee i Sa ee ae Sievele/ cavers statsneters © 106 Wolle, Francis, Desmids of the Pacific Coast....................-5 79 Proceedings. Fein rant ES EDICT ORL ICS OREO CO Sc RE ee 271 eprint Sate A ne RAR ae eae ive elie cece ne woke Seta 343 PAUTLIGHS atie COLTEOUIONSs 6) 6sc%c25<6 atic d Le deedecedudadcdccvace 361 Appendix: Catalogue of the Library ...............c0. cece ce ceeees 363 ee are. Mrmr 1 its eae ree zp a eee * #% J 1 s Pr sgrdi Ee, wed . ie om barat a is 2 J 8g F 5 ti) aS wacg lh? Pelee ae Pree doe lal ta ALatiee ety as . ei We hoe Hn pcitaee Wa nt a 4 ANTHROPOLOGIGAL SOCIETY won SW INGTON, D. C.. % as patch, PROCEEDINGS OF THE Pee FORNTA AGADEMY _ OF SCIENCES. CONTRIBUTIONS TO THE MINERALOGY OF THE PACIFIC COAST. BY W. LINDGREN, U. S. GEOL. SURVEY. 1, Chromiferous, Chlorite. The great serpentine belt crossing the North Fork of the American River, above Dutch Flat, contains several depos- its of chromite, some of which are worked on a small scale. While examining the ore from one of these occurrences— Green Valley, in the cafion of the American River below Towle’s on the Central Pacific Railroad—I found coatings of a scaly mineral of a beautiful peach blossom color, to- gether with smaller pieces of the same mineral in massive state. The substance proved to be a chlorite, and is most closely allied to that variety of clinochlore to which Kok- scharow* has given the rather harshly sounding name of kotschubeite. With a magnifying glass the scaly coatings are seen to be composed of thin hexagonal tables from ~ Bull. Ac. St. Petersburg, 369. 1861. 2p Ser., Vou. I.—1. Issued December 20, 1887. 2 CALIFORNIA ACADEMY OF SCIENCES. ©.2 mm. diameter down to the smallest dimensions, often roughly arranged in rosette form; also of smaller fibrous masses of the same mineral. Under the microscope these latter are mixed with minute grains of uwarowite and chromite. The mineral in its massive state is grown together with chromite, shows a fine parallel fibrous structure and a pale purplish color. The crystals are hexagonal tables with perfect cleavage parallel to the base, apparently bor- dered by oR, and small faces of a striated R, the sign of which cannot be ascertained; it is too small to give sig- nals in the goniometer, and the very slight thickness of the tables prevents accurate measurements with micrometer and micrometer screw. H.-=2. G. (massive variety)=2.69. Streak white. Lustre of cleavage face somewhat pearly. Under the microscope the tables, when resting on their base, are transparent and of a pale purplish color. Between crossed nicols the crystals prove to be double-refracting and biaxial. The apparently strictly hexagonal tables are twinned wonoclinic crystals; the most regular forms are usually divided in six sectors having a common apex in the center and their axes of elasticity in different position. See Fig. 1, in which the shading indicates the position of the plane of the optical axes in each sector. The colors of interference are low on the base, not exceeding the greys of the first order; those of the fibrous aggregates are more vivid, more so in fact than is usual with the chlorites. In convergent light the plane of the optical axes is found in each sector to be parallel to its base; angle of the optical axes quite large, probably about 30°. Double refraction positive. The character of the dispersion cannot be well observed on account of the small size of the crystals. The acute bisectrix stands nearly normal to the base, oP, and the extinction of the fibres is consequently quite small. Pleochroismus distinct: « & b purplish, ¢ (vibrat- ing perpendicularly to oP) yellowish red. MINERALOGY OF THE PACIFIC COAST. » 3 The division in regular sectors, as indicated by Fig. 1, is found among the crystals; but frequently the sectors are more irregular. See Figs. 2, 3 and 4. Pig 4: and is fusible on the edges with great difficulty. With fluxes strong chromium reactions. An analysis of the massive variety which Dr. W. H. Melville, of the U. 8. Geological Survey, had the kindness to make, gave: Loss at 1052 CEO) it .6 Loss above 105° (H, O).. Si or oa CxO: A Ons: 1 Or @ SeOn ae INO Ca On ve igs 5 We a MgO 0.365 % ape RS po lie .31.740 °° AIR, ef 6.741 ** 1.231 ** O:48T** Osis5"* 99.995 4 CALIFORNIA ACADEMY OF SCIENCES. The material was free from chromite.and uwarowite. To facilitate comparison, I give here one of the many analyses of kimmererite (from Dana’s Mineralogy), and y. Leuchtenberg’s analysis of kotschubeite. kammererite. kotschubeite. Texas, Pa. Ural. H, O ....18.20 % 12.63 4% Sian. BL Bie 32.67 * Cr,O .... 2.98 « 4.09 « Al, O,....12.84 * 13.18 « FeO .... 2.46 « 2.22, (Fe, O;) NGO" ooo? Ga O -5. 082 Mg 0 ....35.02 “ 35.65 % 99.08 160.53 The two forms of chlorite, penninite and clinochlore (ripi- dolite in Dana’s Mineralogy) have, according to Rammels- berg, substantially the same composition, and differ only in their crystallization, penninite being hexagonal, and clino- chlore monoclinic. Both have varieties rich in chromium: that corresponding to penninite being kammererite, occur- ring in the Urals, and at the well known locality of Texas, Pa. It has been proved to be uniaxial by Descloizeaux * and J. P. Cook, Jry.t Various authors have also designa-. ted it as rhodophyllite, chromchlorite, rhodochrom. The chromiferous variety of clinochlore kotschubeite, from the Urals, has been described by N. v. Ueuchtenberg,} but has since then, as far as I am aware, not been noticed from any other part of the world, and its occurrence in Califor- nia is therefore not without interest. v. Leuchtenberg arrives at the conclusion that kotschu- beite is a variety of clinochlore, but differing from it in the “Mineralogie. T. I. TAmdaSc. alike SDV. pe 20s { Bull. Soe. Imp. de St. Petersburg. XIII, 34. 1869. ~ MINERALOGY OF THE PACIFIC COAST. 3) inclination of the acute bisectrix to the normal of the base; while this value amounts to 12°-16° in clinochlor, it has been measured to 1°-2° in kotschubeite. v. Leuchten- berg’s description applies entirely to the California occur- rence as to the optical characters, but the crystal form differs. The mineral from Ural had the form of hexagonal pyramids, and was apparently composed of simple crystals, while that here described occurs as thin tables composed of three individuals, twinned according to the well known law of clinochlore, so as to appear as hexagonal crystals. In the position of the optical axis the crystals most resemble those of clinochlore from Texas, Pa., described by Cook. A further and notable difference from Kokscharow’s kotschubeite, and indeed from any known chromiferous chlorite, is in the very high percentage of Cr, O;: v. Leuch- tenberg found 4.09 %, while the highest percentage of of Cr. O3, in kimmereite is 5.50 %. The analysis of the California mineral shows more than twice that amount, viz.: 11.392 per cent, Cr, O, replac- ing Al, O3, but in other respects agrees well with those of kimmererite and kotschubeite given above for com- parison. Nearly half a per cent. of Ni,O also enters into the composition of the mineral; it is similar in this respect to kimmererite, which often also contains a small amount of this metal. In the collection of the State Mineralogical Museum, there is one specimen of chromite from near Jackson, Ama- dor County, covered by a thin coat of massive, peach blos- som colored chromiferous chlorite. Whether it is kiimmer- erite or kotschubeite, is difticult to decide. 2. Uwarowite. Together with the kotschubeite, mixed with it as small grains or lining small fissures in chromite as almost micro- scopic crystals with brilliant faces, there occurs a deep em- erald-green garnet. Under the microscope the crystals 6 CALIFORNIA ACADEMY OF SCIENCES. prove to be almost perfect dodekahedrons. With fluxes, strong chromium reaction. H. above 6. Refraction very strong. Between crossed nicols some grains are isotropic, but the largest number faintly double refracting, some showing an approximate division in sectors. Garnets, as well known, often present abnormal phenomena of double refraction, and, according to Rosenbusch,* the uwarowite always shows these optical anomalies. Uwarowite is known from New Idria, California, occurring on chromite. (See Dana’s Mineralogy. ) A green chromium mineral, also on chromite, from the vicinity of Livermore, Cal., given to me as trautwinite, also proved to be uwarowite, in microcrystalline form. Trautwinite + is a mineral associated with chromite from Monterey County. In chemical composition it approaches uwarowite somewhat, but differs, according to Gold- smith, in its crystal form, it being hexagonal; it is, more- over, very soft, while uwarowite has a hardness approach- ing 7. 3. Scorodite. At Steamboat Springs, Nevada, in the metamorphic series, metalliferous veins occur with arsenopyrites among other minerals. On this, in cavities and cracks, coatings of a leek green scorodite are found, which, under the microscope, prove to be often perfect crystals with the usual combina- tion of pyramid and pinacoids (111.010.100). Refrac- tion and double refraction very strong. This mineral has recently been found by Professor A. H. Ches- ter, at the Hornsilver mine,{ Utah, and as deposit from ar- seniferous thermal waters in the Yellowstone Park, by Mr. A. Hague,? of the U. S. Geological Survey. * Microskopische Physiographie der Mineralien, 2d ed., p. 269. +E. Goldsmith. Proc. Ac. Philad., 1873. 9. 348, 365. tAm. J.Sc., April, 1887. § Am. J. Sc., September, 1887. UNUSUAL NESTING SITES. T UNUSUAL NESTING SITES. II. BY WALTER E. BRYANT. Read December 5, 1887. The entire material, with one exception, which comprises: the present paper, has been received in brief notes or dic- tations from Messrs. W. Otto Emerson, A. M. Ingersoll and Chas. W. Knox, leaving the part taken by the author simply that of editor and compiler. The initials following the cases cited are those of the observers, to whom my thanks. are due for communicating their interesting field observa- tions. Tyrannus verticalis. ARKANSAS FriycatcHer.—A nest was found built upon a fence-post more than half a mile from the nearest tree. It was secured from observation on one side by a board nailed to the post and projecting above it. (A. M. I.) Sayornis nigricans. Brack Pua@pe.—A pair built for two consecutive years in a well four feet below the surface. The first year a second. nest was built after the first had been taken. (W. O. E.) Epidonax difficilis. Batrp’s FLycATcHErR.—A nest was built at the bottom of a hole five inches deep, made by a red-shafted flicker in a live oak. (A. M. I.) Cyanocitta stelleri frontalis. BLUE-FRONTED JAY.—A strange departure from the usual habits of jays was noticed in Placer County, Cal., where they had persisted in building within the show-sheds in spite of the noise and smoke of passing trains. The de- struction of their nests by the men employed on the water 2p SER., Vou.I. Issued December 20, 1887. 8 CALIFORNIA ACADEMY OF SCIENCES. train, which makes two trips a week through the sheds during the summer, sprinkling the woodwork and tearing down the nests of jays and robins with a hook attached to a pole, seemed not to discourage them. So accustomed do the jays become to the passing of trains, that they will _often remain on their nests undisturbed. In one season more than two hundred nests of jays and robins were destroyed, so the train men say, between Cisco and Summit, a distance of thirteen miles. Some of the nests were but partially built, others contained eggs; these latter ones having probably been overlooked on previous trips. The nesting of the jays within the snow-sheds is, so Mr. Ingersoll supposes, to avoid the persecution of squirrels. None, he thinks, however, succeed in rearing a brood, for of more than thirty nests which he found, nearly all were uncompleted. (A. M. I.) Spinus tristis. AMERICAN GOLDFINCH.—In 1884, a grove of young willows that had been occupied the previous season by a colony of tricolored blackbirds, was found deserted by them. Many of the blackbirds’ nests still remained in forks of the wil- lows from four to ten feet above the marsh. Six of these old nests were in possession of American goldfinches. The present tenants had loosely filled the nests about one-half full of cat-tail down, and had formed only a slight hollow for the nest proper. Some were found with eggs, and in others there were ‘‘ birds in last year’s nests.” (A. M. I.) Melospiza fasciata samuelis. SAMUEL’s Sone Sparrow.—A nest containing three eggs was found ina round oyster can, which had lodged side- ways among some driftwood in a willow tree. (W. O. E.) Pipilo fuscus crissalis. CaLiForYIA TowHEE.—A pair constructed a nest in a five- gallon kerosene oil-can that lay on its side in a shallow UNUSUAL NESTING SITES. 9 ditch. Part of one end of the can had been cut open, giv- ing access to the birds. (W. O. E.) Chelidon erythrogaster. Barn Swattow.—A kind-hearted postmaster in the coun- try nailed a shelf-like board against the porch above the entrance to his office, intending to give the crimson house finches a place to build. A pair of barn swallows took possession of this arrangement and built on top of it a nest composed of straw and feathers. This is the only instance I have known where this species used no mud in the com- position of its nest. The position of this nest was less remarkable than the peculiarity of its structure. (A. M. I.) A barn swallow’s nest was built a few feet below the sur- face of a well which was in daily use, water being raised by means of a windlass and bucket. The weight of the growing young became so great that it broke the nest from the moist ground, and the young were drowned. A second nest was speedily begun upon a shelf of rock, nearly thirty feet below the surface, and not high above the water. Unfor- tunately, the result of this second attempt was not learned, for it would bé exceedingly interesting to know how, if at all, the young were brought to the surface from so great a depth. (C. W. K.) The nesting of another pair of these swallows was illus- trative as much of persistency in nest building as it was of the unusual site which they eventually chose, prompted by repeated molestation. Three nests were built in succession; the first, containing five eggs, was taken from a partially abandoned mining tunnel, ten feet from the entrance; later, a second nest and five eggs was found, and taken nearly twenty feet from the entrance of the same tunnel. On sub- sequent visits a swallow would fly out as soon as Mr. Knox entered the tunnel, but the third nest was happily not dis- covered until the eggs had hatched. This nest was about fifty feet from the entrance, and under cover of partial dark- 10 CALIFORNIA ACADEMY OF SCIENCES. ness the persevering pair had built and reared a brood. The finding of the last nest happened by chance. Mr. Knox had descended a shaft connected with the tunnel and was pass- ing along the level with a lighted candle when he saw a bird fly from close before him, and aided by the light which he carried, the nest, with four large young, was found, but left undisturbed. (C. W. K.) Tachycineta bicolor. TREE SwaLLtow.—A few years ago I found a nest with young in a crevice under the projecting and decayed deck of a lumber lighter, moored in Oakland harbor. Vireo huttoni. Hurton’s VireEo.—A pair of vireos built this year in the outer branches of a live oak, only a few feet above the ex- haust pipe from a steam pump, where at times they were compelled to suspend work, owing to the dense vapor which enveloped them. Four eggs were laid in thisnest. (C. W. K.) Cistothorus palustris. LoNnG-BILLED Marsh WRrREN.—A conspicuous nest, con- taining eggs, was woven among the almost leafless branches of a young willow, five feet above afresh water marsh. The false nests were built as usual, but in the coarse grass near Dy p(eAn Mi. Te) WEST COAST PULMONATA. Hh WEST COAST PULMONATA; FOSSIL AND LIVING. BY J. G. COOPER, M. D. (Continued from Bull. 8, page 514.) C.—INTRODUCED SPECIES. Zonites cultellatus Thomson. (See Bull. 4, 246 ) Mr. J. H. Thomson, now of New Bedford, Mass., writes on this species as follows: ‘Tn relation tomy mistaking a specimen of H. mormonum for one of Z. cultellatus, I had, at the time of its discovery, never seen any specimen of the former, or any species found in California except those around San Francisco Bay. I went to the ‘southern mines’ in 1849, but got no land-shells there. In 1852 I lived on San Pablo Creek, 45 miles north of Redwood Peak. I found Z. cultellatus living in consider- able abundance on the road from Oakland to the Redwoods, in a springy little valley.” This locality, as marked on a map by Mr. Thomson, is 2} miles northwest of the summit of Redwood Peak, at the head of a branch of Temescal Creek, and on the zone of calcareous tufa mentioned in Bull. 7, 373. ‘Afterwards, several squatters setiled there, and began raising hogs, poultry, etc. Next year I could not find any shells there; the hogs had eaten them. On my second visit to California, in 1854, [found the shell described by Mr. Binney as H. anachoreta, and three other kinds, on the east slope of the Contra Costa hills, near my house. I have lately received a specimen of H. mormonum var. circumcarinata Stearns, and was struck with the great re- semblance to my Z. cultellatus, only that this was smoother. I have not seen the latter for some years, as it is in the New York Museum.” 2D SER., VOL. I. Issued Dec. 31, 1887. 12 CALIFORNIA ACADEMY OF SCIENCES. Mr. Binney also writes: ‘‘Thomson’s shell was of the European group of Z. acies,” found about the Mediterranean Sea. Mr. Thomson, having found no native species that could be confounded with it, and being so exact as to local- ity, as well as the cause of its extinction there, we must be- lieve that the original stock was brought alive from Southern Europe by ship, about 1850, and was placed where it would probably become abundant, with a view of supplying the market in future, just as P. pomatia and other species have been introduced into many parts of the world for food supply; also, P. aspersa, at Santa Barbara and the Sandwich Islands. Though the location was well chosen, there was not sufficient shelter to protect them from the hogs. As almost any large species may be introduced, by settlers, in this way, we may suppose that the single shell ‘of the Mexican P. buffoniana, found in 1857 by Mr. Holder, on the bay shore at Alameda, was a relic of such an at- tempted colonization. The strange occurrence of Orthalicus zebra on Vancouver's Island, alive, as reported by Mr. Lord, may be another instance. (See P. P. Carpenter’s Report on Mollusea of the West Coast of North America, 1863, p. 607.) As accidentally introduced alive, I may refer to the spec- imen of Athoracophorus found in a bale of the ‘‘ Pulu”’ fern broughtfrom the Sandwich Islands for mattress making. (See Proc. Cal. Acad., v. 195, Nov., 1871.) I also received last year, through Dr. Harkness, a very young Bulimoid shell, living, found adhering to dried plants ina herbarium from Panama. If it had been the warm season, it might have sur- vived and grown in a moist garden or greenhouse, like several other introduced species. ‘Ophiogyra heligmoidea D’Orbigny. (See Bull. 4, 248 ) Mr. H. Moores, of Columbus, Ohio, writes about this shell: ‘‘The specimen was found just as stated, and was WEST COAST’ PULMONATA. 1 (e) kept separate on account of its peculiar form. There was no misplacing of labels or any other mistake about it. I hunted near there a year without finding another. I have tried for many years, in various ways, to have somebody interest himself regarding it.” Specimens of the new Helicodiscus(?) (No. 23, Bali. 7, 367) have been supposed to be young of the above; but, though resembling its inner whorls, cannot be the same species. The discovery of fossil examples of Gonosloma yalesii five hundred miles north of its present range suggests that these sub-fossil shells of a tropical species may be re- mains of a former more northern extension. Pomatia aspersa Miiller. Mr. Binney has received this shell from San Jose, Santa Clara county, where it was doubtless introduced for cultt- vation. ‘The same species was reported from Santa Barbara in 1850, but may have been taken at the Sandwich Islands, where if was introduced by sailors. (See Amer. Jour. of Conch, vy. 211, 1870.) Limax agrestis Mimmns ss (wll aGiaeNow 2.) Living specimens, collected by Mr. W. J. Raymond in his garden in Oakland, were sent to Mr. Binney for com- parison with those now so common in the Atlantic States, where they were introduced with plants from Europe, and he considers them undoubtedly that species, probably brought here with roots of plants. Itis thus likely to be- come a pest to gardeners, like L. hewstond. On the subject of the introduction of the latter, Mr. Bin- ney was formerly doubtful, and thought it might be native, as he had a similar form from Southern and Lower Califor- nia. But on comparing the alcoholic specimens again, he admits that they are not Z. hewstoni, but either new, or L. agrestis. ‘This cannot be decided without more perfect examples. 14 CALIFORNIA ACADEMY OF SCIENCES. Hyalina cellaria Miiller. (Bull. 7, 367, No. 16.) This species, which follows commerce around the world, was found by Mr. W. Sutton numerous in a garden near the centre of San Francisco. The locality produced very large specimens, one measuring 0.55 inch inwidth. It may also spread with roots of plants, hke many other species, and has already been reported by Mr. Binney from Port- land, Or., being sometimes carried on ships’ water-casks, and the eggs thus reaching shore. D.—ADDITIONS TO CALIFORNIA SPECIES. Hyalina subrupicola Dall. (Bull. 4, 254.) One scmewhat weathered specimen was found by me at Alta, Placer County, at about 3,600 feet elevation, and hay- ing sent it to Mr. Dall for identification, he writes that it is his species, ‘‘larger, somewhat, than the largest I had.” Mr. Binney also examined it, and considers it ‘‘ certainly not indentata,” with which he had before combined Mr. Dall’s types of the species from Utah. It was before doubt- fully reported from caves of Calaveras County, Cal., as well as the cave in Utah where first found, but my dead speci- men was from driftwood by a mountain brook, and not near any cave. Hyalina binneyana Morse. The shells from Vancouver's Island, mentioned in Bull. 4, page 253, as near ZH. viridula, were compared by Mr. Dall with the types of above, and found identical. It is thus first reported from this coast, and very likely to occur in the mountains of California, like most of the small boreal species. Mr. Binney reports H. viridwla as found at Port- land, Oregon, and in Utah. (Bull. Mus. Comp. Zool., xiii, 2, p. 42. ‘‘A 2d Suppl. to 5th Vol. of Terr. Moll.,” 1886.) | Selenites ceelata Mazyck. (Bull. 7, 367, No. 15.) I have given this species in the list on page 367 (Bull. WEST COAST PULMONATA. 15 No. 7) as Mesomphix duranti ceelata, as I consider it merely a sub-species, or perhaps only a variety of that species, differing chiefly in a thicker and coarser shell, caused by more abundant lime and moisture in the localities it in- habits on the mainland, than on the islands where the type was discovered. Mr. Binney agrees in this opinion. Succinea avara Say. Mr. Binney (2d Suppl., p. 46), reports this eastern species also from California, but the locality is not given. It oc- curs in some regions east of the Sierra Nevada, but is not known yet on their west slope. Pupa arizonensis Gabb. (P. Vertigo) Ovata Say. These two species, referred to in Bull. 7, page 361, are also additions, the former only found on the east slope of the Sierra; the latter, a common eastern species, apparently strageling west to near San Diego. E.—NeEw Notes on NATIVE SPECIES. Limacoids. Referring now to the table of species given in Bull. 7. p. 367, I may further explain the reasons for grouping them, as there done, in different order from that adopted by the latest classifiers, who have so thoroughly investigated the internal anatomy of these animals. Their external characters, besides being those most easily recognized, are also those by which they are brought under the effects of surrounding influences, and thus they become naturally grouped into series, of which the members ex- hibit similar relations to the laws of nature, independent to a great extent of their internal structure. Although, not claiming that this similarity in habits and appearance constitutes a basis for scientific grouping quite independent of their organization, it becomes necessary to use it in giving their history. 16 CALIFORNIA ACADEMY OF SCIENCES. Thus, all the Limacoid species agree in absence of exter- nal shell, and therefore, while without that slight protec- tion, can better escape their enemies, as well as the effect of droughts, fires, and cold, by crawling into fissures that the others cannot enter. They also suffer less by being washed down in the winter torrents, and follow retreating moisture as the streams dry up; so that some of Nos. 1, 4, 5, 6, 7, 8, 9, 10 may be found active all summer in damp cahons or spring-heads, along weedy brooks and edges of ponds. Near the bay Nos. 6, 7, 8 are, however, scarce after April, and Nos. 9, 10 after June. Limax (Amalia) hewstoniJ.G.C. (No. 3.) There seems reason to believe that this species is of only annual existence, which may be the case also with the other Limaces, though I-have not seen it published. They dis- appear with the first hot, dry weather, and are then found for a while in the burrows of animals a foot or two deep, where their eggs are deposited (also near the surface about wells and cisterns), but after July none can be found even in excavations five or six feet deep. After the ground is well soaked with rain in late autumn they reappear in num- bers, but very few more than half grown, some of late broods, perhaps surviving in the wettest spots. As none are ever found far from gardens, they are absent where the native species survive in the dry season. The variability in color among these shell-less mollusca is well known, but its origin in the principle of self-pro- tection by mimetic accommodation has not been much re- marked upon. In this species, and some others which feed chiefly at night and in cloudy wet weather, the blackish tints chiefly prevail, and seem to deepen after they first come to the surface, but specimens of ZL. campestris and L. agrestisare often found of light shades or with streaks re- sembling the nervations of dead leaves, among which they creep in the daytime. In the large kinds of Ariolimax, Nos. 4 and 5, yellowish WEST COAST PULMONATA. Lyi is also the prevailing color, like that of many dead forest- tree leaves, and they are often blotched irregularly with black, like leaves mouldy or decayed. The small species and sub-species, A. niger, andersoni and hemphilli, are either black, or pale with dark specks irregularly scattered on or between the granulations, while the furrows separating these are of the same light color. In Prophysaon andersoni and hemphilli, the furrows are darker than the tubercles, giving the ‘‘ foliated” appear- ance ascribed by Dr. Gould to an Oregon ‘‘ Arion’”’ (which was probably of this more recent genus). A black variety of the former was also found by W. O. Emerson on the Santa Cruz Mountains among burnt logs, where its color might have also been a protective variation. My reason for making Nos. 7 and 8 only sub-species of No. 6 (4. niger), besides the local limitations before men- tioned, are that they do not seem to differ from that species more than the forms of L. campestris called montanus and occidentalis differ from that species, and Mr. Binney now admits that they intergrade with it. T am inclined to think that all the Limacoids except Limax are biennial, taking most of two years for their growth, and then dying, but some may live longer, especially those hatched late in the season. Mr. Binney considers Prophysaon andersoni as ‘‘ probably a distinct species,” founding this opinion on the genital organs, but the differences in form noted are such as are likely to vary with season, and to change after oviposition. In either case, the specific name is prior to that of P. hemp- hilli, although I described it as an ‘“‘ Arion” before the later genus was separated, noticing its resemblance to A. foliolatus Gould. Vitrinoid. The distinctions separating this group from the Helicoid may be broadly stated as the absence of thickened expanded 2D SER., Vol. I. 2 18 CALIFORNIA ACADEMY OF SCIENCES. lip to the shell, as well as thinner and plainer shells, though these characters differ? from those derived from internal organs. In our native species, however, they serve very well to separate the two groups. Mesomphix (vancouverensis?) simplicilabris Ancey, (No. 13.) Mr. Binney, in 2d Supp., p. 41, calls this a variety of J. voyana, but unless confounded with immature shells of that form, it seems as nearly connected with the above, the shell as found near this bay having the smooth polished surface and undeflected lip of the young of the larger forms, with the small size and wider umbilicus of the small var. of, I. voyana. That itis nota hybrid is shown by its not being found with either of the others, and it is very rare any- where, appearing to be either a survival of an ancestral form, or a proof that JZ voyana is a derivation from the larger forms. Mr. Binney writes about specimens I sent him: ‘«“The lingual is like that of JZ. vancouverensis.” He also remarks on the similarity of the shell to his new form, ‘‘ J. hemphilli,” from Washington Territory; but that is much larger, and with a contracted umbilicus. As M. voyana also intergrades with JL sportella, it seems as if all these forms must yet be considered as only subspecies, like those of some other west slope species. Microphysa conspecta Bland. (No. 21.) The unexpected appearance of these little delicate species in cultivated grounds is shown by the finding of this shell for the first time east of San Francisco Bay in 1886, at Hay- wards. A hanging basket, planted with ferns and covered with mossy bark brought from a cafion near by, was hung in my garden, and some months later I looked under the bark to see if any molluscans had remained init. I found the above species abundant, and in a few weeks took out over one hundred, always finding more, still immature. After being unwatered for four months, and becoming quite dry, they revived on being moistened. WEST COAST PULMONATA. 19 Mr. Binney writes that he once found a similar colony of Hyalina exiqua in a fern garden in Boston. The resem- blance of these two species externally is very great, but on account of differences in jaw and teeth, the former is put lately with the Helicide. The similarity of habits now shown, in addition to their similar shells, is a strong argu- ment for retaining both in one group. Mr. Raymond has also found it in a garden in Oakland. Helicodiscus lineatus Say. (No. 22.) In the ‘‘ Manual of American Land-shells,” Mr. Binney puts this among the ‘‘universally distributed species,” (a division which might better be named circumboreal or boreal for those only North American), since many are not found far south of latitude 49°, or only along mountain ranges. He remarks that the specimens found by Hemphill at Oakland, Cal., and in Idaho, are without the colored lines from which the species was named. In this they re- semble the undescribed form (No. 23), but unlike that, probably have internal teeth. This is a reason for deferring the descriptionyor naming of the latter until living speci- mens can be obtained. I have not heard of the former having been obtained by anyone else in California. Helicoids. Mesodon (4plodon) armigerus Ancey. (No. 25.) The recent separation of this form from the Oregonian M. columbianus, is fully elucidated in Binney’s 2d Supple- ment, with illustrations. Considering, however, the occur- rence of another variety in Plumas County (Bull. 7, 358), and that no specimens from the northwestern part of Cali- fornia have yet been compared with either of them, it seems probable that all the later forms will yet be found to inter- grade with that first described, reducing them to subspe- cies. 20 CALIFORNIA ACADEMY OF SCIENCES. Arionta arrosa Gould. (No. 26.) I have lately seen several more specimens of the small form mentioned in Bull. 7, p. 372, labeled ‘“4. arrosa, Ala- meda Co.” by the finder, and probably from near the same locality. It is therefore not a chance straggler there, though rare and local only in or near ‘‘ Redwood Cafion,” just as it is found near the redwoods west and north of the bay. The size is less than that of coast specimens, but larger than some from Napa County, and though about equal to some of the varieties of A. californiensis, differs in seven whorls and other characters. Connecting links between the two are not found yet, though var. holderiana east of the bay comes very near some of the forms of var.? exarata north of it. A. (californiensis) ram n sa Gould. (No. 32.) This form shows an ability to withstand droughts and heat, great than any other of the large banded species of this coas’ On the east shore of San Francisco Bay it is sometimes found in colonies along the sides of little gullies, washed out by springs which almost entirely dry up in sum- mer, and where no rocks or trees shelter them, the largest vegetation being a c wse grass about five feet high, and annual herbage. Thc may be found in these stations tor- pid during about four dry months, partly concealed in slight cavities, and are remarkable for thickness of shell, derived from the fossiliferous soil of the pliocene terraces jrom which the springs flow. They are not found along the permanent streams near by, where rocks and shade abound, and where the thin form No °0 is rather common. A. anachoreta Binney. In the synopsis (Proc. Cal. Acad., iii, 338.), I classed this bandless shell as perhaps identical with a form found near lat. 42°, and still of uncertain specific standing; but since I have collected near the place where Thomson found the WEST COAST PULMONATA. at type specimen, and for fifty miles or more around there, I am satisfied that it is merely a bandless (perhaps diseased) specimen of one of the large species found there. Mr. Binney remarked its resemblance to A. nickliniana. It is perhaps as near var. bridgesii, which is the prevailing form where it lived, but may have been of var. holderiana. Simi- lar accidental defects being found in all the banded spe- cies, it cannot be considered a subspecies, and scarcely a variety. The same may be said of the form called nicklini- ana Lea., a very uncertain type, while var. bridgesii is well defined. Bandless specimens of other forms are sometimes found. Mr. Ancey, who is inclined to divide both genera and species too much, has recently made new names for the sroups here included in Arionta and Campylea, viz., Hel- minthoglypta Ancey, type A. tudiculata Binney, which he says differs much from Arionta (arbustorun) internally, and Micrarionta Ancey, types C.? rujicincta, C.? gabbi and C.? facta, which I consider of the same group as C.? traskii, C.? fidelis, etc., though he does not include these. From his remarks on these two genera, I conclude that he retains the latter in Lystnoz (but still uses the generic name AGLAIA), giving the two new genera as the parallel series to Arionta and Campylea of Europe. A. exarata (Pfeiffer). (No. 35.) It was intended to have the name of this shell in the col- umn of species, like Nos. 26 and 30, but the printers put it with the subspecies. Should it yet prove to intergrade with either of the two, it will be with 26, not with 30. C. 2 (fidelis?) infumata Gould. (No. 36.) In his ‘‘ 1st Suppl. to Terr. Moll. of U. S.,” Mr. Binney has figured one of the links between the two so-called species here combined as ‘‘ the smooth form of infumata,”’ but does not mention the many intermediate gradations be- tween the two, which perhaps he has not seen. It may, pe) CALIFORNIA ACADEMY OF SCIENCES. however, be as well to continue to regard the original types of west-coast forms as distinct species, the intermediate links being generally confined to narrow geographical limits. It is a still unsettled biological problem whether different species of these low orders of animals may not originate new species by hybridization, some of the progeny being fertile, and perpetuated by natural selection. Mr. Binney, (‘‘Man. of Amer. Land Shells,” pp. 122-3, 1885) con- siders the two as identical, or varieties of one species, but they exhibit such a wide difference in their extreme forms as is not seen in any other known species, and must rank at least as sub-species. O.? infumata presents a curious instance of possibly pre- servative color-variation, as it commences to appear just south of the Oregon boundary, where a drier climate makes. fires more frequent, and from its color is less easily seen by enemies among charred logs and leaves, while its lower, angled form, enables it to crawl under logs or into fissures of rocks, where C.? fidelis cannot thus protect itself. Thus it exists with the depressed forms, 11 and 37, as far east as Solano County. The same may be observed of the angled and hirsute race of C.? mor- monum, called hillebrandi, found in the Sierra Nevada be- tween lat. 37° and lat. 38°. A black variety of C. ? sequoicola has also been found in the Santa Cruz Mountains, where fires are so frequent and destructive, but no angled form of any kind is yet known in the southern Coast Range, al- though a fossil of that shape occurs on Santa Barbara Tsland, which I have referred to the living species A. ? tryona, now only found rounded. (See Proc. Cal. Acad. VI., 17, 1875.) It is not uncommon to find colonies of some of the large Helicoids killed by fire, a slight scorching of the shell being sufficient to kill them. It is thus evident that in the drier localities farthest from the coast where fires are likely to WEST COAST PULMONATA. aS spread most widely, the angled, or even the depressed shells, being best suited to crawl into the deep fissures of the earth, rocks, or under logs, must oftenest escape burning, and thus become the prevailing forms stocking those regions. Thus may be explained the distribution of such forms as A. (cal.) ramentosa (No. 32), which prevails over most of the two counties east of San Francisco Bay, while A. exarata takes the place of A. arrosa in most of Santa Clara valley, but is less common west of the Santa Cruz Mountains. An approach to the angled form is sometimes seen in J. armigerus (No. 25), but as that species only lives in very damp places, this variety may be of different origin. The imperforate and toothed variety is given as Ancey’s type, but the umbilicated toothless form is the most common, being very near that found in Plumas County. Mr. Badger has lately brought from Eel River, Humboldt County, near lat. 40°, and perhaps 1,000 feet elevation, the largest specimens of C. ? infumata Lever saw. One belongs to Binney’s smooth variety, but is nearly destitute of angle, and though very dark has a darker line on the body whorl. and traces of the impressed revolving grooves of jfidelis. It measures in breadth 1.74 inch, axis 0.76, alt. 0.90, being of about three times the bulk of Alameda County shells, and with only traces of their peculiar epidermis, but is a dead shell somewhat worn. Itis the largest land-shell I have seen from the west coast of the United States. A smaller, less worn shell, from the same place, is more angled, but the surface even more like that of jidelis, shining beneath, but band very faint. They might in- deed be almost called a black variety of fidelis, and are truly intermediate. The largest has about the width of the ereat G. newberryana figured in Binney’s 1st Suppl., Pl. IV., but is higher and heavier. With them are some of A. arrosa, very large and approaching the form I called arboretorum Val., also one M. vancouverensis nearer the Oregon shell than 24 CALIFORNIA ACADEMY OF SCIENCES. those southward. These show the effect of the climate to- ward the northwest, and indicate a promise of new forms in that direction. Species are known to extend across the country for 150 miles inland, at short intervals, near lat. 40° to 41°, but are still very imperfectly known. A Portland, Oregon, paper reports that the large species of that vicinity, C.? fidelis, is sold there for food. BIRDS FROM THE FARALLON ISLANDS. 25 BIRDS AND EGGS FROM THE FARALLON ISLANDS. BY WALTER E. BRYANT. Read December 19, 1887. Situated about thirty-five miles west from San Francisco, California, is a cluster, or rather three clusters, of rocky islands commonly known as the Farallones. Upon the charts, these islands are individually designated as North, Middle and South Farallones. The latter island is the largest, and the only one inhabi- ted; all the sea-birds that nest on the others are also found upon the South Farallon. It is regarding this island and the adjacent rocks that this paper will treat. The islands have been occasionally visited by naturalists, but their stay has usually been of short duration, a few hours oftentimes; or, if they stayed for days, the time has been largely devoted to the accumulation and care of spec- imens rather than to the study of the birds which congre- gate in countless numbers during the breeding season. Mr. W. Otto Emerson, who visited the South Farallon from June 14th to July 2d, 1885, and again from May 2d to June 2d, 1887, has generously placed with me his match- less collection of birds and eggs and his notes regarding them, for publication. Mrs. W. H. Rugg, wife of the chief hight-keeper, has contributed interesting notes and data concerning many of the birds, particularly strageling species. The first impression of the island, Mr. Emerson tells me, is that it is low, rounded and knoll-shaped; but on ap- proaching nearer, it discloses its high, rough and rocky character. By the time the anchorage at Fisherman’s Bay is reached, the island and adjacent rocks are seen to be fairly alive with birds, whose clamor and rushing wings, as 2D SER., VOL. I. Issued January 19, 1888. 26 CALIFORNIA ACADEMY OF SCIENCES. they rise—alarmed by the boat’s whistle—can be compared only to a swarm of bees ona grand scale, at the time of leaving the hive and undecided what direction to take. Arch Rock, Sugar Loaf and East End seem, in places, to be covered with snow, as the sunlight falls upon the white breasts of the murres. The South Farallon extends about a mile in length from east to west, and is nearly half as wide as long. It is of granite formation, with a broken ridge extending length- wise and interrupted, in places, by precipitous bluffs and ragged, rocky points. At the west end is the highest bluff, Indian Head, one hundred and five feet above the sea-level. From there a fine view can be had of part of Breaker’s Bay, and along the whole side of the main island. The shore-line is very uneven, long and narrow fissures running in from the sea and often extending under the rocks. One of these under passages has an opening on the island, and is supplied with a fog-horn which is blown by the rush of air driven through by the waves. From the summit of Tower Hill, the most elevated point, 343 feet above the sea, the North Farallones, seven miles away, can be plainly seen, with Middle Rocks lying between—a small group, one hundred and sixty-six feet at the greatest height. They are seldom visited, being dan- gerous to land upon. No shrubs of any kind grow on the island. The princi- pal vegetation consists of the Farallon weed (Beria mari- tima). This plant, and also alfileria, and the roots of Lep- igonum macrothecum are eaten by the rabbits which inhabit the island. They were introduced from England by a sea captain, who brought over a few as a present to a former light- keeper; and they multiplied until there was not sufficient food to support them, many dying of starvation during a dry season. At times, they furnish the only fresh meat which the light-keepers and their families have during BIRDS FROM THE FARALLON ISLANDS. 27 weeks or months of stormy weather. Other plants have been introduced in the hay brought for ‘‘ Jerry,” the island mule, whose duty it is to haul the car over the tramway from the landing to the store-house. The only fresh water to be had on the island is that which is caught during the rainy season and saved in cisterns. A few springs near the shore line are too strongly charged with guano to admit of their being used for drink. Birds rarely or never fly against the light-tower, but oc- casionally strike the bell-wires running from the house to the tower. The abundance of the breeding water-birds is estimated to be in the following order: 1. Murre. 2. Western Gull. 3. Cormorants (including three kinds). 4. Pigeon Guil- lemot. 5. Tufted Puffin. 6. Cassin’s Auklet. 7. Ashy Petrel. It is, of course, understood that the land birds of this catalogue are, in most cases, to be regarded as strag- glers either driven off shore by high winds, astray in heavy fogs, or resting on their migrations to the north or south- ward. During the bird-wave in May, 1886, the weather was un- usually fair. No stragglers were noticed by Mr. Emerson in 1885. Mrs. Rugg saw a few that year. 1. Lunda cirrhata. TurreD PurFin.—Arriving at the island in pairs during the latter part of March, they associate in pairs until their single egg is laid, after which the one not sitting re- mains near the entrance to the burrow; the presence of this sentinel—which may be either male or female, as both birds take part in incubation—indicates a nest with a bird sitting. They burrow when possible; otherwise, make use of nat- ural cavities anywhere on the island for a nest. Sometimes _a few pieces of weeds are carried in, but often no material is used in the nest. 28 CALIFORNIA ACADEMY OF SCIENCES. It is always safe to wear a heavy glove when taking a bird from its nest. They will strike a few times at one’s hand, and then retreat, if there is room; but if not, they become angered and bite viciously. One caught Mr. Em- erson by a finger and bit it to the bone, holding on with bull-dog tenacity until killed. When alighting, they hold the body and head low and then straighten up, adjust the wings and draw the head back proudly. On the rocks, or flying about, they are si- lent, but when two are in the same burrow they keep up what sounds like an angry quarreling and scolding. The sentinels, if alarmed, circle about; but when a person re- mains perfectly quiet they will soon return and light with- ina few yards. Their bright and oddly-shaped bill, white eyes and yellow nuptial tufts, which fiutter in the wind, make them birds of peculiar interest to observe in life. The food of the puffins was found by dissection to con- sist mainly of small fish, together with some alge. Young, with feathers plainly appearing through the down, were taken July 26, 1886. I have carefully measured fifty puftin’s eggs, which avyer- age 70.2 mm. x 48.4mm. The individual proportions of eight eggs, showing the greatest and smallest extremes of both diameters, are 81 x 50; 77 x 48; 74 x 50; 71.5 x 51; 71 x 46; 65.5 x 45; 64 x 50; 63.5 x 50 millimeters. 2. Ptychoramphus aleuticus. Cassin’s AUKLET.—A strictly nocturnal species, flying during foggy, stormy or moonlight nights, but never at dusk. They arrive early in the year, coming in great numbers in the night of January 14,1887. The auks do not fly until it is quite dark, and are supposed not to fly very high; otherwise more, it would seem, would strike the lighthouse. One, attracted by a lantern carried by Mr. Emerson, flew with characteristic swiftness directly at it, but missed and struck against the side of a house, BIRDS FROM THE FARALLON ISLANDS. 29 when it was picked up stunned. Auks have struck persons walking without a light, but always below the shoulders. From the balcony of the light-tower they could be heard below calling; their note is the most noticeable of the night cries, seeming to predominate over all other bird sounds, especially during stormy and windy nights. On the Farallon Islands they do not burrow, but lay in natural cavities over the entire island, particularly where a pile of rocks afford concealment. The inhabited places may be known by the excrement accumulated about the entrance. They use no nest material, laying a single un- marked ege. Several young are supposed to be raised dur- ing the season. Many nests were found occupied by young in down and one adult bird sitting upon a fresh egg; in some nests the egg was kept warm by contact with the young. In no instance were two old birds found in the same nest, and no birds were found at the time search was made without an egg or young or both. The majority of adult birds taken were females, although both sexes were found sitting. If provoked, either young or old will seize afinger and hold on. The old birds are silent when on the nest, but the downy young make a faint peeping when disturbed. When taken from the nest, they endeayor to crawl out of sight, and if tossed into the air they descend quickly and hide themselves from the light. They com- menced flying this year as early as April 2d; and eggs have been found as late as November 20th, 1886, showing a breeding time extending through eight months. The eggs, when held to the light, show a pale shade of emerald green. Fifty specimens measured average 45.3 x 32.5 mm. The extreme largest and smallest eggs were found to measure respectively 49 x 35; 48.5 x 33.5; 48 x 36; 47.5 x 36; 46x 32; 43 x 32.5; 48 x 82; 42.5 x 34 millimeters. I have taken measurements of examples of both sexes, the averages of which, in millimeters, are here appended. 30 CALIFORNIA ACADEMY OF SCIENCES. Three males.—Wing, 127; tail feathers, 33.3; exposed culmen, 19.2; tarsus, 23.3; middle toe and claw, 36.2. Five females.—Wing, 127.8; tail feathers, 32.5; exposed culmen, 19.8; tarsus, 23.9; middle toe and claw, 35.1. 3. Cepphus columba. Pickton GuImLLEMoT.—This species is one of the last to arrive at the island, and is to be found in pairs after May Ist. When at rest they squat, duck-like, upon the rocks, hold- ing the bill in horizontal position; occasionally a few will rise to a standing position, with their bills considerably elevated. Previous to laying they gather in small groups, usually near the water’s edge, and when disturbed rise to the stand- ing position, and in this attitude, with open bills, they salute each other or returning fellows, by making a pevuliar, whistling cry, that can be heard at all times of the day. After pairing they fight a great deal among themselves, chasing each other on the wing and engaging in combat on land and water. Mrs. C. M. Crowell, of Haywards, Cal., witnessed a fight in the water, which continued for ten minutes; they first held fast to each other’s bills, then sepa- rating, one dived and the other swam about watching closely for the appearance of the first, which, on coming to the sur- face, was at once attacked and tormented until it finally dived and came up beyond reach of its adversary. If caught, they show a disposition to fight, but first will endeavor to escape by crawling away. They lay wherever concealment can be found, in dark caves and fissures, under boulders, and the tramway wall; most of the nests were found on flats near the water’s edge. Two eggs, the largest number found in one nest, are laid without any attempt being made at nest-bnilding. Some observers have credited them with carrying small pebbles to their nests, but as these are not always found where the BIRDS FROM THE FARALLON ISLANDS. ol eggs are laid, their presence may be due to some extent to accidental accumulation. When a bird leaves the nest, the mate at once pursues, as though trying to drive her back, chasing her along the ground, with upraised wings, mean- while making the usual hissing noise, and following her even out to sea. Upon each side of the abdomen of the specimens which I have examined are bare spots against which the eggs are held. Although noisy at most times, they are invariably silent while sitting, differing markedly from the puffins in this respect. In no place were two birds found occupying the same cavity. If their eggs are taken, they will lay again in the same place. Their food consists principally of fish, which they obtain by their expert diving. Egg laying begins early in May and continues into July. Young in down were found July 16th, 1886. Twenty-five sets of two eggs each average 61.6 x 41.4 mm, Selected specimens of the two extremes in size measure 66.5 x 41.5; 66x42; 64x43; 60x48; 60x 39.5; 59.5 x 39.5; 58.5 x 41;5; 58.5 x 41 millimeters. 4, Uria troile californica. CatirorniA Murre.—By far the most abundant species, outnumbering all others combined. They begin to arrive on the island in myriad numbers by the first of April, and commence departing about the middle of September; by the first of October they have all ieft the island but can be seen upon the water. Their arrival usually occurs at night when great numbers come suddenly, and perhaps leave the next day; especially are they likely to leave soon after coming—and before mating—if a storm occurs, returning of course later. The food of the murres consisted, as far as was learned, wholly of fish. 32 CALIFORNIA ACADEMY OF SCIENCES. ‘When on the wing the murre sometimes makes a peculiar grunting sound; on the rocks they are very noisy, bowing their heads continually. They are much more clamorous and make greater demonstrations when they are joined by others, and like all the island sea fowls, are more so before a storm. Distributed over the entire island and also the isolated rocks, and North and Middle Farallones, but in places much more abundant than in others; these densely populated rookeries are the ones worked by the eggers, who, during three months of the year, send the eggs of the murre and western gull to the San Francisco market, where they find a sale at from twelve to twenty cents a dozen. Between 1850 and 1856 there was reported to have been brought to San Francisco between three and four millions of eggs. For the last few years the number of eggs mar- keted has averaged from one hundred and eighty thousand to two hundred and twenty-eight thousand. In 1886, two men who were left on Sugar Loaf collected one hundred and eight thousand eggs. The South Farallon being the principal island, and the only one inhabited, is the one from which, with neighbor- ing islets, the entire egg crop is harvested—most of the eges coming from the Great Rookery at West End. The - business is carried on by twelve to fifteen men, who inspect the rookeries occasionally to learn when the murres have commenced laying, as the time varies with different seasons. When ready to begin picking, all the eggs that can be reached are broken, so as to insure only fresh ones being taken, for the eggers maintain that an egg sat upon for ,a day is unfit to market; the gulls willingly assist in this de- stroying work. The entire main island, or rather those parts of it which are picked, is gone over every other day. In the forenoon of the first day they take the West End, commencing with Main Top, going out Great Rookery and to Indian Head, then across the flat to Murre Bridge. In ee BIRDS FROM THE FARALLON ISLANDS. 30 the afternoon, if not too tired, they collect on Seal Rock; but if this is omitted, all the eggs there must be destroyed, and a fresh lot laid, before any can be saved for market. The second day’s picking will be made on the main part of the island, including all the rookeries at East End, also on the North Rocks and Sugar Loaf. -Picking usually com- mences on Sugar Loaf, an isolated rock 185 feet high. It is warmer there, and more protected from prevailing winds, and, being less disturbed than the main rookeries, the murres lay there first. 'This rock is reached by means of a boat, which is left in charge of one man, while four or five others begin gathering, working from below upwards. Another man starts at once to the top and collects there, meanwhile keeping the gulls away. The boat seldom reaches the Sugar Loaf unnoticed by the gulls; but when it does occur, the first to make the discovery raises the cry, when all the gulls in the vicinity rise screaming, and fol- low the men—hovering, screaming and cackling over them. The Sugar Loaf is the most difficult place to collect eggs. In several places ropes are made fast to enable the eggers to reach the most inaccessible places. Two men have lost their lives on this rock: one of them fell, last year, into the sea and was never found. He had injudiciously stowed too many eggs in the front of his shirt, and in passing a narrow Shelf the eggs actually crowded him off. Minor accidents and mishaps are of almost daily occurrence. One egger, having about ten dozen eggs in his shirt, fell and rolled a distance of twenty feet; although uninjured, he was com- pletely soaked with egg. Asa rule, the eggers get consid- erably spattered by the excrement from the birds whirling overhead. Before proceeding further it will be well to notice closely the men who engage in this nest robbing extraordinary, and the methods they employ. The eggers are Italians and Greeks, usually those who have been engaged in fishing 2D SER., VOL. I. Issued January 19, 1888, 3 34 CALIFORNIA ACADEMY OF SCIENCES. about the islands: the first party to take possession each year manages to hold their position against all comers and to even defy the United States authorities to remove them. Being trespassers, they have, on more than one occasion, been taken away, but only to return the following year. This season the party secreted themselves in Murre Cave while the revenue cutter Corwin hovered about the island for hours. living in caves or tents improvised from old sails and spars their requirements of life arefew. A cotton flour sack (100 1b. size) is made into an egg shirt by cutting out a hole in the bottom for the head and one on each side for arm holes; a gathering string is passed around the mouth of the sack which, when it is put on, is drawn tightly about the waist; a slit down the front of the shirt from the neck makes an opening for stowing the eggs, while a padding of Farallon weed inside on the bottom forms a cushion for them. Arriving on the picking ground or in a murre cave, they set to work, using both hands when possible. The birds take wing by pitching from their narrow shelves or scram- ble pell-mell out of the cave past the eggers. Highteen to twenty dozen murres eggs is considered a good load for a shirt, and about as heavy and bulky a load as a man will care to pack. Having filled the shirt the eggs are emptied into a basket to be carried to the landing. If overtaken by night, the eggs are dumped into a pile until the next day. The baskets filled with eggs, or the heaps which sometimes contain as many as one hundred and fifty to three hundred dozen, are most carefully covered with old sacks or weeds and weighted with rocks to prevent the gulls from getting at the contents. A few of these persistent thieves invaria- bly hang around the egg heaps and baskets, and if an open- ing is espied revealing to sight an egg, they pull and tug at the weeds or sacks till the egg is obtained, when the lucky bird flies away with it, followed by others whose screams soon attract a large congregation about the eggs, which in an incredibly short space of time are all taken. BIRDS FROM THE FARALLON ISLANDS. 39 When sending the eggs to San Francisco, they are simply tumbled into the fishing boat; many are thereby dented or slightly cracked, but they are seldom broken enough to in- jure their market value. At San Francisco they are boxed and taken to market. The single large egg laid by the murre is deposited at any place upon the rocks furnishing space enough for the bird to sit. Its pear shape, an all-wise provision no doubt, preventing it from rolling off of a slightly inclined plane. The egg is constantly guarded by one of the pair, for if left exposed it would be at once carried away by some watchful gull. Sometimes two or three gulls will combine to rob a murre that is separated a little from the colony; by feints and tormenting from all sides the murre, in an unguarded moment, or when striking ata gull, is certain to expose the egg, which is immediately snatched by one of the gulls, who, flying away, is pursued by the others for a share of the spoil, which they seldom get. The gull in possession of the egg retires to a convenient spot and breaks the egg, not by dropping it, for by that he would lose the contents, but by rolling it around against the rocks till itis dented sufficiently for the gull to insert his beak. The gulls pick murres’ eggs up bodily, and carry them away in their capacious mouths, but do not stick their bills into them to get hold, as stated by some writers, whose observ- ations must have referred to eggs already broken by the gulls or eggers. When the young is hatched the parents must be ever as watchful, for the gulls will steal a young murre as soon as they would an egg, and will swallow as large a young as they can get down their insatiable throats. As soon as the young can care for themselves, that is, af- ter the pin feathers are started, the murres leave the island, or if the gulls are particularly numerous or troublesome, old birds, in order to save their young, crowd them from the rocks down and into the ocean where by diving they can 36 CALIFORNIA ACADEMY OF SCIENCES. escape the gulls. If the young when crowded off a shelf by the parent falls upon another rock it is quickly followed by the old bird, who broods over it until another oppor- tunity and another shove launches the young one in the water. The parents share equally the duties of incubation, one re- maining on the egg during the day the other by night. During the day few are seen flying unless disturbed by the eggers or the report of ‘a gun, but when the relief is made morning and night they are especially noisy, quarreling and scream- ing in great masses of bird life. The young sea lions have a great fondness for murre’s egos, and as soon as they are large enough to know what an egg is, and once get the taste of them, they become another factor in the destruction of eggs. Mr. Emerson has seen young sea lions with their muzzles slabbered with egg. The old sea lions do not trouble the rookeries, but spend their time ashore basking about the water’s edge. The island mule has also found that eggs make an agree- able variation to his diet. He hunts nests very assiduously, erowing fat and sleek in the breeding season. The large series of eggs of this species in Mr. Emerson’s collection exhibits a wide range of variation in coloring, they being specimens selected from thousands. In size, fifty ordinary sized eggs average 82.9x 50.8 mm. The following measurements show the extremes of normal specimens which were examined, 92x53; 91x51; 85x54; 85x63; 79x47; 77 x47; 75x49; 75x 50.5 millimeters. Some diminutive eggs, and those noticeably of different shape or size, measured 95x 46; 87x45; 77x48; 73x 46.5; 68 x 44.5; 63.5x 37; 60x 38.5 millimeters. 6. Larus glaucescens. GLAUCOUS-WINGED GuLL.—A few were seen about the island as late as May. BIRDS FROM THE FARALLON ISLANDS. at 6, Larus occidentalis. Western Guitu.—This is the only gull which breeds on the Farallon Islands, and eggs of other species, such as California gull and American herring gull, purporting to have been collected there, have undoubtedly been wrongly iden- tified. In numbers the gulls rank about second during the summer. In winter they distribute themselves along the coast and into San Francisco Bay, congregating at the island by the first of April. The gulls are indiscriminate feeders; in addition to their usual articles of diet, they sub- sist largely upon eggs during the summer. They do not eat the eggs of their own species, nor do they trouble the cor- morants after the murres have commmenced laying. Sea urchins, crabs, voung murres and rabbits, and fish stolen from the cormorants’ nests are eaten. Not being quick enough to swoop upon the rabbits, they catch them by pa- tient watching at their burrows, and will patiently try for fifteen minutes to swallow a squealing young rabbit, and finally fly away with the hind feet protruding. The dead bodies of murres are also eaten; they detach pieces of flesh by backing away and dragging the body, meanwhile shaking their heads, till a piece breaks off. A young gull raised by Mrs. Rugg has returned to the island annually for four years; although not tame enough to take food from her hand, he comes to the house and picks up any scraps of cooked food thrown to him and flies away screaming. He will not take raw meat, and would not eat it when young. He does not appear for food oftener than once a day, and sometimes only once a week. They nest in small colonies of ten pairs or less, either on low flats or the high points. Two weeks are occupied in the rebuilding of last year’s nests; on many days they appa- rently do not work at all. Even after the first egg is laid they continue to carry fragments of dry Farallon weed to their nests. The first eggs of the season were taken May 9th, in 1885 and 1886, and May 13th, in 1887. The eggers 38 CALIFORNIA ACADEMY OF SCIENCES. collect the gulls’ eggs for the San Francisco market during the first four weeks of their laying, after which they give their whole attention to the murres. Usually three eggs constitute a setting, but from repeated robbing often only two are laid. When more than one egg is found in a nest by the eggers they are destroyed on sus- picion of having been incubated. Gulls’ eggs are collected in the same way as the murres’, but having thinner shells fewer are carried for a load in the shirt or basket. A young gull from the islands, apparently but a few days old, differs from the downy young described in ‘‘ The Water Birds of North America” (Vol. II, p. 230), in having the abdomen and sides mottled, although the marks are not as clearly defined as those of the back and head. The hundreds of gulls’ eggs which I have seen vary con- siderably in intensity of color. The average size of fifty egos, taken indiscriminately, was 70.2x49.4 mm. The largest and smallest measurements found in this series of fifty specimens were 76x51; 72x52; 72x48; 71x52; 71x46; 66x46; 65x48 millimeters. 7. Larus californicus. CaLirornia GuLL.—More or less common during the fall. 8. Larus heermanni. HEERMANN’S GULL.—Occasionally seen by Mrs. Rugg. 9. Diomedea albatrus. SHORT-TAILED ALBATROSS.—An adult male which I have not seen, but which has been identified by Mr. Emerson, is now in the possession of Mrs. Rugg. The bird followed a ship to the island and was shot by one of the men on a tow boat, March 20th, 1887. 10. Oceanodroma homochroa. AsHy Pretret.—The last to arrive on the island is this small petrel, which is also the rarest of the birds which breed there. BIRDS FROM THE FARALLON ISLANDS. 39 They nest anywhere on the island in natural cavities, particularly those under loosely piled rocks. No nest is made and only a single egg laid, although it sometimes hap- pens that an egg and downy young will be found in the same place. The egg of this petrel was first found by Mr. A. M. Ingersoll on the South Farallon, and noticed by him in the ‘‘ Ornithologist and Odlogist,’’ of February, 1886. The egg is accurately described as ‘‘dull creamy white in color, with a circle of reddish dottings so fine as to be almost in- distinct, around the large end, which is somewhat flattened like the large end of an acorn.” It measured 30x24 mm. (1.18x94 ins.) Two sets of petrels’ eggs collected by Mr. Emerson, June 15th, 1885, and June 8th, 1886, measure re- spectively 30x25 and 31x23.5 millimeters. A young bird which was taken from the nest June 11th, 1886, has a heavy covering of light slaty down on the un- derparts, across the interscapulars and on the head. With the exception of the parts mentioned the bird is feathered. The tips of wing coverts are light gray. 11, Phalacrocorax dilophus albociliatus. FaraLton Cormorant.—Although not the most abundant of the cormorants they gather in the largest rookeries for nesting. About the first of May this species begins to col- lect dry kelp and stalks of the Farallon weed, which they carry 1n large mouthfuls to their rookery, usually on some rough ridge side. They are very dilatory in the work of construction, resting a great deal upon the nest. whereby a depression is formed on the accumulated material, for the eges. By May 10th some of the nests contain eggs, and a week later every nest is covered by a sitting bird; both sexes share the work of incubation, one staying on the nest while the other is off fishing. While sitting the birds purloin material from each other’s nests to add to their own. They will lay repeatedly in the same nest if robbed. A0) CALIFORNIA ACADEMY OF SCIENCES. Being the tamest of the cormorants, their nests can be ap- proached within a few feet before the birds reluctantly crawl off, and if no furthur advance is made, they crawl awkwardly back upon the eggs, knowing that the gulls will take them at once if deserted. If frightened by the too close approach of any one or the report a gun they take wing, but soon re- turn. The flight of the cormorants from their rookeries is spoken of by Mr. Emerson and others as a wonderful sight. The eggs of this species, fifty examples of which have been measured, average 60.7x39.5 mm. The most noticeable sizes are 64.5x38.5; 64x40.5; 63.5x41; 63x41; 57x38; 56.5x38 millimeters. 12, Phalacrocorax penicillatus. Branpt’s CoRMORANT.—They commence laying about the same time as the Farallon cormorant, selecting places rather accessible to an odlogist, where they congregate in large rookeries, principally at West End and Sugar Loaf. Like the other species they often have great difficulty in getting a nest built, as the thieving gulls will carry away the weeds about as fast as they can gatherthem. The eggs are usually four, sometimes five or even three in number, and vary somewhat in shape as well as size. Some eggs are rather oblong and pointed at the ends, others shorter and blunt at one or both ends—characters which are noticed in all the cormorants’ eggs from the island, there being no rule of shape to guide one in identifying the species. Eleven sets of four eggs and two sets of three give an average size of 62.5 x 38.5 mm. Special measurements of eight examples to show the ex- treme variations offer the following results, 66.5 x 38; 66x 37:5;, 65x36; 64x40; 615x371; 59.5x41- ioe 40. 56 x 38 millimeters. The young are hatched entirely naked, their skin resem- bling a greasy, black kid glove in appearance. In this con- dition, and even after the down is on them, they are irresisti- BIRDS FROM THE FARALLON ISLANDS 4] bie morsels to the hungry gulls. The young are abundantly supplied with small fish, which may be seen lying in and about the nests. 13. Phalacrocorax pelagicus resplendens. Barrp’s Cormorant.—They are less common than the two foregoing species, with which they do not associate. The nests are built usually in the most inaccessible places, and at all altitudes; some were found so close to the water’s edge that they were splashed by the highest waves beating against the rocky shore. The same rookeries are used from year to year, and the same nests are occupied after being robbed, the owners simply adding a few more pieces of weeds before laying. They congregate in colonies of eight or ten pairs, nesting on natural shelves of perpendicu- lar or overhanging rocks. Three or four eggs are laid in a nest of the same material as is used by the other cormo- rants. Incubation commences after the first egg is laid, in or- der to keep it protected from the gulls. The birds may be seen on the nests for days before the first egg is laid. The eggs of Baird’s cormorant may usually be known by their small size, the average dimensions of thirty-six speci- mens (nine sets of four), are 55.5x36.5 mm. By the fol- lowing examples it will be seen that the largest ones are nearly or quite the size of some of the smallest eggs of the other two cormorants, 59x27; 59x36; 58x38; 58x 37.5; Meo a <OOs OLX D0, O25 OD. The young are bare when hatched, and do not appear with down until they are, as Mr. Emerson judged, about three weeks old. 14. Pelecanus californicus. CALIFORNIA Brown PELicAN.—Seven were seen on the rocks one morning. 15. Fregata aquila. Man-o’-W ak Birp.—A skull of this was found by Gruber on the Island in 1861. 42 CALIFORNIA ACADEMY OF SCIENCES. 16. Merganser serrator. RED Breasted MrrcaNser.—One bird was taken by Mrs. Rugg. 17. Oidemia perspicillata. Surr Scorer.—A few seen feeding near the island. 18, Plegadis guarauna. WHITE-FACED Guossy Ipts.—One bird was shot in the spring of 1884, from a flock of half a dozen. 19. Ardea herodias. Great Brut Heron.—A few are seen on the island every winter. 20. Ardea candidissima. Syowy Hrron.—They have been noticed only on one oc- casion, when three were seen. 21. Rallus obsoletus. CALIFORNIA CLAPPER Rait.—A female bird was shot No- vember 18th, 1886. 22. Porzana jamaicensis coturniculus. Faratton Ratu.—Regarding the claim of the unique specimen of this bird as a valid variety, there does not appear to be sufficient grounds for its retention as a sep- arate form. In the light of Mr. Emerson’s investiga- tions, and that of others who have visited the islands and found no rails or suitable places for them to live, there seems to be but one conclusion to draw concerning it, and that is that it is an abnormal specimen of P. jamaicensis. 23. Fulica americana. AMERICAN CooT.—One was caught alive in the spring of 1884, and kept in a coop for several days before it was al- BIRDS FROM THE FARALLON ISLANDS. 43 lowed to go. They are very common on the water about the island during the fall. 24. Phalaropus lobatus. NORTHERN PHALAROPE.—They have been seen in the fall. 25. Tringa minutilla. Least SAnppIpER.—This species occurs in flocks on the island, and doubtless Hreuwnetes occidentalis does also, but specimens were not obtained, 26. Tringa alpina pacifica. RED-BACKED SANDPIPER.—Seen in flocks in the fall. 27. Heteractitis incanus. WANDERING T'ATTLER.—Several were seen every day in May, 1886, associated with the black turnstone. More were observed in June, 1885. They would not follow the turn- stones inland, but remained at the water’s edge. When flushed they rose with a short, shrill whistling note, and made a long flight before alighting. They appeared to be extremely shy at all times. At high tide they sat motionless upon the rocks, which they resembled in color, making it difficult to detect them unless flushed. When the tide was out they fed at the water’s edge; following a receding wave they searched hur- ridly amongst the kelp for food, running or making short flights to escape an incoming breaker. Several times a laggard was overtaken and covered by a breaker; when it receded he flew to the rocks and after shaking the water from his plumage, returned to feed. A male was taken July 22d, 1886, and a female later the same month (27th.) 28. Numenius longirostris. LONG-BILLED CuRLEW.—Stragglers occur during migra- tion. 44 CALIFORNIA ACADEMY OF SCIENCES. 29. Agialitis vocifera. KiLipEER.—Common at times during the fall. 30. Arenaria interpres. TurNsToNE.—An adult male taken by Mr. Emerson, May ‘th, 1887. 31. Arenaria melanocephala. Buack 'TurnstoneE.—Flocks of five to ten were seen in May and June, 1887. They were very tame, feeding slowly and quietly aloug the water’s edge in compact flocks. Upon higher ground they moved more rapidly and spread out more. Individuals were seen as late as August, but they do not, Mr. Emerson states, breed on the island. A female was taken July 26th, 1886. 32, Columba fasciata. BanD-TAILED Pickon.—A single bird came March 25th, 1886, and stayed for a day, seeming very tired. A number of pigeons visited the island in the winter of 1884-5. 33. Zenaidura macroura. Movurninc Dove.—As many as nine of these birds have been seen in a single day. Mr. Emerson observed five May 25th, 1887. Single birds were seen at different times dur- ing the same month, and in August, 1887. 34. Elanus leucurus. WHITE-TAILED KrrE.—Several seen by Mrs Rugg in the winter of 1886-7. 35. Accipiter cooperi. Coorrer’s Hawk.—A specimen taken in 1886. 36. Circus hudsonius. MarsH Hawk.—One seen in May, 1885. BIRDS FROM THE FARALLON ISLANDS. 45 37. Buteo borealis calurus. Western Rep-tat..—Every spring the island is visited by numbers of these hawks. In 1882 they came in April, about the time of the arrival of the murres, leaving again in May. During their short stay they fed almost exclusively upon the murres, killing, in the estimation of Mr. Emerson, several dozen a day. In 1887 the lighthouse men killed about seventeen of these hawks, and during the month of May, 1885, twenty-eight, mostly of this species, were destroyed. 38. Archibuteo ferrugineus. FrErrucineus Roucu-tec.—One specimen killed. 39. Falco mexicanus. PRAIRIE FALCON. 1886. An adult male taken December 18th, 40. Falco peregrinus anatum. Duck Hawx.—A specimen was shot December 15th, 1886. 41. Pandion haliaetus carolinensis. AMERICAN OsprEY.—One was shot December 15th, 1886; no others have been seen since. 42, Asio accipitrinus. SHORT-EARED OwL.—One was seen in the spring of 1885 and one in May, 1887, by Mr. Emerson. 43. Speotyto cunicularia hypogea. Burrowinec Owt.—I'wo were seen in the spring of 1887, sitting on the rocks. 44, Ceryle alcyon. BELTED KINGFISHER.—One was seen August 16th, 1887. 45, Colaptes cafer. RED-SHAFTED FLIcKER.—Occasionally seen in the winter. 46 CALIFORNIA ACADEMY OF SCIENCES. A specimen, yellow on under surface of wings and tail and with red cheek-patches, is in the collection of Mrs. Rugg. 46. Trochilus anna. Anna’s HumMineG-BirD.—A single bird has been seen at the island. 47. Trochilus. Humming-birds, supposed to be either rufus or alleni, have been seen in the spring, and also on August 15th, 1886. 48. Tyrannus verticalis. ARKANSAS KINGBIRD.—Seen May 28th, 1887. 49. Contopus richardsonii. WESTERN Woop PEWEE.—Seen on a number of different days during the month of May, 1887. 50. Empidonax difficilis, Barrp’s Fiy-carcHerR.—One specimen was taken by Mrs. Rugg. 51, Corvus corax sinuatus. AMERICAN Raven.—Old nests of this bird are mentioned by a visitor to the island as having been shown him by the © light-keeper. As they are known to build among the rocks in the Santa Barbara Islands, the reference is probably correct. 52. Corvus americanus. AMERICAN Crow.—One pair seen during visit of May, 1887; two pairs in June, 1885. 53. Agelaius vheeniceus. RED-WINGED Buacksirp.— A fall visitor. 54, Sturnella magna neglecta. WesterN Merapow-Lark. — Stragglers come during the winter. BIRDS FROM THE FARALLON ISLANDS. 47 55. Scolecophagus cyanocephalus. Brewer's Biackpirp.—The shyest of the straggling land birds that visit the island. They appear in large flocks during the winter. 56. Carpodacus frontalis rhodocolpus. Crimson Hovusr Fincno.—They occur at times during the winter A young of the year was seen by Mr. Emerson May 26th, 1887. 57. Spinus psaltria. ARKANSAS GOLD-FINCH.—Small flocks appear at intervals during the winter; they feed on the seed of the Farallon weed, and often remain on the island for several days. 58. Zonotrichia gambeli. GAMBEL’s SpaRrRow.—This and the next following species visit the islands in flocks. 59. Zonotrichia coronata. GOLDEN-CROWNED SPARROW.—An immature female taken September 16th, 1886, on which day large numbers of this species were seen. 60. Spizella socialis arizone. WESTERN CHIPPING SpaRRow.—One seen May 22d, 1887. 61. Junco hyemalis oregonus. OrEGon Junco.—Have been frequently seen in large flocks during the winter. 62. Habia melanocephala. BLACK-HEADED GROSBEAK.—I'wo birds have been seen; one was taken May 28th, 1887. 63. Piranga ludoviciana, Lovutstana TanaGeR.—A male in immature plumage was A8 CALIFORNIA ACADEMY OF SCIENCES. taken from a flock of nearly a dozen, on September 13th, 1886. 64. Chelidon erythrogaster. Barn Swattow.—A pair came to the islands on May 21st, 1887, and after vainly trying to find mud for a nest, departed on June 2d, 65. Lanius ludovicianus excubitorides. WHITE-RUMPED SHRIKE. — One seen on the island for several days in 1886. 66. Vireo bellii pusillus. Least VirEo.—Two male specimens, one of which was caught alive in the house on May 28th, 1887, are in Mr. Emerson’s collection. 67. Mniotilta varia. Buiack aND WHITE WarsLer.—A male in good plumage was found on the rocks below the bell wires that lead to the tower May 28th, 1887; it is now in Mr. Emerson’s collection. It measures, wing, 68 mm.; tail feathers, 50 mm ; bill from nostril, 8.5 mm.; tarsus, 16 mm.; middle toe and claw, 15.5 mm. 68. Helminthophila celata lutescens. LutTesceNtT WARBLER.—One was seen in company with other small land birds on May 22d, 1887. 69. Dendroica estiva marcomi. WESTERN YELLOW WarRBLER.—Iwo seen May 22d, 1887. 70. Dendroica czrulescens. BLACK-THROATED BLUE WARBLER.—A female of this species ~ remained in and about the house for three weeks, dying November 17, 1886. It is believed to be the first occurrence of this warbler in California, for it has never been observed aaa aa BIRDS FROM THE FARALLON ISLANDS. AQ on the mainland. It measures, wing, 59.5 mm.; tail feath- ers, 49 mm.; bill from nostril, 7 mm.; tarsus, 18 mm. 71. Dendroica townsendi. TOWNSEND’Ss WARBLER.—TIwo males were seen on May 22d, 1887. 72. Sylvania pusilla pileolata. PILEOLATED WaARrBLER.—Immature male taken September 12th, 1886. A female seen May 22d, 1887. 73. Galeoscoptes carolinensis. CatpirD.—The occurrence of this bird on the South Farallon has been noted by Mr. C. H. Townsend, who secured a specimen on September 4, 1884. 74, Salpinctes obsoletus. Rock Wren.—Mr. Emerson found this species to be quite common on the island, and it is the only land bird which is resident. They were found all over the South Farallon, but more numerous about the southern exposure of the light- house hill. They build in suitable crevices a nest of Faral- lon weed, lining it with chicken feathers and hair shed from the island mule. About the entrance they accumulate small bits of rock, coal, mussel shells, and small bones from dead sea-birds and rabbits. Nests containing eggs were found on the 5th of April, 1886, and on June 16th, 1885. Five and six eggs constitute an average set, but one nest was found with ten eggs, two of which were fresh, and the others in various stages of incubation. Aswell as can be judged from the specimens at hand, which are in worn plumage, the island rock wrens are iden- tically the same as those from various localities in Califor- nia. The average measurements of five males, taken in May, 1887, give the following results: 2p SER., Vou. I. Issued January 19, 1888. + 50 CALIFORNIA ACADEMY OF SCIENCES. Wing, 72.5 mm.; tail feathers, 55.1 mm.; bill, from nos- tril, 14.3 mm.; tarsus, 20.4 mm.; middle toe, 14.8 mm. Three females, shot at the same time, average: Wing, 70 mm.; tail feathers, 52 mm.; bill, from nostril, 13.8 mm.; tarsus, 20.6 mm.; middle toe, 15 mm. 75, Sitta canadensis. RED-BREASTED NutHatcH.—A number were seen in Sep- tember, 1885, and a single female secured for identifica- tion. 76. Regulus calendula. Rupsy-CcROWNED KineLer.—A single specimen was taken and prepared by Mrs. Rugg. 77, Turdus aonalaschke. DwakF Hermit THrusH.—One specimen seen by Mrs. Rugg in the winter of 1886-7. 78. Merula migratoria propinqua. WesTERN Rospin.—Stragglers are occasionally seen dur- ing the winter. 79. Hesperocichla nevia. VarieD THRUSH.—Several came to the island in the win- ter of 1886-7, and a specimen was secured. 80. Sialia mexicana. WesTERN BLUEBIRD.—A few seen occasionally. 81. Sialia arctica. Mountain Buiuepirp.—They appear in flocks in the win- ter time, but remain only a few days. AMERICAN GOBIIDA) AND CALLIONYMID As, 51 A List of the American Species of Gobudee and Callionynide, with Notes on the Specimens Contained in the Museum of Comparative Zoology, at Cambridge, Massachusetts. By Cari H. E1cgenmMaAnN and Rosa Smira EIGENMANN. In the Proceedings of the United States National Mu- seum, 1886, pp. 477-518, Jordan & Higenmann have given descriptions and the synonymy of the species of Gobiide of North America. In this paper we give descriptions of all American forms not mentioned by Jordan & Higenmann. The nomenclature has been corrected, and the following forms described as new: 1, Gobius lucretie; 2, Gobius gar- mani; 3, Gobius hemigymnus; 4, Microgobius eulepis; 5, Boarbulifer papillosus, gen. et spec. nov.; 6, Clevelandia longipinnis, gen. nov.; 7, Callionymus calliurus. We wish to express our thanks to Dr. Alexander Agassiz, Curator of the Museum, and to Mr. Samuel Garman, for the privilege of studying the American Fishes contained in the collection of the Museum of Comparative Zoology. Family GOBIIDA. Suab-family OxyMETOPONTINE. I. lIoauossus Bean. 1. Ioglossus calliurus Bean. Gulf Coast of the United States. Sub-family ELEoTRIDINm. II. Goxstomorus Lacépede. 2. Gobiomorus lateralis (Gill). Panama. Nos. 13,430 and 13,431, three specimens. Cape St. Lucas, Xantus. The specimens have a broad lateral band and a large dis- tinct spot on base of caudal. 2p SER., VOL. I. Issued January 25, 1888. 52 CALIFORNIA ACADEMY OF SCIENCES. 3. Gobiomorus dormitor Lacépéde. Eleotris longiceps Giinther, Proc. Zool. Soc. Lond., 1864, 151 (Lake Nicaragua). West Indian Fauna, East Coast of Central America and Surinam. The specimens in the Museum are from St. Thomas, Surinam, Barbadoes, Fort de France, Martinique, St. Lucia, Hayti, and Lake Nicaragua. The largest specimen is a head from Hayti, which meas- ures 5} inches from tip of lower jaw to end of opercle. The color in adult individuals varies from light brown to blue black. The young somewhat like Gobiomorus lateralis in general coloring; having profuse dark markings, the belly plain; an interrupted black bar on middle of sides; no regu- lar spot at base of tail. A specimen collected in Lake Nicaragua by Freebel is presumably to be referred to Hleotris longiceps Giinther, but no tangible differences can be detected between it and speci- mens of Gobiomorus dormitor from other localities. A comparison of the species G. lateralis and G. dormitor shows them to be very closely related. We append the differences observed: lateralis. dormitor. Scales in median series uniformly, 52. Scales in median series, 55—57. In a cross series, 17. In a cross series, 18-—22. Scales from nape to first dorsalspine, Scales from nape to first dorsal spine, 27. 27—34. Head in length, 24—3. Head in length, 3—3}. Depth, 5—6. Depth, 5%—6. Dorsal vi--10. Dorsal vi—10. Anal 1a Anal 10. 4. Gobiomorus maculatus (Giinther) Ecuador. III. Dormrraror Gill. 5. Dormitator maculatus (Bloch). Eleotris latifrons Richards. Voy. Sulph. Fishes 57, plate 35, figs. 4-5, 1837. Dormitator micropthalmus Gill, Proc. Acad. Nat. Sci. Phila. 1863. 170. AMERICAN GOBIIDA’ AND CALLIONYMIDA, 53 West Indian fauna from South Carolina to Para; Panama. The specimens in the Museum are from Charleston, S. C., Mobile Bay, Barbadoes, Sao Matheos, Cuba, Hayti, Porto Rico, St. Thomas, Martinique, Par4, Gurupa and mouth of Rio Grande. Specimens 1} inches long, collected by Mr. Samuel Gar- man at Martinique, have interrupted cross bars of light and dark. Numerous specimens of the same length from canals in Para have similar, but fainter, markings; while many of them have longitudinal lines formed by the occurrence of a black spot on the centre of each scale. There seem to be two forms of the adult—one with the profile gibbous, the dorsal outline forming a regular curve; the other having the profile depressed over the eyes, the anterior portion being sub-horizontal. The specimens from Gurupa and the Rio Grande have the profile depressed; all the other specimens have a gibbous profile. A comparison in detail of the two forms is appended. Only extreme differences are given. West Indian inches. Profile regularly curved from first dorsal spine to snout. Head 3% to 4; depth3 to 34; depth always greater than length of head, Highest anal ray 1} to 12 in head. Distance from first dorsal spine to snout greater than distance from first dorsal spine to first anal ray. Scales in median series 29 to 32. Color usually dark brown, a black spot above base of pectoral, a short bar on base of pectoral. specimens, 5 — 7144 Dr. Jordan’s description of Rio Grande specimens, 5, 6% and 7344 inches. Profile depressed over eye, becom- ing horizontal anteriorly. Head 3; depth 3 to 3%; depth usu- ally less than length of head. Highest anal ray 12 to 2in head. Distance from first dorsal spine to snout equals distance from first dorsal spine to base of last anal ray. Scales in median series 30 to 34. Color gray, a jet black spot above base of pectoral; a black bar at base of pectoral; a black line from eye to mouth; longitudinal black lines on cheeks and opercles; dark spots on back; some silvery scales on sides. the types of Hleotris grandi- squama agrees well with the specimens of Dormitator mac- 54 CALIFORNIA ACADEMY OF SCIENCES. ulatus from the Rio Grande. If it is-identical with them the species grandisquama cannot be separated from macula- tus. The characters given by Kigenmann & Fordice* to dis- tinguish maculatus from the Pacific Coast species, latifrons, are not of specific value. The numerous specimens of mac- ulatus examined by us differ in every character more than maculatus is said to differ from latifrons. Thus the number of scales between the ventral and the vent differs from 12 to 20 in specimens of maculatus which we have examined. 6. Dormitator grandisquama (Cuv. & Val.). America. Probably another synonym of the preceding. ITV. Guavina Bleeker. 7, Guavina guavina (Cuv. & Val.) West Indian fauna, south to Rio de Janeiro. The specimens in the Museum are from Ceara, Victoria, Sao Matheos, Cuba, Rio Janeiro, Rio Grande do Sul, Turbo, Atlantic Coast of Central America, Barbadoes, Goyaz, Brazil. 8. Guavina brasiliensis (Sauvage). Brazil. Eleotris (Eleotris) brasiliensis Sauvage, Bull. Soc. Philom., Paris. Series 7, IV, 53; 1880. (Bahia. ) Depth 63 in total length; head 42. D. VII—I, 9; A. I, 9; Scales, 80. Snout equal to eye, 5 in head; 2 in interocular space. Lower jaw a little produced, maxillary to below anterior third of eye. Teeth of outer series enlarged. Preopercle without spines. Scales small, ciliated, 35 series between origin of second dorsal and anal. Scales on top of head as large as those of body, and reaching to tip of snout. Uniform brown; length 0.115 m. ~ *Proc. Philad. Acad. Nat. Sci., 1885, 71. AMERICAN GOBIID# AND CALLIONYMIDA, 55 V. Eerorris Gronow. 9, Eleotris amblyopsis Cope. Surinam; Atlantic and Gulf coast of the United States. We have examined fifteen specimens, 23 inches or less in length, all from Surinam. The scales in a median series vary in these specimens from 40 to 44; in a transverse series between second dorsal and anal, from 12 to 14. These are the only characters which separate this species from -/. pisonis (Gmelin). The American species of Hleotris may be distinguished as follows: a. Teeth sub-equal, those of the inner or outer series enlarged. b. Cheek entirely scaled. e. Teeth of the inner series of both jaws enlarged. d. Seales in a median series 40 to 44; in a cross series 12 to 14. amblyopsis, 1. dd. Scales in a median series 57 to 66; in a cross series18 to 20. pisonis, 2. ce. Teeth of the outer series enlarged...............-+.+0: beliziana, 3. Op Howerhnltof cheels maked. weass. 6 4 aes nneeve cso -ler perniger, 4. Mee CURL ALL TEC AI eee sateroycts aioe leis, sonsie alate a fin Sales cote are equidens 5. 10, Eleotris pisonis (Gmelin). West Indian fauna; Florida to Rio de Janeiro. The specimens in the Museum are from Tajaperu, Para, Cuba, Rio de Janeiro, Victoria, Rio Doce, Nazareth, Pernambuco, Ft. de France, Martinique, St. Vincent, Dominica, St. Kitts, Hayti, Florida. There is considerable variation among the specimens; the young have narrow longitudinal lines, and lines radia- ting from the eye; the teeth of the inner series are enlarged —more in young, less in adult specimens. 11, Eleotris perniger Cope. West Indian fauna south to Rio de Janeiro. 12. Eleotris beliziana Sauvage. Atlantic Coast of Central America. 13, Eleotris equidens (Jordan & Gilbert). Panama. 56 CALIFORNIA ACADEMY OF SCIENCES. VI. Erore.is Poey. 14. Erotelis smaragdus (Cuv. &;Val ). West Indian fauna; Key West. VII. Gymneteorris Bleeker. 15. Gymneleotris seminuda (Ginther). Panama. Sub-family Gopiinz. VILL. Stcyprium Cuv.’& Val. 16. Sicydium plumieri (Bloch). West Indian fauna. The specimens in the Museum are from Hayti, Dominica, Martinique, Cuba (Poey’s types of siragus), St. Vincent. Mr. Samuel Garman collected several hundred specimens of this species at Kingston, St. Vincent. Most of these specimens are less than an inch in length, the longest 1} inches; they differ considerably in coloration from the adult; most are entirely naked, a few of those examined having scales only on the posterior part of the body. Caudal deeply emarginate. There are traces of about seven dark vertical bars; a black bar at base of pectoral; dorsals with several series of black spots; an H-shaped figure on base of caudal, a black bar on the posterior half of caudal fin; belly and lower part of body plain; everywhere else with black points. The specimens collected by Mr. Garman may be the types of a new species. No large specimens were collected at the Island of St. Vincent. Specimens an inch and a half in length from Hayti have the fins plain and a series of blotches along the middle of the posterior part of the body; the body, except the belly, is entirely covered with scales which are plainly ctenoid. IX. Sicyoprervus Gill. 17. Sicyopterus gymnogaster (Grant). Panama. 18. Sicyopterus salvini (Grant). Panama. AMERICAN GOBIIDA AND CALLIONYMIDA. Shy X. Evortsopus Gill. 19. Evorthodus breviceps Gill. West Indian fauna. XI. Lorxocosius Guill. 90. Lophogobius cy prinoides (Pallas). West Indian fauna. XII. Gostvus Linnzxus. 21. Gobius lucretiz spec. nov. No. 3957, one specimen, 134 inches. Pearl Island, Gulf of Panama. Head 33 in length (43 in total); depth 53 (7); D. VII-10; A. 12. Scales, in a median series, about 28; in a cross series, 8. Body slightly compressed posteriorly; head little wider than high; eye placed high, its diameter equal to length of snout, 4} in head; profile little decurved; mouth large, oblique, maxillary extending below posterior margin of orbit, 2 in head; intermaxillary anteriorly on a level with the centre of the pupil; teeth all recurved, large, those of the upper jaw in a narrow band, the teeth of outer and inner series enlarged, those of the lower jaw similar, largest in front. No dermal flaps on shoulder-girdle. Scales large, very weakly ctenoid; becoming cycloid and very much crowded above and below pectoral; head, breast and ante- rior part of nape naked. As seen through a lens, these regions seem to be covered with minute, imbedded scales: this effect is, no doubt, due to light reticulations on a darker ground. Dorsal spines slender, not filamentous; caudal pointed, 3 in length of body; ventrals 1} in head; pectorals longer than head. Color ight brownish, with four oblique, dark cross bars as wide as the interspaces; four narrower transverse bars on nape and back; a large dusky spot at base of caudal; upper 58 CALIFORNIA ACADEMY OF SCIENCES. half of base of pectoral black; a black spot on opercle, margined below and behind with silvery; fins dusky. We have named this species for Mrs. Lucretia M. Smith, of San Diego, California. This is the only American species of Gobius proper hav- ing 7 dorsal spines. 22, Gobius soporator Cuv. & Val. All tropical seas. The specimens examined are from Panama, Barbadoes, Para, Itapu- ana, Cuba, Galapagos, Sambara, Bahia, Orange Key, Bahamas, Pernambuco, St. Thomas, Tortugas, Florida Keys, Martinique, Sao Matheas, Curuca, Rio de Janeiro, Rio Doce. The color variations among examples of this species are very great, specimens from one locality varying from plain sand color, or gray, to greenish black: some dark brown specimens have light bars across the back; in others the scales have light centers forming horizontal series of light lines; sometimes there are light spots on sides of head and cheek; some specimens are conspicuously marbled with light and dark brown; and white spots occur in the centres of some of the scales on specimens of any ground color, these white spots being brighter on some of the scales than on the others, forming interrupted longitudinal lines. If any value could be placed upon the coloration, almost every specimen would be a distinct species. The color variation is irrespective of locality, some localities having all the above described variations. The types of Poey’s mapo, lacertus and brunneus prove to be color varieties of Gobius soporator. We have compared these various forms of soporator with Gobius albopunctatus collected by Dr. Klunzinger in the Red Sea; with others collected by Mr. Garrett in the Society Islands, Sandwich Islands and Kingsmill Islands; with specimens collected by C. L. Salmin at Cape York, Aus- tralia, and with those collected by B. G. Snow on Ebon AMERICAN GOBIIDA AND CALLIONYMID A. 59 Island, Marshall Group. We are not able to detect any speci- fic differences. Some of the American specimens might readily be identified as albopunctatus, but, on account of the perfect intergradation of these specimens with the typical soporators, we prefer to refer all the specimens be- fore us to the one species soporator, which seems therefore to be universally distributed in tropical seas. 93. Gobius andrei Sauvage. Ecuador. Bull. Soc. Philom., Paris. Ser. 7, LV, 44; 1880. Depth 6 in total length; head, 4. D. VI-I, 9; A. I, 7. ' Seales, 35 in a median, 12 in transverse series. Head much wider than high; snout equal to diameter of eye, 43 in head; interorbital space narrower than eye; max- illary reaching below centre of eye. Teeth of outer series larger than the others; no canines. first dorsal not quite as high as body; caudal rounded; upper rays of pectoral silky; ventrals not to vent. Scales ciliated. Uniform brown; lower part of spinous dorsal tinted with black; soft dorsal and anal with large brown spots; anal blackish. Length 0.160. (Guayas, Ecuador.) An examination of the type of this species will doubtless prove it to be a soporator. . 24, Gobius nicholsi Bean. Californian fauna. 25, Gobius glaucofrenum (Gill). West Indian fauna. There are four specimens of a Gobius in the collection labelled ‘‘ No. 13,212, Tortugas, Lyman.” ‘They agree with Dr. Gill’s description of glaucofrenwm, which is said to be found on the coast of Washington Territory. Evidently a mistake has been made in the locality of the type or of the specimens in the Museum of Comparative Zodlogy, as it is not probable that this species occurs in the waters of Wash- 60 CALIFORNIA ACADEMY OF SCIENCES. ington Territory and also in the Tortugas. As no Goby of this type was found on the Pacific Coast by Jordan & Gil- bert or by Dr. Bean, in very thorough explorations of that region, it is likely that the locality originally assigned is an erroneous one. The following is a description of the specimens from the Tortugas : Length 17, 1,4, 14, 14 inches. Head, 34 in length (42 in total); depth, 4% (53); D. VI—10; A. 10. Scales, ina median series, 23, in a trans- verse series, 8. Kye as long as snout, 34 in head; jaws equal; maxillary barely reaching pupil; teeth in bands in both jaws, those of the outer row of the lower jaw enlarged. Dorsal spines scarcely filamentous; the third highest and equal to depth of body; posterior dorsal rays highest, as high as spines. Anal similar to soft dorsal; pectoral long and narrow, longer than head, 33 in body; ventral reaching past vent. Scales large, thin, finely toothed, reduced on breast; nape naked. Color in spirits, ight yellowish-brown. A light spot on each scale, the spots especially conspicuous near shoul- der; six dark spots on middle of back; fainter but similar spots along middle of sides; a conspicuous dark spot above opercle, a wavy light line extending forward from it through lower rim of eye to snout; a straight pale blue bar extending parallel to it across preopercle and cheek to corner of mouth; a narrow faint bar below it; a triangu- lar dark spot at corner of mouth. Cheeks and preopercle purp- lish chocolate; opercle and snout plain yellowish; two brown spots at base of caudal. The smaller specimens differ from this in having the markings more distinct. AMERICAN GOBIIDA AND CALLIONYMIDA. 61 26, Gobius quadriporus Cuv. & Val. Surinam. Cuvier & Valenciennes, Hist. Nat. Poiss., XII, 87, 1837. (Surinam.) D. VI—1, 9. A.1,9. Scales as in G. caninus. The two pores on the vertical arm of the preopercle very open; two smaller ones above them; teeth of the outer series small, two small canines on each side of lower jaw; dorsal spines not prolonged as filaments; color yellowish, with lighter lines which follow the rows of scales; brown spots on dorsal; two lines on cheeks. 27. Gobius stigmaturus Goode & Bean. West Indian fauna; Key West. 28. Gobius shufeldti Jordan & Eigenmann. Gulf Coast of the United States. 29. Gobius garmani sp. nov. The types of this species are: No. 26,081, one specimen, 3; inches long; Dominica. 8S. Garman. No. 26,062, three specimens, 2 inches, 1$ inches, and 1 inch; Fort de France, Martinique. S. Garman. No. 27,122, four specimens, 3 inches, 23 inches, 2§ inches, and 2? inches; St. Kitts. S. Garman. Head 4 in length (53 in total); depth 4 (53). D. VI—11; A. 11. Scales in a median series 30, in a cross-series 7. Body robust, head short and blunt; profile in front of eye abruptly decurved, rounded much as in Gobius boleo- soma; mouth inferior, horizontal; lower jaw included; max- illary extending to below pupil, 2} in head; lips thin; teeth short and thick, in a single series in each jaw. Dorsals contiguous; dorsal spines filamentous, the second and third longer than the rest, reaching past first third of second dorsal; last dorsal rays reaching base of caudal; pectorals equal to the head in length; ventral short and broad, 5 in body; caudal rather long and pointed, 3 in body. Scales large, slightly reduced and cycloid on nape. Color yellowish, marbled with brown; a series of irregular 62 CALIFORNIA ACADEMY OF SCIENCES. blotches along the sides; a light spot at base of caudal, partly or wholly surrounded by a broad ring of dark brown; head slate color, white below; three dark bars extending forward and downward from eye to mouth, a triangular spot on opercle; dorsals, caudal, and pectorals finely barred with black; a chocolate bar on base of ventral; anal margined with white; an irregular black bar on shoulder and upper half of pectoral; everywhere more or less blotched with darker; the blotches at times forming numerous bars across the back. This species is most nearly related to Gobius boleosoma Jordan & Gilbert, and Gobius fasciatus Gill, from which it can readily be distinguished by its proportions and its color. 30. Gobius boleosoma Jordan & Gilbert. Gulf Coast of the United States. 31. Gobius fasciatus (Gill.) West Indian fauna. No. 13,231, one specimen, 1} inches; Hayti. Weinland. This species has so far been known only from an insufii- cient description of the types which seem to have been lost. We add a description of the specimen before us: Head 4 in length (42 in total); depth 6 (7); D. VI—12; A. 10. Scales in a median series about 30; in a cross- series 7. Body slender, elongate; head somewhat pointed; profile rounded, not as much as in G. garmani and G. boleosoma; eye large, slightly longer than snout, 3? in head; interorbi- tal area scarcely wider than pupil; mouth slightly oblique, maxillary extending to below anterior margin of pupil, 3} in head; lower jaw thin and flat; teeth strong, recurved, in a band in each jaw; the teeth of the outer series of the upper jaw enlarged, several times as large as those of the inner series. 9 AMERICAN GOBIIDA AND CALLIONYMID. 63 Scales finely serrate (fallen off anteriorly); ante-dorsal region and breast naked. Dorsal spines slender, filamentous near tip, not reach- ing second dorsal, 1} in length of head; second dorsal of moderate height; caudal (tips broken) about 5 in length, 12 in length of head; ventral not reaching vent. 1} in head: pectorals pointed, equal to the head in length. Color yellowish, marbled with darker above; four oblong dark blotches along middle of sides; a darker spot at base of caudal; narrow dark stripes across nape; a faint dark stripe along upper margin of opercle, through lower margin of eye to snout; another extending from angle of mouth to edge of preopercle, then extending down along the margin of the preopercle and ending in a dark blotch on the lower part of the cheek; a dark spot on opercle; first dorsal with two curved bars; caudal with three rather broad dark bars; anal dusky; connecting membrane of ventral white, its first rays blackish, outer rays yellowish; lower parts yellowish. 32 Gobius stigmaticus Poey. West Indian fauna; Southern United States to Rio Janeiro. The specimens in the Museum are from Cuba and Rio Janeiro. We have compared the specimens collected by Prof. Agassiz with the typeof Poey. The Rio Janeiro specimens are darker than the type; the bars on the cheek are con- spicuous, edged with white; the shoulder spot is larger than the eye. The third dorsal spine sometimes is very elongate, reaching the fifth or sixth dorsal ray; the last dorsal ray oc- casionally extends to the base of the caudal. Posterior canines of lower jaw conspicuous, two on each side in some examples. The canines are also present in the typical specimen. 33, Gobius lyricus Girard. (No. 13,109, Type of Gobionellus costalesi Poey.) Depth 55 in length; head 4; D. VI—12; A. 12; pectoral 64 CALIFORNIA ACADEMY OF SCIENCES. slightly longer than head. Scales on head much reduced, toward the caudal enlarged. Caudal long, lanceolate, 23 in length of body; highest dorsal ray 3}; highest anal ray 3}. Color brownish, the first dorsal with dots on its spines; a light stripe near the upper and lower margins of the caudal fin. The bottle No. 13,109 contains two specimens; the smaller one is stigmaticus, the larger we describe above, and we con- sider it identical with Gobius lyricus Girard. Three other specimens of Gobius lyricus were collected on the Island of St. Kitts, W. I., by Mr. Garman. The largest is 3 inches long. Depth 5 in length; head 4. The second and third dorsal spines extend to base of caudal; dorsal rays scarcely less than length of head, the last rays reaching past base of caudal; the caudal fin is two-thirds longer than the head, 21 in body. Color light brown, faintly marked with darker; the first dorsal with minute derk points, the lower fourth of the spines with simple dark spots, above which are jet black spots ocellated with white; the second dorsal fin dusky, darker posteriorly, the basal portion of the last half of the fin evenly black, the anterior four rays marked with dark points similar to the spots on the lower parts of the spines of the first dorsal; caudal dusky, with two light bars; anal plain, darker than body; ventral fins blackish, edged with white; pectorals blackish, with many series of white spots on the membrane, and short, white bars at base; branchiostegal membrane black, with a light margin. 34. Gobius encaeomus Jordan & Gilbert. South Atlantic Coast of the United States. 35. Gobius smaragdus Cuv. & Val. West Indian fauna to Rio Janeiro. The specimens in the Museum were collected by Prof. L. Agassiz at Rio Janeiro. AMERICAN GOBIIDA AND CALLIONYMIDA, 65 36. Gobius badius (Gill). Gobius bosci Sauvage, Bull Soc. Philom., Paris, series 7. iv. 44. 1880. Gill, Aun. Lyc. Nat. Hist. New York, vii, 47, 1859. (Amazon). Head 6 in total; depth about 7; D. VI—1,10; A. 1.10; scales in a median series about 50, in a transverse series 18. Anterior profile very oblique; a line of pores behind eye; & pore above each eye; two on upper half of ascending margin of preopercle; eye 4 in head; interobital 1+ of or- bit; caudal 5 in total length; pectorals 6; color reddish brown or ‘‘ dark bay, with a posteriorly straight hoary dot in the center of each scale; on the back and sides above, the head is plumbeous, with two livid blue bands extending from the eye to the upper jaw.”’ 37. Gobius poeyi Steindachner. West Indian fauna. 38. Gobius strigatus O'Shaughnessy. Gobius Kraussi Steinduchner. Ichthyol. Beitr. VII[. 19. 1879. West Indian fauna. 39. Gobius sagitulla (Ginther). Panama. 40. Gobius uranoscopus Sauvage. Brazil. Sauvage. Bull, Soc. Philom. Paris; series 7, IV, 170. 1880. Depth 7 in total length; head 44. D VI—10; A. 9; Scales 98. Head 2 times as long as high; snout obtuse, shorter than eye; eye 4 in head, interorbital space very narrow, the eyes almost touching. Teeth of outer row much longer and stronger than the others. Soft dorsal high; caudal rounded; upper rays of pectoral not silk-like; ventrals long; brown, marbled with black; some brown points on caudal. Length .065 m. (Brazil). 41, Gobius oceanicus Pallas. West Indian fauna; south to Rio Janeiro. The specimens in the Museum are from Rio Janeiro, Pernambuco, Porto Alegre, Nazareth, Sao Matheos. 2D SER., Vou. I. Issued January 25, 1888. 5 66 CALIFORNIA ACADEMY OF SCIENCES. 42. Gobius hemigymunus spec. nov, No. 27,124, one specimen, 1%4 inches. Dredged in the West Indies. Head 32 in length, (44 in total); depth 43 (53); D VI—10; A 8. Scales smaller than in G. paradoxus, 17 series devel- oped, 7 in across series on tail—fewer further forward. Body compressed; depressed anteriorly, the greatest depth in this specimen being at the origin of the anal and second dorsal fins. Head wider than deep, rounded; pro- file much decurved from eye to mouth as in paradoxus; eye perfectly round, smaller than in paradowus, 15 in the rounded snout, 5 in head; interorbital space scarcely wider than orbit; mouth somewhat oblique, larger than in paradoxus, the maxillary reaching beyond posterior rim of orbit; lower jaw slightly shorter than the upper; teeth in the upper jaw in a band, the outer series remote, and the teeth several times as large as in the inner row, all more or less movable; teeth in the lower jaw similar, a recurved canine on each side near the front. Seales very weakly ctenoid, covering only the sides of the posterior half of the body, not extending quite to base of dorsal or anal fins even at their posterior insertion, the upper and lower edges of the caudal peduncle being lke- wise free from scales, the scaly region, however, being widest on the peduncle, and tapering forward to the central point opposite beginning of anal, where the scales are smallest. First spine of the dorsal not elongate as in paradoawus, 14 in head, the 3d, 4th and 5th spines slightly exceeding the first in height, equalling the posterior rays of soft dor- sal, which are little higher than the anterior rays of the soft dorsal; caudal rounded, about 4 in length of body, 1} in head; ventral not reaching vent, 13 in head; pectorals rounded, rather short and broad, 14 in head. Color light olivaceous, without distinct markings, every- where with minute dark punctulations; eight faint cross AMERICAN GOBIIDAH AND CALLIONYMIDA, 67 bars from dorsal to middle of sides, which, close under dorsal fins, are formed of two blackish dots; eight black dots along lateral line, the last being at base of caudal; fins all smutty, the pectoral lightest, white on its anterior half, two dusky spots at its base; opercle ashy; a light bar at base of caudal; iris blackish blue, a short straight streak of same color from eye to upper lip; an irregular bluish mark on cheeks formed of punctulations closely crowded. Related to Gobius paradoxus Ginther, from which this species differs in the size of the mouth, the greater width of the head, the arrangement and number of scales, the first dorsal fin, the color, and other minor characters. The specimen was dredged. Its exact locality is not known. 43. Gobius paradoxus Giinther. Panama. 44, Gobius seminudus Giinther. Panama. XIII. Cxonornorvs Poey. 45. Chonophorus flavus (Cuv. & Val.) Brazil. The Museum contains two specimens (No. 13,076) from Bahia, collected by Professor Agassiz, and one (No. 13,- 094) from the Rio Doce, which may be referred to this species? Our description is taken from the specimens from Bahia. D. VI—12; A. 11; Seales 55. Body very elongate, depth 63 in length; mouth horizon- tal, lower jaw flat; teeth in two series in upper jaw, the outer ones enlarged; in the lower jaw the teeth are in a band. Color yellow, with a row of faint ocellated spots along middle of sides; dorsal and caudal fins faintly barred; lines ra- diating from eye; a line along opercle and half way across pectoral; all the markings obscure. 68 CALIFORNIA ACADEMY OF SCIENCES. The specimen from Rio Doce, like most of the speci- mens of other species from this river, is plain blue-black 46. Chonophorus taiasica (Lichtenstein). Euctenogobius latus O'Shaughnessy. Ann. & Mag. Nat. Hist., Series 4, XV, 145. 1875. Atlantic and Pacific Coasts of tropical America. The specimens in the Museum are from Bahia. Rio Doce, Barbadoes, St. Vincent, St Lucia, Dominica. Mendez, Martinique, Panama. The color and form in this species vary much with the stage of growth of the individual; the young resembling the species of Hadropterus in color and shape. In the Panama specimen the head is narrow and pointed, eye large; the series of spots along the sides, conspicuous and well defined; the caudal has three black bars instead of the usual nine or ten narrow lines; a black spot at base of caudal. Three specimens, collected by Mr. S. Garman at Dominica, are similarly marked, except the caudal, which has wavy black lines; specimens collected by him at St. Vincent are of a plain sand color. In the older forms the head becomes much widened, the premaxillary widens in front and projects over the lower jaw, the coloration often be- comes less defined, varying from plain sand color to dark olive. The teeth in the young are proportionately larger; many of the teeth are lost with age, and they seem in much less regular series. 47. Chonophorus mexicanus Ginther. West Indian fauna. XIV. Lepimocosivs Gill. §. Lepidogobius. 48, Lepidogobius lepidus Girard. Californian fauna. The specimens in the Museum are from San Francisco. AMERICAN GOBIIDA AND CALLIONYMIDA. 69 §. Hucyclogobius Gill. 49, Lepidogobius newberrii Girard. Californian fauna. XV. Microcosius Poey. 50. Microgobius signatus Poey. West Indian fauna. The types are in the Museum. 51. Microgobius eulepis spec. nov. No. 27,123, one specimen. Fortress Monroe, S.C. Mrs. C. N. Wil- lard. Length 13 inches. Head 4 in length (54 in total); depth 54 (7); D. VII—15; A. 16. Scales, 50 in a longitudinal series, 14 in a cross-series. Body elongate, scarcely compressed; head slightly higher than wide, the depth 14 in its length; eye large, longer than snout, 34 in head; snout 5 in head, rather broad, not pointed as in thalassinus; preorbital narrower than pupil; mouth very oblique, maxillary not extending beyond ante- rior margin of pupil, 2} in head; teeth in the upper jaw in a very narrow band, slightly enlarged in the outer series, largest towards angle of mouth; teeth of the lower jaw in a similar band, some of the outer ones in front long and slen- der. Seales cycloid, rather large, crowded anteriorly, regu- larly arranged, not imbedded, as in signatus, not deciduous as in thalassinus;: breast, nape and region along spinous dorsal naked. The first dorsal spine is equidistant from tip of snoutand first anal ray; longest dorsal spine 1} in head; caudal fin about 4 in body; ventral not reaching vent, equal to the length of head, the basal membrane one-fourth of its actual length: pectoral equals length of head. Color yellow or very lght brown, dotted with minute dark points above; scales along back with a dark margin; 70 CALIFORNIA ACADEMY OF SCIENCES. head and nape with minute points; spinous dorsal transpa- rent, a marked black spot on the upper part of the mem- brane between fourth and fifth dorsal spines; other fins plain; a light vertical bar on the posterior margin of preo- percle; no other bars or stripes anywhere. This species is closely related to Microgobius signatus Poey, and WMicrogobobius thalassinus Jordan & Gilbert. It differs from the former chiefly in the number of fin rays and number of scales; it differs from MW. thalassinus in having its scales firm, though not imbedded as in signatus, and in the length of the caudal and ventral fins. It differs from both in coloration. 52. Microgobius thalassinus Jordan & Gilbert. Atlantic coast of the United States. 53. Microgobius emblematicus Jordan & Gilbert. Panama. 54. Microgobius gulosus (Girard.) Gulf coast of the United States. XVI. BaARBULIFER gen. nov. Body entirely naked; numerous barbels around the mouth and on the chin. Dorsal spines 7. Otherwise as in (obio- soma. 55. Barbulifer papillosus spec. nov. No. 26,186, one specimen. Key West, Florida. Length 1 inches. Head 33 (42 in total); depth 4} (52); D. VIT—9; A. 9. Body short and robust, deepest below first dorsal spine; head blunt, profile straight from first dorsal spine to eye, much curved in front of eye; eye longer than snout, 33 in head; interorbital area two-thirds diameter of eye; snout blunt; mouth small, oblique; maxillary 3 in head, reaching to below anterior margin of pupil; lips thick. AMERICAN GOBIID4 AND CALLIONYMIDA. (a About 21 barbels, in length one-half orbital diameter or longer, are arranged as follows: a series of seven cross the snout from one angle of the mouth to the opposite angle, the anterior three on the snout rather thick and colored (two of them nasal), all the others yellowish, the barbel nearest each angle of the mouth longer than any of the others; on the lower jaw a barbel near each rictus, two on the chin, behind which are two pairs of barbels; posterior to these and below the rictus are two barbels on each side; one slender barbel on each side of preopercle below the posterior margin of the eye. Numerous rows of pores or papille on the head; one series extends straight downward on the anterior part of the opercle, from the upper end of which another series ex- tends perpendicularly backward; other pores irregularly scattered on the opercle; a double series extends along edge of preopercle, the pores becoming larger and especially conspicueus below, meeting on the chin; six or seven series radiating from eye, extending to snout, maxillary, and oper- cular series below; a row of pores nearly surrounding the mouth, curving backward, encircling the nasal opening; one series about the eye posteriorly, otherwise none on top of head or nape. Fins high and rounded; second dorsal higher than first, + in head, caudal very broad and rounded-—equal to the head in len:th; anal lower than soft dorsal; ventral reach- ing two-thirds to vent, 14 in head; pectoral 14 in head. Color yellow; upper half of body with a broad band of purplish spots; six diamond-shaped spots of darker cross the band, extending above and below it; nape, top of head and upper part of cheek covered with dark points; opercle light yellow, cheeks darker; an oblique bar of black points on upper half of pectoral base, a curved bar of fainter spots on base of caudal; fins otherwise colorless and transparent. A second specimen, also from Key West, with fins broken 72 CALIFORNIA ACADEMY OF SCIENCES. and the barbels and papille of the head more or less torn and obliterated, may perhaps be referred to this species, the proportions being the same. Gobiosoma ceuthecum Jordan and Gilbert, should be re- ferred to this genus. Barbulifer papillosus is most readily distinguished from ceuthecus by its deeper form, the depth of ceuthecus being 7 in its length. The barbels in ceuthe- cus, Dr. Jordan informs us, are quite small. 56. Barbulifer ceuthcecus (Jordan & Gilbert). Key West. XVII. Gosiosoma Girard. 57. Gobiosoma histrio Jordan. Guaymas. 58. Gobiosoma molestum Girard. Gulf coast of the United States. Bahia. The specimens in the Museum are from Louisville, Ky., Pensacola, Florida; and No. 13,143, one specimen 13 inches, from Bahin. This specimen agrees very closely with the specimens of Gobiosoma molestum from Pensacola, Fla., the fins being somewhat higher. D. VII—11; A. 11. Head 3 in length; depth 33. High- est dorsal spine 1} in head, highest ray 1}; caudal very broad and rounded, 1 in head; ventral reaching vent, 14 in head; pectoral slightly longer. Body highest below first dorsal spine, tapering to snout and to caudal peduncle; head slightly wider than high. Color yellowish, with traces of dark cross shades; head, body, vertical and ventral fins with minute points; a black spot on the membrane of dorsal between first and second spines; fins otherwise uniform in color; pectorals plain. 59, Gobiosoma bosci (Lacépéde.) Atlantic coast of the United States.. The specimens in the Museum are from Florida; Somer's Pt., New Jersey; Charleston, S. C.; Fortress Monroe; Amelia Island, Florida; Hilton Head, S. C. AMERICAN GOBIIDH AND CALLIONYMIDA. 73 60. Gobiosoma multifasciatum Steindachner. Gobius lineatus Poey. Mem. de Cuba, ii., 392. 1860. West Ind'an fauna. No. 13,317, one specimen. St. Thomas; Hassler Expedition. This specimen, seven-eighths of an inch long, agrees with Dr. Steindachner’s descriptions of G. multifasciatum. We have not been able to find the type of G. lineatus, but Poey’s description agrees with the specimen before us. As the combination, Gobius lineatus, has been used before, we have retained Steindachner’s name. 61. Gobiosoma zosterurum Jordan & Gilb:-rt. Panam. 62. Gobiosoma ios Jordan & Gilbert. West coast of the United States. XVIII. CLEVELANDIA gen. nov. Body long and slender; maxillary much produced, but not extending to the gill-opening; mouth horizontal. Dor- sal spines four, very weak. Minute cycloid scales. We have dedicated this genus to Daniel Cleveland, Esq., of the San Diego Society of Natural History. Mr. Cleve- land has done much towards making known the fauna and flora of Southern California. 63. Clevelandia longipinnis (Steindachner.) robiosoma longipinne Steindachner. Ichthyol Beitr. vili, 27. (Las Animas Bay, Gulf of California.) No. 13,151, oue specimen. San Francisco. Cary. Length 1g inches. Head 4 in length (4} in total); depth 62 (7). D. IV—16; A. 17. Scales in a median series, about 70; in a cross-series, about 18. Body very much elongate, slender; head long and slen- der, depressed anteriorly much as in sox; profile straight; eye moderate, slightly shorter than snout, 44 in length of head; interorbital area about as wide as pupil; ante-orbi- 74 CALIFORNIA ACADEMY OF SCIENCES. tal area scarcely half diameter of eye; mouth large, maxil- lary extending much beyond orbit; lower jaw flat, slightly curved upward anteriorly; mouth very much as in Ksoz; teeth all small, in narrow bands in each jaw; the outer ones of the upper jaw slightly larger than the others. Scales minute, slightly enlarged posteriorly; their mar- gins plain, the anterior part of the exposed area lengthwise striated; breast and ante-dorsal area naked. Distance from snout to insertion of first dorsal spine is contained 23 in body; the spines slender and short, 3 in head; inter-dorsal area equals snout and eye; dorsal rays slightly longer than the spines, the last ray not extending half way to caudal; caudal pointed, scarcely shorter than head; ventrals not reaching half way to vent, 1? in head; pectoral 13} in head; vent slightly behind middle of body. Color light brownish; numerous darker spots of aggre- gated points along nape and upper half of body; belly white; head slightly darker than body; posterior edge of opercle white; an oblique silvery bar on the lower half of opercle, and a light blotch at the upper corner of opercle; cheek with black points; some light areas below eye; lower surface of head and posterior part of maxillaries plain; two dark bars on spinous dorsal; second dorsal with three or four dark bars; a curved black bar at base of caudal; re- mainder of caudal irregularly barred with dark; other fins plain. The specimen we have described agrees well with Stein- dachnev’s description, except that he states that his speci- mens were entirely naked. XIX. GuILLIcHTuys Cooper. 64. Gillichthys mirabilis Cooper. Californian fauna. Or AMERICAN GOBIIDA AND CALLIONYMID”®. XX. TypHnoconius Steindachner. 65. Typhlogobius californiensis Steindachner. Californian fauna. Sub-family GoBIoIDINz. XXI. Tynrnastes Ginther. 66. Tyntlastes brevis Giinther. Panama. 67. Tyntlastes sagitta Giinther. Panama. XXII. Gostoiwres Lacépede. 68. Gobioides broussoneti Lacépéde. West Indian fauna; south to Rio Janeiro. Specimens of this species were collected by Hartt and Copeland at Sao Matheos, Rio Janeiro, Rio Pard. There is also in the collection the type of Cobioides barreto Poey from Havana. 69. Gobioides peruanus (Steindachner). West coast of South America. Steindachner, Fisch-Fauna des Cauca & Fliisse bei Guayaquil. 42. 1880 (Guayaquil). Head 5 in length; depth 11. D.7,17. <A.TI, 16. Ante-orbital portion of head 23 in post-orbital portion; inter-orbital 5 in head; lower jaw slightly projecting; a series of large slender teeth in each jaw, behind which in each jaw is a narrow band of fine teeth. Caudal 44 in body, connected with both dorsal and anal by amembrane. Dorsal rays violet, the membrane yellowish; narrow, angular cross- bars on body. XXIII. Cayvennia Sauvage. Body much elongate, dorsals united, caudal free from doisal and anal; ventrals united, not adhering to belly; 76 CALIFORNIA ACADEMY OF SCIENCES. teeth small, outer enlarged; anterior part of body naked, posterior part covered with cycloid scales. 70. Cayennia guichenoti Sauvage. Cayenne. Sauvage. Bull. Soc. Philom. Series 7, iv, 57. 1880. Head 9 in total; depth 17; D. VI, 17; A. 1, 16; vertebrx about 36. Head deeper than wide; eye small, placed well forward; maxillary reaching to below posterior margin of eye; a low membrane connecting dorsal and caudal; caudal 7 in leneth; ventrals 14 in head; brownish, marbled with black an- teriorly; length 0,400 (Cayenne). Perhaps the most interesting fact connected with the study of the American Gobies is that six of the seventy species, distributed in four, probably five genera, are half naked. Three of these were known at the time Ginther’s catalogue was written, allfrom Panama. Since then, Sauvage has discovered another half naked species at Surinam; a sixth species, whose habitat is unfortunately not definitely known, has been added in this paper. The American Gobies fall into four more or less natural groups—Oxymetopontine, Hleotridine, Gobiine and Gobioidu- ne; and what is most remarkable, three of these sub-families have each one or more half naked representatives. How widely these half naked species may be distributed cannot be told from the few data. Nothingis known of them but that they exist. An explanation why they are found only in the waters of Central America and the West Indies, should not be attempted until it has been proved that they are con- fined to these waters. Family CALLIONYMIDA. XXIV. Catuionymus Linnezus. 71. Callionymus calliurus spec. nov. No. 26.265, one specimen. Off South Beach, Key West, 5 fathoms. Pourtales, 1869. : AMERICAN GOBIIDH AND CALLIONYMIDA. Th Head, to tip of opercular spine 34 in length (5 in total); depth 7 (9). > D. TV—6; A. 4. Body flat below, the ventral surface bordered on each side with a fold of skin which is wider than the pupil; a single lateral line; diameter of eye equal to length of snout, 33 in head; maxillary not extending to the eye; opercular spine with two barbs above, the anterior one larger and turned forward; gill-opening a minute foramen opening upward. The last dorsal ray equals the length of the head, and the first dorsal spine reaches its tip when the fin is depressed; ventral fins connected by a broad membrane to the middle of the outer pectoral region; pectoral fins.as long as the head. Cheeks, opercles, connecting membrane of ventral fins and ante-pectoral region with milk-white spots; lower jaw black near the rictus; a series of black dots on branchios- tegal membranes, one or two similar dots in front of pec- torals, two on the cheek forming a series with the second branchiostegal spot; four black spots on the marginal mem- brane of the belly, other black spots above it; lower half of body with numerous dirty-white spots; pectorals trans- parent, ventrals dusky; membrane of anal sprinkled with minute black points aggregated into black spots in places, and with opaque white spots; caudal transparent, having minute points, its upper half with opaque milk-white bars running obliquely downward and backward from ray to ray; lower half with interrupted longitudinal lines of opaqué white, alternating with black spots; dorsal transparent, with white and black dots most conspicuous between last rays; body marbled with light and darker. This species is nearly related to Callionymus pauciradiatus Gill. The only description of that species is: ‘‘D. III, 6, A. 4. The preopercular spine is armed with three teeth above and terminates in an acute point.’’ Ann. Lyc. Nat. Hist., N. Y., vili, 143, 1865. (Matanzas). The speci- 78 CALIFORNIA ACADEMY OF SCIENCES. men before us has four dorsal spines and only two barbs on the preopercular spines; how they may agree or differ in other respects we are unable to tell. Until Dr. Gill’s type is more fully described we feel justified in considering the specimen in the Museum the type of a new species. 72. Callionymus pauciradiatus Gill. West Indian fauna. 73. Callionymus bairdi Jordan MS. “D. IV, 9. A. 8. Preopercle with an antrorse basal spine above and 8 sharp teeth near the tip below.” 74, Callionymus agassizii Good & Bean MS. DESMIDS OF THE PACIFIC COAST. 19 DESMIDS OF THE PACIFIC COAST. IDENTIFIED BY REV. FRANCIS WOLLE. Collected at Donner, Truckee, and Reno, Nev., Septem- ber, 1887. The following is a list of the plants found, not contained in the report published in Bulletin 7, June, 1887: 1. HyYALorHEcA DISSILENS Breb. 2. SPHHROZOSMA PULCHRUM Bail. 3. PENIUM INTERRUPTUM Breb. 4. CLOSTERIUM ACUMINATUM Kg. 5. ANGUSTATUM Keg. 6. STRIOLATUM Khrb. iB Detponti Klebs. 8. Doctpium TRABECULA Naeg. 9. CaLOCYLINDRUS coNnNATUS (Breb.) Kirch. 10. Cosmarium Broomet Thwaites. 11. CosmariumM OvaLE Ralfs. 12. Evastrum GEMMATUM Breb. 138. AMPULLACEUM Ralfs. 14. vERRUCOSUM (Ehrb.), Ralfs. 15. ANSATUM (Ehrb.), Ralfs. 16. STauRASTRUM comMuUTATUM, Kg. ive CORONULATUM, Wolle. _ Beside these desmids, the vials contain the following forms: 18. SprroGyra GREVILLEANA (Hast), Kg. 19. TENUISSIMA, Kg. 80 CALIFORNIA ACADEMY OF SCIENCES. 20. ULOoTHRIA SUBTILIS, Kg. 21. STAUROSPERMUM QuUADRATUM, Kg. 22. OSCILLARIA PRINCEPS, Vauch. (Vauch). 23. ZYGNEMA STELLIUM, Ag. sterile. 24. CEDOGONIUM The following, already noted in Bull. I, No. 7, plicate in the Nevada vials: HYALOTHECA MUCOSA. PENIUM DIGITUS. CLOSTERIUM ROSTRATUM. COSMARIUM CRENATUM. CUCUMIS. ORNATUM. PACHYDERMUM. TETRAOPHTHALMUM. TUMIDUM. XANTHIDIUM CRISTATUM. EUASTRUM BINALE. ELEGANS. STAURASTRUM ARCTISCON. AVICULA. DEJECTUM. DICKIEI. ECHINATUM. INCONSPICUUM. MUTICUM. SEBALDI. are du- BROMINE IN SEA WATER. 81 DETERMINATION OF BROMINE IN SEA WATER BY FRACTIONAL TITRATION. BY F. GUTZKOW. Read befor? the California Academy of Sciences, February 6th, 1888. This method of quantitative determination of bromine consists of three operations— 1—Separation of the bromine as cuprous bromide. 2—Conversion of the cuprous bromide into zinc bromide. 3—Titration by hypochlorite of sodium. 1st OpERaTion.—Separation of the bromine as cuprous bro- mide.—I mix 250 ¢c.c. of filtered sea-water from the Pacific Ocean, to which a drop or two of sulphuric acid has been added, with 100 ¢.c. of a solution containing 25 grm., more or less, of crystallized cupric sulphate. To this clear solu- tion, in which the copper may be assumed to exist as cupric chloride, I add from a graduated glass a solution of ordinary good sodium sulphite, the strength of which I need not know, until the brown flocculent precipitate which each addition produces will dissolve more and more slowly. It is easy to find a point when the brown precipitate has dis- solved by digesting, but the solution remains slightly turbid from shining crystals of cuprous bromide. The separation ot crystals increases rapidly, cuprous chloride being also separated. An excess of sulphite of soda does not matter much. By heating to about 40° C. the green liquid be- comes blue again. Then, after cooling by water, another addition of sodium sulphite is made, say one-third of the volume previously used, again heated until the blue color distinctly reappears (about 70° C.); the flask is once more cooled by water and allowed to settle, after removing any traces of sodium sulphite remaining in the neck by shaking 2D SER., VOL, I, Issued Feb, 28, 1888. 6 82 CALIFORNIA ACADEMY OF SCIENCES. with the liquid. It is advisable to loosely cork the flask. It will be found that the precipitate of mixed cuprous chlo- ride and bromide is an unusually heavy one, settling rapidly from a perfectly clear liquid, from which any little particles floating on the surface can be made easily to sink (by a glass rod or judicious shaking), so that the liquid may be decanted to almost the last drop after short settling. The solution has become strongly acid by the sulphuric and sulphurous acid formed through the reaction between cupric chloride and sodium sulphite. For the 250 c.c. of sea-water employed about 3 grm. of copper have been sepa- rated, varying in amount somewhat according to the tem- perature after heating and after cooling. Heating to the boiling point only increases unnecessarily the copper sepa- rated. The second addition of sodium sulphite I found necessary for removing the last trace of bromine. Now, in order to test if all bromine has been separated, I add to the decanted liquid another portion of sodium sulphite, say as much as the secund time, heat and cool as before, treat the precipitate by zine and hydrochloric acid, as will be described hereafter, filter some of the solution of zine chlo- ride obtained into a test tube and test by one drop of the standard chlorine solution and chloroform, whether after shaking, the chloroform turns yellow. If it does not it is certain there is no trace left in the decanted liquid. By these operations bromine can be very conveniently and completely separated in about ten minutes. Enough cupric sulphate ought to be added to form cupric chloride also from the sodium sulphite, as this is partially converted into sodium chloride by the formation of cuprous salts. In analyzing saline solutions other than sea-water, I ascertain the specific gravity and the corresponding percentage in sodium chloride from the tables, and reckon three parts of crystallized cupric sulphate for one part of sodium chloride, assuming all dissolved salts to be sodium chloride, as an excess of copper does not matter. BROMINE IN SEA WATER. 83 2D OPERATION.—Conversion of the cwprous bromide into zine bronude.—This is accomplished by zine with the addi- tion of a little hydrochloric acid. The operation is an easy one, and would seem to require no explanation. But for the 3d operation, the titration, I require a small volume, say 25 c.c., in our case. A judicious economy of water and acid is, therefore, imperative. The precipitate in the flask is di- gested with about 100 c. c. of cold water, to which about 1 grm. of sulphuric acid has been added. The settled liquid is decanted as before, the last portion over a very small filter. Any iron which the cupric sulphate may have con- tained is now reduced to a no longer appreciable amount, and the sulphurous gas much reduced. But in order to destroy the last trace of the latter, I add a few drops of sodium car- bonate and spread the crystalline powder around the flask. After one or two minutes enough oxychloride will be formed to answer the purpose afterwards. Now TI rinse the con- tents of the flask into a porcelain dish, using only a few cubic centimeters at one time; heat in a test-tube about 5 ¢c.c¢. water, with $ c¢.c. hydrochloric acid, and filter over the small filter I had been using into the flask. A little more water applied to filter and flask will clean both sufficiently. Thedish is heated on a water bath, and when by smell I detect no sulphurous gas, I add one large amply sufficient piece of zinc. The white crystals will soon be replaced by copper and a clear solution of cuprous chloride in hydrochloric acid. The finished reduction I recognize by touching with the point of a glass rod first the liquid and then a drop of sul- phuretted hydrogen water spread on a porcelain plate. Enough acid must be present to prevent the formation of oxychloride of zinc. The contents of the dish, copper and all, are filtered over a small filter, directly into a small graduated cylinder, and ought not to give more than 10 or 15 c.c. filtrate. The copper is then fully sweetened into another porcelain dish with about 25 ¢.c. of hot acidulated 84 CALIFORNIA ACADEMY OF SCIENCES. water and the filtrate evaporated to about 5 or 6 ¢.c., with the addition of one or two drops of sodium carbonate, then made acid again and added to the 10 or 15 in the graduated cylinder, and the water used for rinsing the dish utilized to fill up to 25 c.c. 3D OPERATION. — Titration by hypochlorite of sodium.—I must suppose that the chemist who takes interest in the subject treated here is acquainted or makes himself ac- quainted with what Fresenius in the sixth edition of his celebrated book remarks on titration of bromine. He will find that all these titrations are based on liberating bromine by chlorine and on the color imparted by free bromine to water, or chloroform or other absorbent. My method offers nothing new in that respect. Generally chlorine water is used as standard solution, which is the most changeable liquid employed in volumetric analysis. Figuier removes the bromine liberated by boiling; Reimann by chloroform. In the first case it requires better eyes than mine to recog- nize coloring when there are only one or two milligrams of bromine left, and in the second, frequent removal of the chloroform is necessary; for, while the decoloration of a small drop of chloroform slightly colored is a sure and deli- cate test, the change of color ina highly colored drop of chloroform is a very coarse one. As standard solution I use hypochlorite of sodium or po- tassium, that is, ordinary ‘‘ Kau de Javelle,” as prepared by druggists (by treating ‘‘chloride of lime” with sodium or potassium carbonate), containing, generally, from one to two per cent. of chlorine, which may be set free by an acid. The slight excess of the carbonate employed does not interfere. I dilute one part of the commercial liquid with five parts of water, and test ils strength by a normal solution of potassium bromide, acidulated by sulphuric acid. This normal solution I prepare by dissolving 1 grm. potassium bromide in water, add diluted sulphuric acid containing about 2 grm. SO, and fill up with water to say one-half liter. BROMINE IN SEA WATER. 8) Then I ascertain the bromine in a given volume either by argentic nitrate and chlorine gas or by titration with the standard solution of hypochlorite of sodium, which has been previously assayed for chlorine by one of the usual methods. Deducting the volume used for assay I dilute until the normal solution contains one milligram per cubic centimeter. Although the standard solution of hypochlorite of sodium is by no means unalterable, still it is much less changeable than chlorine water. [For instance, a solution prepared three weeks ago and indicating 100 milligrams of bromine by 18.8 c.c., requires to-day 19.6 c.c., weakening by 4 per cent. This weakening is, however, no matter, as it amounts to only .01 c.c. per day for 25 miligrm. of bromine, and I recom- mend, in all cases, to test its strength immediately previous to the assay on the unalterable normal solution of bromine. This is preferable to its test by an iodide of potassium, ete., solution, which Fresenius (Quant. Anal.,§ 145, %) recom- mends, deviating in this one little instance from the funda- mental principle of volumetric analysis expressed by him elsewhere, that the test ought to be made, if possible, al- ways in the same manner as the assay. The novelty of my method is a kind of fractional titration which allows to foretell the final reaction on an aliquot por- tion of the liquid, and—in the case of bromine—to compare the color before and after addition of the chlorinated standard solution. This kind of titration is advantageous also for other determinations, especially those which have a very sudden final reaction, as, for instance, the analysis of caustic soda. For the last named kind of analysis I proceed as follows: The apparatus consists of a flask provided with a well-fitting cork with two holes. Through one of these holes passes a little bent glass tube, stopping short at the cork inside and connecting outside with a rubber tube and clamp or ‘‘clip.” Through the other hole passes a long funnel to nearly the bottom of the flask. I prefer the kind 86 CALIFORNIA ACADEMY OF SCIENCES. of funnel called thistle-shape, or, better, poppy-head shape. It must hold 30 to 40 c.c. A mark by file or paper glued on shows when the funnel is filled with 20 c¢.c. I measure the solution of caustic soda to be tested and dilute to a round figure, say 100 c.c., add litmus and fill into the flask, opening the clamp. By blowing through the tube I raise 20 c.c. into the funnel and close the clamp. I note the stand of the burette, say = 0.00 c.c., add one drop of stand- ard acid; the solution remaining blue, I may add four more drops for the 80 c.c. in the flask, because + = 20 ¢c.c. were not changed. I do so; find that I have been using .25 c.c. in five drops, that the liquid remained blue, and that I may add safely 4 X .25 or 1.25 ¢.c., including the five drops al- ready given. I bring the burette to 1 c.c. to get at round numbers; still blue color. I may add four more c.c., and do so, } ¢.c. at a time, stirring with a little glass rod which I leave always in the funnel, and observe that with 3 c.c. the contents remained blue, but become red with 5}.¢.c. Now I know that the final reaction w:ll occur between 3 X 5 = 15, and 33 X 5= 173 c.c. LIopen the clamp and run standard acid to the 15 c.c. mark of the burette, shake and wash the funnel by raising and lowering repeatedly the mixed liquid. I raise again to 20 ¢.c. in the funnel, close the clamp and begin my second test, adding, as before, 1 drop; the liquid remaining blue, [add 4 more. Still blue. I may, conse- quently, give 5 X .25 ¢.c. = 1.25 ¢.c. I doso, but only .10. c.c. or 2 drops at a time, because I know that I am ap- proaching the final reaction below 17.5 c.c. Having added .30 ¢.c., the liquid remains blue, but turns red at .40 ¢.c. I may add 5.x .80 but not 5 X .40c¢.c. I open the clamp and proceed as before, bringing the burette to 16.50 ¢.c. I raise again 20 c.c., which I still, to be on the safe side, count as ;, although it has become now nearly ith of the whole. I add one drop of acid: blue, another: red. Igo no further, because I know I am within 3 drops of the final reac- tion. I remove cork and funnel and add the next drop BROMINE IN SEA WATER. 87 directly to the flask, find that the third drop will turn red, and read 16.9 ¢c.c. off the burette. The following table will show how the observations may be conveniently noted, + and — signifying respectively blue and red: ADDITION STAND OF To z T0 THE WHOLE. Rinne: MADE. 0,00 c ec. ALLOWED. FORBIDDEN. MADE. aa drop | |=) 0108 =.20 6.Ce [ioe case) || eels owe || fesle ts vinte .25 cc. | UV COM! scoacoos || oogntiedde leobuamoode .50 SO Sete so ococrd Nines cococ ll qoceMeoc om ilenca cratic 1.00 + HOO I tosonveed] occas aanel || aosncococ 3.00 -- 100) |) ak'derads 15.00 c.c¢ 15.00 3.50 | SSuhisocgee aobebr NAT HOS I soasoosoA.|) Gabe Sader No. 2.. | ldrop | + 5 asl Wines hearer Meee Ploralcs at ears S18 Se eee |e eacige S heal (eicecicg ae fetes ron ate as .20 + UU IW Acbacdée || aosospccd |) Moccecnes .30 ol 1a) ||. Goemasee 1.50 16.50 .40 = SeepGoscasns PAU. Haas Sorc orca huctaeeicie pitet No. 3.. | ldrop | + a ewe ll hee orth Bec 25 16.75 -- 3 drops 16.90 c.c. For the analysis of bromine I modify my apparatus as follows: The flask holds about 70 c.c. and has a round bot- tom, if possible, ege-shape, and short and wide neck, so that a cork with three borings may be fitted. One of these holes serves for a small bent glass tube, with rubber tube and clamp attachment as described above. Instead of one funnel I have two, as much as possible of equal size and of similar shape as described, each holding 10 or 15 ¢.c. Into one of the two cups I place a short and thin glass-rod, bent so that the heaviest part is inside the cup, and graduate (with the little glass-stirrer inside) for one and for five cubic centimeter. The tube of the other funnel is bent above the cork, so that 88 CALIFORNIA ACADEMY OF SCIENCES. both funnels will rest side by side in equal and least possi- ble height above the cork. Both tubes are, now, cut off, so that they reach nearly to the bottom of the flask. They ought to be as narrow as a reasonably fast flow of the liquid from flask to cups will allow. As I could not procure fun- nels to satisfy me in this respect I shoved a narrow tube through the straight tube, flanging it a little so that it found a support in the conical bottom of the cup, and closed the lower opening of the bent tube partially. Thus, when the air is compressed in the flask the liquid will travel in both tubes with tolerably equal speed. The graduated cup with the straight tube serves for fractional titration, the other one to compare the original color with the changes of color pro- duced by the additions from the burette to the graduated cup. Everything being prepared, I pour 25 c.c. of my normal solution of potassium bromide through one of the funnels into the flask, raise 5 c.c. or 1 to the corresponding mark in the cup, close the clamp and assay in similar manner as described above. No. 1.—I add one drop of the solution of hypochlorite of sodium of unknown strength. A yellowish tint is no- ticed, faint but unmistakable. I am allowed to add four more drops. I doso, and observe deepening of the color with every drop. Having, thus, spent .25 c.c. 1 am allowed to add 1 c.c. more or 1.25 ¢.c. in all. I add .25 c.c. at one time, stir and observe that the color becomes more intense by stirring. The next .25 c.c. do notseem materially to in- crease the color although it does not decrease. Therefore, I add only .10 ¢.c. at one time, note .85 as doubtful, observe with the next drop a probable decrease of color, which de- crease becomes more marked with the second drop, and quite decided and unmistakable with the third. I know, now, that the assay will be finished between 5 x .75=3.75 c.c., and 5 X 1=5c.c., probably already 5 X .90 = 4.5 ¢.c. I lower the stand in the burette to 3.75 c.c., which makes BROMINE IN SEA WATER. 89 the contents of the graduated cup nearly as light as those of the other cup, open the clamp, mix and wash the cups as described above. As I know that I am very near the finish, I raise only to the 1 ¢.c. mark. No. 2.—I add one drop: the 1 c.c. gets quite light compared with its partner in the other cup. I blow, now, into my rubber tube, open the clamp while blow- ing, in order to prevent the sample from sinking back into the flask, and raise to the 5 ¢.c. mark. This permits a better observation than adding the drop at once to the 5 e.c. I compare the two colors at my leisure. The result is doubtful. The color seems to have neither increased nor decreased. The next drop shows a decided decrease of color on comparing both cups. I cannot go any further with fractitional titration, being, prob- ably, within three drops of the end of the assay. I open the clamp, remove the cork and funnels after washing rod and cups with water, and boil the contents of the flask until the color has become ‘quite white and no more smell of bromine can be noticed. I cool by water, add one c.c. of chloroform, close the flask with my thumb, shake vio- lently and observe whether the chloroform has become col- ored. If not, as it will not be if all the bromine has been evaporated, I add one drop from the burette and shake as before. In the settled little drops of chloroform (the advantage of a pointed egg-shape of the flask will show now) I recognize distinctly a yellowish tint. The next drop leaves me uncertain if there is any change in color. The third drop shows without fail an almost complete decoloration. The assay is finished. By deducting the last drop and one- half of the former drop, I learn that 25 milligrm. bromine are indicated by 3.925 c.c. of my standard solution. Hach drop of .05c.c. will, henceforth, disclose (*:33% being = 78.5) 7s or about $ of a milligram of bromine. The limits of accuracy depend of: (1) the least possible number of drops for testing by chloroform; (2) the least possible amount of 90 CALIFORNIA ACADEMY OF SCIENCES. chloroform, that is: (3) the least possible volume of solu- tion, because, the more solution the more chloroform must be given to obtain drops of sufficient quantity for a good observation of color. The following table will record the results of this assay in a better shape, + expressing increase, — decrease of color, ? doubtful. ADDITION STAND OF To 1 To THE WHOLE. BURETTE. ; SESS ALLOWED. | FORBIDDEN. MADE. 0.00 c.c. { |1 drop = D) Meee INO MAS See ches Viana + BD IGECes Il) torsheeeies s. | erete ere aera | |e eee .25 SB 25) eee se | eee .50 2250" OSes Ne oe 75 SF STB edi Rata es cere 85 Rol cc ctchoss 3, eee | eapoeySatoro-caalll oi steenaaers a 1.00 es eee DOO G6. || sisihe-sices JM ete See ait 51 ill, once ete aiate eu Ml watashae er aea| CO moti Call MeeanttCel INOR 2) Sate wake -05 eal pees ene’ [05.8 > Re 1051 tees) ak 5 } aoe ale 3.85 IN OU oes ects eel eee eee aa WITH CHLORUFORM: | .--.ee-- eee a .05 ae ee = Saye ? .05 Seen AAR Aa — .05 4.00 ete: less the last 144 drops: 075 3.925 ¢.¢. It will be observed that the one dangerous calculation in this titration was the fourth of No. 1, which brought me on a not perfect observation too near to the result. I choose intentionally this example for pointing out the danger of adding too much from the burette on the observation of the first fractional titration, when the second cup contains still colorless liquid and offers no chance for comparison of colors. The increase of color becomes easily fallacious BROMINE IN SEA WATER. 91 after a certain intensity has been obtained, but not the de- crease. The record of ‘‘addition allowed to the whole” ought to be based on certainty, which may be won by an- other fractional titration. The following table will explain how the titration ought to have been made: ADDITION STAND OF To 1 To THE WHOLE. BURETTE. ee ALLOWED. |FORBIDDEN.| MADu. 0.00 c.c. iy { RO ele henola MING. 22.<tiers ort sly wuton .25 STRAUS EMSs Sarae tere wctey (Paw eee omens 90 See tOU MME Ae ohokekte. Soe Eaters 75 Bi Bee Cee (MOI ae | A PPeeee SANE me hee fe eke 65 E\ \pRGoa ose 1 adercarcctic| WER Moxie a Sut EF Reis sa 1.00 Sa ile & Sans aere DOORGC tT io ol incn ae eee MOGEaOe 39-|) aaoageod || o5cooodo. || CAN eyos al Chal: IN Os CASA BAOCOOEE .05 a PEE || ancuctcarel PRM eel tm ake Mae .10 =. ADT ys goth eae Bote || Mase ee Semele cadre .20 =i (ele OO) ame uated | Wee eceee es tl thes ee ll Mere ee .30 ee aera eral ert rete tet || hey cees hare lh were epee ee 40 =) |b es Soest DROOE = SR Aas fet Oe eae Ss Beers S504 Shemini ced wohl eee a eet OL 0) 3.50 INO essere eee ss 05 + OE OF Ic ena Wika as Sel eal Ge nae 10 DI erreyater ye || eieieia ses ha | cree etn. will Saw ate Sake: 15 =P ILS cat ceeiraleievs OWE eed erselereraerevsc|lisce an atcre se eee 225 3.75 etc., like No. 2 of the foregoing table. To sum up, this kind of titration is based at the begin- ning on coloration, and toward the finish on decoloration, of the fraction in the cup. As long as three drops are re- quired tojproduce a decoloration, another fractional titra- tion may be safely undertaken; but when already two drops change the color decidedly, it is better to stop and proceed to the final test with chloroform. These remarks refer, of course, only to a standard solution of about the strength indicated here. 92 CALIFORNIA ACADEMY OF SCIENCES. The titration of the 25 c.c. obtained from the 250 c.c. of sea-water is done in exactly the same manner as that of the normal solution of potassium bromide, except that the ob- servation of the first drop from the burette requires special attention. In the presence of iron (from the zinc) or cop- per, both being under the circumstances ferrous or cuprous salts, no bromine will be set free until they have been con- verted by the chlorine into ferric and cupric salts. There- fore, before commencing on the fractional titration, I empty the 25 ¢.c. into the flask, add one cubic centimeter chloro- form, also one drop from the burette, shake, and observe whether the chloroform shows coloring. If it does, I evaporate the drop of chloroform by boiling the contents of the flask while keeping them in agitation, cool, return the solution to the graduated cylinder, fill up to 25 c.c. and titrate in the manner described above, adding the one drop from the burette used for coloring the chloroform to the account. The more experienced analyst. will, however, be able to recognize the first color without requiring chloroform, and thus avoid boiling the solution previous to titration, which may cause a slight loss by evaporating some hydrobromic acid, unless sodium carbonate is added. For the titration, the solution has then to be made acid again. Whilst by this preliminary test traces of iron, manganese, copper and sulphurous gas can be detected and made harmless, the presence of hyposulphite of zinc (formed by the action of sulphurous gas on zinc) affects the assay more seriously, as it is less readily oxydized, completely only when the titrated solution is boiled to evaporate the free bromine. This causes formation of hydrobromic acid—hence the necessity of guarding against the presence of sulphurous gas before zinc is added. Every analyst will judge what other sub- stances may be present in his solution which might inter- fere, and how to remove them. It remains to state that the titration as described may also BROMINE IN SEA WATER. 93 be applied to a solution of zinc bromide and chloride ob- tained by reducing a precipitate of argentic bromide and chloride by zinc. But more zinc will be required than for the reduction of cuprous bromide and chloride. Argentic nitrate will precipitate 5 grm. of chlorine from 250 c. c¢. of sea-water, while in the precipitate obtained in the manner described there are only 1} grm. For this reason, and on account of the comparatively large volume of the argentic chloride, it is less easy to obtain a small volume of solution for titration. Fehling’s method of partial precipitation by argentic ni- trate, with its difficult settling and sweetening of the pre- cipitate, and absence of a ready test for ascertaining whether the bromine has been completely separated, I leave out of consideration here. The amount of bromine in the sea water near the en- trance to the Bay of San Francisco, determined in the man- ner described, is 67.5 milligrams in one liter. 94 CALIFORNIA ACADEMY OF SCIENCES. CHANGES IN THE FAUNA AND FLORA OF CALIFORNIA, BY H. H. BEHR, M. D. A residence of nearly forty years in this vicinity, during which the country has been rapidly settled, has necessarily made me a witness of many changes inthe Flora and Fauna, and it seems to mea matter of importance that some record of these changes be made. Such records of the physical changes abound and are of very great interest, but the mu- tations of organized life are certainly equally so, and they have had no historian. From time to time, in the progress of a busy life, I have taken notes of various such changes, and propose to give them to the Academy in the shape of a series of short papers. 1, Serpents. The venomous serpents of this vicinity, of which only one species (Crotalus lucifer) occurs, have, as is well known, ‘very greatly decreased in number, indeed it is probable that very few, if any, now exist on the peninsula of San Fran- cisco. They are, however, still found in considerable num- bers on Tamalpais, that wild and only half-explored mount- ain at our doors, and in the range of hills back of Oakland and Berkeley, and in certain sections there has certainly been a local increase in their numbers. As two of the places where they are increasing are to me well-trodden grounds, I have made some investigations of the causes of this increase, which is sufficiently marked to be noticed by casual visitors, and of course still more by residents of the districts. Statistics in such matters being out of the question, I have only to mention the reasons which have convinced me of the local increase, to which my attention was first called 2p SER., VOL. 1. Issued Feb. 28, 1888. CHANGES IN FAUNA AND FLORA. 95 by resident friends. First—Comparative frequency of snake bites to cattle in one locality and to dogs in the other. Second—The appearance of very young rattlesnakes, so called ‘‘one-buttons” before dark. I do not know the habits of the Crotalides of other regions, but our Crotalus lucifer, is, in the first year of his existence, even more noc- turnal in his habits than full grown specimens, and the ap- pearance of one before sunset is always exceptional. The first locality to which I refer is in the Contra Costa mountains, whose eastern chain is there broken by a deep ravine, a transverse valley carrying the waters of the table lands into the alluvial plains about Alvarado. My authority is a landed proprietor, blessed with a taste for zodlogy, as far at least as that branch of science is con- nected with sportsmanship. The greater part of the land surrounding his place is very hilly, and the declivities partly naked and rocky, partly wooded, only accessible by a few trails, most of them im- practicable for saddle-horses. The upland, used only for grazing, is open grass-land, dotted occasionally with live- oaks, but having’no chapparal whatever. There is but little agriculture in the immediate neighbor- hood, but a considerable number of ‘‘ chicken ranches,” the owners of which carry guns and possess dogs of some- what roaming dispositions. There is but little game left excepting quail. Hawks, eagles, owls and herons, were formerly abundant, but have perceptibly diminished. The grazing country seems to contain few reptiles excepting rat- tlesnakes. Ophibolus, Pityophis (magnificient specimens), and several species of EKutainia are found in the gulch and along the sides of the creek, but no Crotalus, as far as the observations of my friend extend. The pasture grounds of the upland are well stocked with cattle and horses. There are but few goats, and at present no sheep, and none of the farmers keep hogs to any extent. 96 CALIFORNIA ACADEMY OF SCIENCES. The other locality to which my attention has been spe- cially called is in the Coast Range. Its drainage is partly to the Bay of San Francisco, partly into the ocean. The elevation varies from 1,000 to 3,000 feet above the level of the sea. T have several statements from inhabitants of this district, all agreeing as to the fact of an increase of rattlesnakes, but differing in regard to the causes. The informant upon whose statement I chiefly rely has seen a great deal of bor- der life, some military service, and has made explorations in Mexico and the Indian Territories. His observation is keen and is the more reliable because not influenced by preconceived ideas. The lower part of this district must have been originally a dense forest of redwood (Sequoia sempervirens), Douglas spruce (Pseudotsuga Douglasii), madrono (Arbutus Menzesir), and chestnut oak (Quercus densiflora); the upper part, chap- paral (Ceanothus and Manzanita) and chemisal (Adenostoma), varied occasionally by groves or forest strips of Pinus in- signis and Pseudotsuga Douglasti, live-oak (Quercus agrifolia), and Kelloge’s oak (Quercus Kellogit). At this day tracts of cultivated land, chiefly vineyards and orchards, form frequent interruptions of the original forms of vegetation. There is still game enough to cause much inconvenience to the vine-growers, but it has, accord- ing to the statements made to me, perceptibly diminished. The hills are steep, but not so much so as to prevent the growth of chapparal upon them; the gulches accessible, but not broad enough to be of much use to the agriculturist; and though there are uo large water-courses, there are plenty of springs. Among reptiles, Gerrhonotus and Phrynosoma are the most common. The serpents are not well represented. I have not seen Ophibolus; the blue racer (Bascanion) has been only occasionally seen; Pityophis is exceedingly rare, and even CHANGES IN FAUNA AND FLORA. 97 the species of Kutainia, otherwise so universally distributed, are not often met with; the rattlesnake has, however, dis- tinctly gained ground. The population consists chiefly of the owners of large tracts of land, most of them in easy circumstances, and employing a good deal of manual labor in orchards and vineyards. The farmers keep some cows and horses on their premises or a neighboring hillside, but the region can hardly be called a grazing country. The professional sportsman or pot-hunter rarely visits the district, but boys and dogs of roaming disposition abound, and account in some measure for the diminished number of eagles, hawks and owls. Let us now consider the causes which have led to the increase of so very undesirable a neighbor as Crotalus luci- fie First.—The protection given to it by the nature of the country; the inaccessible hillsides and uninhabited upland abounding in gopher and squirrel holes of the Contra Costa, and the impenetrable thickets of chapparal in the Coast Range. These conditions change only very slowly with the settlement of the districts. Second.—Destruction of its natural enemies. There are very few persons carrying a gun who can resist the temptation of killing an eagle, a hawk or an owl, and many of them are probably ignorant of the fact that by so doing they preserve a rattlesnake or even a whole brood of them. Eagles and hawks will, I know, occasionally help themselves to a chicken or a young turkey, but their war- fare is usually directed more against gophers, squirrels, snakes, etc., avoiding as long as possible the dangerous vicinity of the hencoop with its hysterically noisy inmates and attendant shot-gun. 2p SER. Vou. I. Issued Feb, 28, 1888. 7 98 CALIFORNIA ACADEMY OF SCIENCES. Owls are even more useful in keeping in check the undue increase of these animals, especially the rattlesnake, as their nocturnal habits bring them in frequent contact with the ‘‘one-button,” the hopeful descendent of the mature Crotalus. There are some other allies still to be mentioned, who give us no inconsiderable assistance in our attempts to get rid of the rattlesnake. One of these is the king snake, Ophibolus getulus. An incident which occurred in a drug store in this city, where various living reptiles are kept in a glass case, proves that the statement of the prowess of this snake made by old rancheros are founded on actual obser- vation, and not as has been supposed, on vague tradition. A specimen of Ophibolus being one day added to the happy family, attacked a rattlesnake immediately on its entrance, and winding himself by an exceedingly rapid mo- tion round the latter strangled him, but not before receiy- ing a wound from the fangs of his foe from which he died the next day. The confined space, preventing freedom of movement, in which the conflict took place, was probably the cause of this mishap. Similar statements have been made in regard to Pityophis. As to Hutainia we may be certain that these defenseless species do not disturb the rattlesnake, and as their food is chiefly frogs and small fish they do not inter- fere with his food supply as Pityophis certainly does, by feeding on ground squirrels, gophers, rats, mice, etc., and Ophibolus still more, by climbing shrubs and devouring the eges and young of birds, a cheerful nocturnal custom of the rattlesnake also. In relation to the amount of help given by these enemies of the rattlesnake, they rank about as follows: First.—Owls, which feeding on the strictly nocturnal “ one-button,” destroy a greater number of individuals, needing more for food on account of their small size. CHANGES IN FAUNA AND FLORA. 99 Second.—Hawks and eagles, which destroy many of the mature snakes moving about in the day-time. The diminution in the number of these birds, at least as far as the Coast Range is concerned, is probably largely due to the forest fires, which are, I am sorry to say, very frequent in some parts. The owl builds usually in old hollow trees, which are of course more easily destroyed by fire than sound ones. The eagle builds every year, if undisturbed, in the same tree, and is very apt to leave a region where its nest has to be frequently changed. ‘Of course these fires do not directly injure the birds, for they occur late in the year, after the fledging of the young. Ophibolus and Pityophis are comparatively trifling factors in the destruction of the rattlesnake, nevertheless it would be wiser to leave inoffensive creatures unharmed, for they are his implacable enemies, and perform a very consider- able service to the farmer by lessening the number of ver- min, and itis after all a very cheap kind of heroism to wantonly kill an animal which is incapable of doing harm. Nature not interfered with by man is perfectly able to take care of herself. When once we do so all the ‘bisul- phides” and ‘‘ insect washes ” in the world will not restore the balance which has -been disturbed; natwram expellas Jurca tamen usque recurret,” says Horace. We can only make Nature our ally, never our servant. 100 CALIFORNIA ACADEMY OF SCIENCE'S. A NEW GENUS AND SPECIES OF N. A. SCARABAIDZA., BY J. J. RIVERS. The insect about to be noticed, would in general appear- ance remind any one of Pentodon, but an examination of the mouth parts would at once dispel the idea, the strongly toothed outer margin in Pentodon, together with the highly developed teeth of its maxillary galea, would seem to re- move those insects to different associates. The one I now mention has a history; it adorns several collections either without a name or else labeled Aphonus clunalis Lec. A reference made to LeConte’s description of A. clunalis, Proceed. Acad. Nat. Sci., Phila., Vol. 8, p. 23, 1856, and at the end of the description the following phrase ‘‘maaxillarum galea bidentata,” is sufficient to show that LeConte did not refer to the present insect, for its maxillary is entirely unarmed. There appears to be no description of this insect any- where in the literature of N. A. Coleoptera; and thinking that as this species was from El Paso, Texas, Mr. W. Bates, of England, who is working upon the Mexican Scarabeida, had perhaps received and already incorporated it in his mon- ograph, I wrote to him concerning it. He replied: ‘‘ Pro- ceed and describe it; send me a copy of the publication containing the description, and I will adopt your naming if I find the species among those occurring across the line.” This insect was in a large capture of Scarabeide taken last year at El Paso, Texas, by G. W. Dunn, after whom the specific name is formed; the generic name is derived from the unarmed condition of its galea ov maxillary. The group it belongs to is very puzzling, as it appears to possess diverse characteristics; the tarsal form and lack of stridulat- ing organs throw it with Aphonus; the style of sculpture of the elytrz is of the common oblique pattern, while the shape 2p Ser. VOL. I. Issued June 5, 1888. NEW GENUS AND SPECIES OF SCARABAIDA. 101 of the middle and hind tibis, together with the bulging upon the pygidium, point towards Oryctes, and the rounded form of the club of the antennz also favors Oryctes, though it matches Pentodon in that character very much better and it agrees with none in its unarmed galea. ANOPLOGNATHO gen. nov. Form oval, convex; dorsum gently rounded; under wings well developed. Clypeus triangular, with sides slightly emarginate, apex rounded and curving upwards; in fresh unworn examples the margin is continuous, and meets a depression on either end of the sutural line; this transverse line rises towards the center, but the center itself is slightly depressed. Antenne nine-jointed, the basal joint very large, equal to all the joints of the funicle; the fifth and sixth joints being connate, makes this truly a nine-jointed species. Labrum membranous, covered with stiff hairs. Mandibles much produced, not toothed externally, but with a slight emargination. Maxillw much flattened and widening from the middle to the front edge, the inner angle is sharply defined and with a projecting joint, the outer angle much rounded; the under surface has stiff bristly hairs pointing forwards. Maxillary palpi four-jointed; the first joint narrow, the second and third about equal, and the fourth nearly as large as the second and third combined. Labium bilobed, with well rounded margins. Mentum connate with the several adjoining parts, the labial palpi appearing at the sides, they are three-jointed, the first joint being shorter than the second and the third equal to both. Prothorax shining black, wider than long and narrower in front, front margin recurved, smooth, side margins slightly reticulated and reflexed, but becoming flattened before reaching the hind angles; hind margin well developed from the angles by an impressed line, which dying out before reaching the center the margin is continued by coarse puncturings only. Elytra elongate-oval wider towards the apex, evenly rounded on top, falling to the sides in a gentle curve, Anterior coxe large, with the trochanter oval in outline, the outer end fit- ting into an emargination of the epimera, Middle coxz nearly the size of the former, of irregular form, with the 102 CALIFORNIA ACADEMY OF SCIENCES. apex somewhat truncate. Posterior coxz small and subtriangular, oblique. Abdomen: ventral segments six; fifth and sixth movable. Spiracles situated upon the inner edge of the inflexed portion of the ventral segments except the posterior one, which is in the suture, where the propygidium meets the fifth ventral segment. The middle and hind tibiz have each two oblique carinw across the middle portion, and their apices are expanded and digitate. Middle tarsi moderately stout, first joint more robust, the second and third of equal length, and the fourth shorter, Hind tarsi shorter than those of the middle pair; more robust and more equal in length. Propygidium regularly rugose, what might be called reticulate. Pygidium very obtusely and evenly rounded in the male. Anoplognatho Dunnianus sp. nov. —Elongate-oval, convex, bruader behind. Clypeus triangular, with the sides slightly emarginate, tip rounded, upturned; the edge carries a raised margin that meets the sutural line, which increases in hight towards the middle, but is eut by a slight depression in the center, giving the outline of a double curve. This character, however, is not constant. There is no other armature on the head, Vertex rugosely punctate, as is the depression forming the area of the clypeus. Prothorax transverse, narrower in front, angles pointed, sides reflexed, partly reticulated but flattened before reaching the hinder angles; hind margin well defined part of the way across from the angles, but con- tinued over the middle by punctures only. . The middle of the disc towards the hind margin is free from punctures, but towards the side margins they become closer and merge into each other, while at the front angles they form a densely rugose surface. Elytre elongate-oval, wider behind the middle, and covering the propygidium; the sutural margins near the apex appear depressed, but it is caused by the bulging on either elytron by the usual broad tubercle; there are three series of obliterated striz, the two inner sets being double, and occu- pying an oblique course from base to apex. Pygidium very obtusely rounded. Front tibiz strongly sinuate. Be- low, the various parts are clothed with hairs placed in lines; the color is deep chestnut. Above, the surface, particularly of theelytra, becomes almostsmooth and shining; the color is brownish black, inclining to a chestnut tone towards the hinder part of the elytre. Length .95-1 inch. Female generally smaller, with the sides of the elytre more parallel; it ismore convex than the othersex, the ely tre shorter not covering the propygidum, and smoother. Pygidium obtuse but not evenly bulging, because the Ci lower part is incurved and margined by a boldedge. Front tibia strongly tridentate. Length .75-.95 inch. Locality: El Paso, Texas. CALIFORNIAN LEPIDOPTERA. 103 A NEW SPECIES OF CALIFORNIAN LEPIDOPTERA. BY J. J. RIVERS, This insect is from the neighborhood of Truckee, a mount- ainous district of California. It has been mistaken for a variety of JM. anicia Doubl., but as this has a black eround color while JZ. anicia has a red one, they must belong to different groups of the same genus. Mr. C. F. McGlashan has for several years bred this insect in numbers and dis- tributed it generously, therefore I have great pleasure in dedicating the species to him. M. Chalcedona Doubl. is the type around which numerous forms congregate, some as mere varieties, while others, though showing strongly their ancestry still exhibit suf- ficient likeness to each other in the imago state, to puzzle and perplex the student, and make it difficult to say whether these various forms are simply variations or true species. In the larval state, however, abundant difference is seen which easily points out the genuine species. Dr. H. Behr in his description and remarks upon MW. Cooperi, Proceed. Cal. Acad. Sci., Vol. 3, p. 90, 1863, says: ‘‘ With all the well ‘marked differences, the species look so much alike, that it would have been a long time before M. Cooperi would have been recognized as a distinct species, if it had not been for ' the striking difference of its caterpillar.”’ MELITZAA Fabr. M. Macglashanii sp.nov. General appearance like M. Chalcedona, but still more like M. Cooperi Behr. Outer margins of upper wings rounding in both sexes. The general disposition of the spots forming the bands being equidistant, gives to the whole surface a more checkered appearance than is seen either in M. Chalcedon or M. Cooperi. Expanse of wing of the largest @ specimen 24 inches; in the largest Q specimen of M. Chalcedon 24 inches. Vhe murginal band of red spots to all the wings always present. Male.—Upper side blackish; outer margins of all the wings with reddish spots, a character more constant in this species than is found in either Cooperi or Chalcedon. The next two rows are pale straw or even white, and the spots are inclined to be quadrate; the usual bifid band has the third 2p Sen. Vou, I. Issued June 6, 1888, 104 CALIFORNIA ACADEMY OF SCIENCES. and fourth spots from the costa wholly red; three spots in cell. two yellow and one red; hind wings have first a row of marginal red spots, followed by three rows of yellow, then, in some, an elon- gated twin spot of red; the basal area has four angulate spots of yellow, three forming a triangle. Under side of anterior wings asubdued brownish- red, the marginal row scarcely defined; in some examples, however, there is a black zigzag line that cuts off the marginal row; the second and third rows are repeated faintly, and the markings of the upper surface are faintly vis- ible; hind wings have an unbroken marginal border of red which continues firmly along a third of the inner margin. The second band consists of seven conical spots of yellowish white; the third row is made up of rounded or angulate spots of red, attended more or less by a yellow halo, and in some specimens the spot is circular, and a dot or pupil adorns the center; the fourth row contains eight large elongate spots of whitish-yellow, cut unequally by a black line, but the first spot on the front margin is almost equally divided; the basal area is red and contains three yellowish spots, the largest of which is sometimes divided by a reddish spot. Female exhibits the same markings as the male; with the exception of the marginal borders and a faint tinge of red on one horn of the bifid band, the whole surface is of sombre black. Palpi reddish; antennz yellowish; abdomen with five dorsal segments fringed with rusty brown, and a white row of spots on each side of the dorsal ridge; beneath, through the center, it is whitish. Expanse: male 13; female 21 inches. Larva.—Deep black, sparingly irrorated with white. Seven of the dorsal spines light yellow, with a fine line of deeper yellow forming a ring around the base of each. A double line or band of white occupies the dorsal region, interrupted by the base of the spines, and longitudinally separated by a fine black line that runs along the dorsal ridge; the sub-dorsal row of spines black on a black ground, the next row of spines also black, but at the base below each of seven spines is a yellow crescent-shaped spot which interrupts a broad spiracular band of white; the next row of spines black, then a narrow longitudinal line of scattering white spots; then follows, above the feet, a row of twin diminutive yellowish spines, which interrupts a line of white. Beneath is a double line of white. The hairs below the spiracular line brownish. ‘The head, prolegs, base of feet and spines (except the seven dorsal) black and shining. Caterpillar looks very like that belong- ing to WM. Leanira, and therefore more attenuated than that of the robust M. Chalcedon. Food plant Pentstemon, but will feed upon the flowers of Castilleia, and upon the leaves of Plantago major. Chrysalis.—Dull white, prothorax with two yellow tubercles, the first dorsal segment has also two, the second three, the fourth to the tenth have five tubercles on each dorsal segment. In the front of each tubercle is a black curved spot often uniting. There are also some black spots on each CALIFORNIAN LEPIDOPTERA. 105 segment placed in regular order, but there are no black dots on the apex of the tubercles as seen in the allied species. Length %{ inch. CLISIOCAMPA. C. thoracica Stretch. Larva.—Body obscure brown, dorsal ridge orna- mented with a row of conspicuous ochre-colored spots, two spots on each segment, the one placed in front always much the larger; the spots are attended by numerous short waving lines of reddish-brown, running longi- tudinally, and there is a sub-dorsal line of the same color, and a spiracular line of a much lighter shade. Its whole length above the feet is occupied with numerous tufts of grayish hair. The dorsal line of spots is its most particular feature, as it contrasts with the sombre tone of the body color. Length 1%4 to 2 inches. Natural food plant the native willow, upon which it is found feeding at Berkeley, but it has taken to the foliage of different kinds of orchard trees, upon which great damage has been done. 106 CALIFORNIA ACADEMY OF SCIENCES. CONTRIBUTIONS FROM THE SHAW SCHOOL OF BOTANY. No. 4. SYNOPTICAL LIST OF NORTH AMERICAN SPECIES OF CEANOTHUS. By WILLIAM TRELEASE. In its present limitation, Ceanothus is an exclusively Amer- ican genus, and all but three of the species known to me occur north of Mexico. It is quite puzzling to the student of our flora, partly because of the variability of some spe- cies, and partly because in some instances what are evi- dently distinct species approach each other so closely as to render their characterization difficult, although they may be pretty surely recognized by the practiced eye. ‘To the evo- lutionist, these groups of interlocking sub-species and barely distinguishable species, are interesting as represent- ing different stages in the mutations of their prototypes; but they are exceedingly trying to the systematic botanist, whose task is to so arrange and limit them as to render their recognition by others easy. Aside from the monograph of the genus (in the broader sense in which it was formerly understood) in De Candolle’s Prodromus, which characterizes very few of the species that are now known, and the fuller revision in the first volume of Torrey and Gray's Flora, it has been very judiciously and completely treated by Mr. Watson in a revision of the en- tire group (Proc. Amer. Acad. x. 1875, 332-9), and a separate elaboration of the numerous and difficult Californian species (Bot. Cal. i. 102-104; ii. 439). His views on the syno- nymy of species are also very fully indicated in his Biblio- graphical Index to North American Botany. Nor should the field notes of the keen-sighted Nuttall, on a number of the Pacific Coast species, in Torrey and Gray’s Flora, be overlooked. 2p SER. VOL. 1. Issued June 15, 1888. NORTH AMERICAN SPECIES OF CEANOTHUS. 107 In the main, the limitation and descriptions of species in Mr. Watson’s monograph, and his references to literature, in the Index, are so accurate as to require little alteration; but after repeatedly going over the genus as represented in the Gray, Engelmann, and Torrey herbaria, and in the ex- tensive collections of the California Academy of Science, and the United States Department of Agriculture, as well as the fine private herbarium of Mr. J. C. Martindale, it has seemed to me that Ceanothus ig capable of subdivision into more natural groups than have yet been proposed, and the following is otfered as a tentative arrangement of our spe- cies, I shall be very grateful for specimens and notes eluci- dating doubtful species, and I desire here to express my obligation to the botanists who have already placed public or private collections in my hands for study. A. Leaves alternate, not spinescent, glandular-toothed or occasionally entire; stipules thin, often subulate, fugacious; fruit sometimes keeled or crested on the back of each segment, but not bearing prominent dorsal horns.— Huceanothus. a. Inflorescence ‘on leafless lateral peduncles borne on the old wood. 1. C. SANGUINEUs, Pursh. Fl. Am. Sept. i. 167; Wat- son, Proc. l. c. 334, Index, 166.—British Columbia to Idaho and California. b. Inflorescence on leafy shoots of the present season’s growth. * Flowers white, in small simple corymbose clusters terminating mostly leafy spineless twigs: leaves very small (2 to 8 mm. long), 3-nerved.—Atlantic species. i 2. CC. MICROPHYLLUS, Michx. FI]. 1. 154; Watson, J. c. 335. Georgia to Florida. 108 CALIFORNIA ACADEMY OF SCIENCES. 3. C. SERPYLLIFOLIUS, Nutt. Gen. i. 154; Watson, J. c. 335.—Georgia. The locality in Nuttall’s Genera, and on an original specimen in the Gray herbarium, is Florida. * * Flowers white: inflorescence rather simple and mostly compact, at the ends of slender usually leafless or nearly leafless peduncles: twigs subte- rete, not spinose: leaves ample or medium-sized (15 to 75 mm. long), thin, toothed, 3-nerved: fruit about 4 mm. in diameter. — Atlantic or Rocky Mountain species. 4. (C. Americanus, L. Spec. 195; Watson, J. c. 333, In- - dex, 163.—Ontario to Manitoba, south to Florida and Tex- as. 5. C.ovatus, Desf. Arb. ii. 881; Watson, /. c. 384.—Can- ada and the lake region to Texas. VAR. PUBESCENS, Watson, Index, 166.— Rocky Mount- ain region. * * * Flowers blue: inflorescence compound, ample, on leafy branches: twigs conspicuously sulcate: leaves medium-sized (25 to 50 mm. long), prominently 3- ribbed, minutely glandular-serrulate: fruit mostly 5 or 5 mm. in diameter. 6. C. THYRSIFLORUS, Esch. Mem. St. Petersb. Nema ee 285; Watson, /. c. 334.—Mountains of Western Califor- nia. * * * * Flowers blue (often. pale) or white: inflorescence compound, rather loose. on few-leaved branches: leaves entire, mostly medium-sized.—Pacific spe- cles. + Twigs (more or less spinose in the first) usually somewhat angled and often twisted: leaves rather firm or even coriaceous. NORTH AMERICAN SPECIES OF CEANOTHUS. 109 «+ Leaves rather broad, not at all 3-nerved: fruit 5 to 6 mm. in diameter. 7. C. sprnosus, Nutt. in Torrey & Gray, Fl. 1. 267; Wat- son, /. c. 337.—Middle and Southern California. 8. OC. Paumeri, n. sp. Glabrous throughout, or a very few hairs on the leaves and petioles: branches greenish, becoming brown: leaves mostly on short spurs, slender-peti- oled, about 40 mm. long, elliptical or ovate-oblong, rounded at both ends, mucronate or emarginate, entire, thinner than in the last: flowering branches ascending, naked or few-leaved: inflorescence oblong, nearly simple: exocarp of fruit rather fleshy.— Mountains of Southern California (Palmer, 1875, No. 42). Intermediate between C. spinosus and C. integerrimus. ++ ++ Leaves narrow, 3-nerved, the nerves often conceal- ed by the revolute margins; fruit about 3 mm. in diameter. 9. C. Parryi, n. sp. Branches glabrate or sparingly villous, strongly sulcate, more or less papillate : leaves nar- rowly elliptical-oblong, obtuse, 15350 mm. or less, gland- ular-serrulate, glabrous above, the lower surface rusty-to- mentose, at least along the veins: inflorescence oblong, interrupted, terminating recurved - ascending slender, few- leaved branches: flowers blue.—Known to me only from specimens found in cultivation at Calistoga, Cal. (Parry, 1881, No. 33). +— + Twigs not spinose nor rigid, very slightly angled while young: leaves thinner, at least the largest 3- nerved: inflorescence rather dense for the group; fruit about 4 mm. in diameter. 10. C. INTEGERRIMUS, Hook. & Arn. Bot. Beechey, 329; Watson, J. c. 334, Index, 165. C. thyrsiflorus, var. macro- thyrsus, Torrey, Bot. Wilkes, 263! California to Arizona. 110 CALIFORNIA ACADEMY OF SCIENCES. 11. C. parvirouius. C. integerrimus, var.? parvifolius, Wat- son, J. c. 334.—California to Oregon. *%%**%* Flowers blue or white: inflorescence mostly com- pound and ample: leaves generally medium-sized to large 3-nerved; margin various. + Twigs not spinose nor conspicuously glaucous, the youngest angled in the first species: flowering- branches few-leaved or leafless: leaves broad, usu- ally large (25 to 75 mm. long): fruit about 5 mm. in diameter. 12. C. arBorEuS, Greene, Bull. Cal. Acad. ii. 144. @C. sorediatus, Lyon. Bot. Gaz. x1. 204, 333.— Islands off the Californian Coast. 14. C. veLuTinus, Dougl. in Hook, Fl. Bot.—Am. i. 125, pl. 45; Watson, J. c. 334.—British America to California, Colorado and Nebraska, chiefly in the mountains. VAR. L&vIGATUS, Torr. & Gray, Fl. i. 686; Watson, Index, 167.—Range of the species. + + Twigs terete, often very divergent and rigid, some of them ending in firm spines. ++ Twigs short, often very glaucous: leaves large in the first, mostly medium-sized, rather small in the last: fruit 4 to 6 mm. in diameter. 14. C. incanus, Torr. & Gray, Fl. i. 265; Watson, 1. c. 336.—California. 15. C. EGLANDULOSUS. C. divaricatus, var. eglandulosus, Torrey, Pac. R. R. Rep. iv. 75. C. divaricatus, Watson, l.c. in part. — Mountains of California and Lower Cali- fornia. NORTH AMERICAN SPECIES OF CEANOTHUS. eel 16. C. pivaricatus, Nutt. in Torr. & Gray FI. i. 266, 686; Watson, J. c. 336 in part.—-California and Lower Cali- fornia. VAR. GROSSE - SERRATUS, Torrey, Pac. R. R. Rep. iv. 75. —California. 17. C. corDuLatTus, Kellogg, Proc. Cal. Acad. ii. 124, f. 39; Watson, /. c. 337.—Mountains of California. ++ ++ Twigs slender, usually slightly if at all glaucous: spines slender and sharp: leaves rather narrow for the group: fruit about 4 mm. in diameter. 18. C. FEenDLeErI, Gray, Pl. Fendler 29; Watson, J. c. 337. Mountains of Colorado, New Mexico and Arizona. VAR. VIRIDIS Gray, in Herb. — Arizona (Lemmon. Greene). * £ ¢ Flowers deep blue (except sometimes in the first), in rather compact nearly simple corymbose or ob- long clusters, on leafless or nearly leafless (some- times abbreviated), usually scaly peduncles: twigs terete, mostly spineless, and not very rigid.—Pa- cific species. * Leaves medium-sized, all or nearly all 3-nerved: twigs rather more rigid in forms of the first two than elsewhere in this group: fruitabout 4 mm. in diameter. 19. C. sorepiatus, Hook & Arn. Bot. Beechey, 398. Watson, J. c. 336. — Coast Range of Southern California, extending into Lower California. 20. C. nirsuTus, Nutt. in Torr. & Gray, Fl. 1. 266; Wat- son, J. c. 336. C. diversifolius Kellogg, Proc. Cal. Acad. 1. 58, 65.—Mountains of Southwestern California. VAR. ? GLABER, Watson, J. c. 336. — East Humboldt mountains, Arizona. (Watson, No. 212.) tA CALIFORNIA ACADEMY OF SCIENCES. 21. C. DECUMBENS, Watson, /. ¢. 335.—Mountains of Cen- tral California. ++ Leaves usually small, commonly only the largest 3-nerved, not papillate: fruit 3 mm. in diameter. 22. C. DENTaTUS, Torr. & Gray, FI. i. 268, Watson lI. c. 335.—Coast range of Southern California. SUBSP. ? FLORIBUNDUS. C. floribundus Hook. Bot. Mag. pl. 4806; Watson, J. c. 338.—Known certainly only from plants grown in European gardens, from Cali- fornian seed. suBsP. ? Loppranus. C. Lobbianus, Hook. Bot. Mag. pl. 4810.—Coast Range of Southern California. 23. C. IMPRESSUS, n. sp. Villous, with short spreading hairs: leaves broadly elliptical to nearly orbicular, 6 to 8 mm. long, loosely villous, especially on the veins below, the upper surface deeply furrowed over the midrib and several pairs of lateral nerves, the slightly glandular mar- gin very revolute, appearing there as if crenate: pedun- cles about 10 mm. long, scaly toward the base: inflorescence sub-globose, compact: fruit not seen.—Santa Barbara County, Cal. +++ Leaves medium-sized, oblong, not at all 3-nerved, papillate. 24. C. papituosus, Torr. & Gray, FI. i. 268; Watson, J. c. 337.—Mountains of Western California. 25. C. Vearcutanus, Hook. Bot. Mag. pl. 5127; Watson, 1. c. 338. —Decribed and known only from European plants raised from Californian seed. B. Leaves opposite or alternate, pinnately veined, coriaceous, of- ten pungently toothed; stipules thick and spongy, taper- pointed persistent; inflorescence densely corymbose, on short spurs from the new wood; fruit usually large for the genus, each carpel commonly bearing a dorsal horn, an alternating set. of 3crests or horns frequently at or near the apex.—Cerastes Species of the Pacific Coast or Southwest, mostly with rigid almost spinose twigs. NORTH AMERICAN SPECIES OF CEANOTHUS. tis a. Procumbent and radicant: flowers bright blue (ex- cept occasional albinos or hybrids), mostly on long, slender, colored pedicels. 26. C. prostratus, Benth. Pl. Hartweg, 302; Watson, 1. c. 8339.—Washington Ter. to California and Nevada. b. Erect or spreading, the branches mostly rigid: ped- icels usually stouter or rather short. * Leaves opposite. + Flowers (always?) white. 27. ©. cunEatus, Nutt. in Torr. & Gray, Fl. 1. 267; Wat- son, J. c. 338 and Index, 164, in part. —Oregon to Lower California. 28. C. Greeai, Gray, Plant. Wright. ii. 28; Watson, l.c. 338.—Utah, Arizona and New Mexico to Mexico. 29. C. crassiFoLius, Torr. Pac. R. R. Rep. iv. 75; Wat- son, /. c. 338 and Index, 164, excl. synonym and variety. —Coast range of Southern California, Lower California and Californian islands. ++ Flowers blue (or rose-purple ? ). 30. C, Ricipus, Nutt.in Torr. & Gray, Fl. i. 268; Watson, l. c. 339.— Coast region of Southern California. VAR. GRANDIFOLIUS, Torr. Pac. R. R. Rep., iv. 75; Wat- son, Index, 164, under C. crassifolius. C. crassifolius, var. glabratus Gray, Cat. Bolander, 8; Watson, Index, 164.—Range of the species. ** Leaves alternate; flowers white. 2p SER. VOL. I. Issued June 15, 1888. 8 114 CALIFORNIA ACADEMY OF SCIENCES. 31. C. veRRucosuS, Nutt. in Torr. & Gray, Fl. i. 267; Greene, Bull. Cal. Acad. 11. 81. C. cuneatus, Watson, l.c. 338, and Index, 164, in part.—Southern California and Lower California. 32. C. MacrocaRPus, Nutt. in Torr. & Gray FI. i. 267; Greene, /. c. C. cuneatus, Watson, J. c. 338, and Index, 164, in part.—Coast Range of Middle and Southern California. NOTES. C. sanguimeus appears to differ from all other species of the genus in flowering from wood of the preceding season’s growth; otherwise, it is related in several respects to Nos. 4 and 5. C. microphyllus and serpyllifolius—very closely related to each other—show no very great affinity for other groups. For convenience they are placed where they now stand, al- though in the character of the inflorescence, they approach the dentatus group. I have no fruit of either. While C. ovatus is well marked in its typical Eastern form, it passes gradually into var. pubescens in the West, and through this approaches Americanus in its leaf-forms. The leafiness of flowering branches is sometimes quite variable. In C. spinosus, the firm leaves commonly turn brown in drying, especially the upper surface; the branches of the ample somewhat leafy loose thyrsus mostly spread at right angles or are even recurved; and the flowers are scarcely more than lilac-colored. C. parvifolius appears to be distinct from integerrimus in its loose low habit, smaller leaves scarcely exceeding 25 mm. in length, the majority of them not 3-nerved, and in its smaller oblong or (from the falling of the lower fasci- NORTH AMERICAN SPECIES OF CEANOTHUS. 115 cles) subcorymbose inflorescence of blue flowers—those of integerrimus appearing to be always white. C. arboreus is, as Professor Greene has shown, the largest of our species. It was collected on Sta. Catalina many years ago, by Nuttall, who notes its arboreous habit on the label accompanying a specimen in the herbarium of the Philadelphia Academy. Numbers 14 to 17 of the list are quite difficult to dis- tinguish. C. incanus, with large, rather thin, nearly smooth ellip- tical leaves, often rounded in abruptly to the lateral nerves, a short distance above ths base, usually entire, pale below; rather large white flowers (often 5 or 6 mm. in expanse); and large depressed fruit with a more fleshy exocarp than usual, is generally recognized without much difficulty. 'The same is true of C. eglandulosus, which is nearly glabrous and very glaucous, with rather small strongly ovate entire or © nearly entire leaves, usually brown above, in herbarium specimens; and smaller dingy blue flowers. But divaricatus and cordulatus approach each other so closely that it is hard to draw the line between them. The specimens with smaller, thicker leaves are commonly referred to the latter, as being evidently what Kellogg figured, while the former species usually has large and more flexible leaves. C. Fendleri, with rather thin narrower leaves, silky-can- escent in the type, green and nearly glabrous in the variety; stands out quite well from its congeners, in geographical distribution, also. C sorediatus is a species which I do not at all understand. As it is here accepted, it includes plants with slender rather simple twigs, and others that are quite rigid and intricately branched; the leaves of some are very broadly ovate, while others are narrower; and the pubescence varies from silky 116 CALIFORNIA ACADEMY OF SCIENCES. ormore or less hirsute to very densely white-or rusty- tomentose on the lower surface of the leaves. The inflor- escence is intermediate between that of the preceding and following groups of species. In some of the rather slender- stemmed plants that I regard as most typical, the twigs are very rough with crowded small warts. Probably one or more species may ultimately be separated out, and it is not certain that a number of specimens now referred here do not really belong to divaricatus on the one hand, or hirsutus on the other. The dentatus group is one of the best circumscribed, but it is not less puzzling than others, for it is doubtful whether the forms it comprises are best described as species or va- rieties. Lobbianus is chiefly characterized by its conspicu- ously unequal leaves—those on the primary shoots 25 mm. long, mostly acute at both ends, the larger 3-nerved. While I have scarcely felt that this form and floribundus are worthy of specific rank, I have pretty confidently sep- arated out the plant with rather broad hairy leaves, deeply furrowed over the veins. It is known to me only from two flowering specimens in the Gray herbarium, collected at different places. In Crrastes the species are quite as perplexing as in EKucranoruus, and the difficulty of separating them is in- creased by the occurrence of what appear to be hybrids. This is especially true of prostratus and cwneatus, typically very distinct in habit, foliage, flowers and fruit; but nu- merous specimens have been collected over a large area, showing various combinations of the characters of both. So marked are these that Mr. Thomas Howell writes me that from his field observations he is disposed to regard prostra- tus as only a variety of cwneatus.* “On these supposed hybrids, see Garden and Forest, i. 7. NORTH AMERICAN SPECIES OF CEANOTHUS. UT The fruit of prostratus is usually considerably longer than broad, about 8 mm. in diameter, rather fleshy, with 3 nearly erect prominent horns, and an accessory set of apical crests or smaller horns. Cuneatus produces subglobose or very slightly elongated fruit, about 6 mm. in diameter, with thin flesh, and usually 3 rather small horns, an accessory set of smaller ones being occasionally seen. C. Greggii, closely related to cwneatus, is marked by its leaves, often white-tomentose below, mostly equally rounded and acute at both ends, frequently with one or two small teeth on either side, and its rather pointed fruit, about 4 mm. in diameter, seemingly nearly or quite horn- less, the white calyx-lobes more persistent than usual. C. crassifolius, as it is now accepted, is very heterogen- eous, embracing plants with entire or toothed smooth green leaves, and others with the leaves pungently toothed, revo- lute margined, and very white-woolly below. The fruit is as much as 8 or 10 mm. in diameter, depressed-globose, smooth or with 3 low deeply dorsal horns, and the base (in- vested by the adnate calyx) very prominent, and indurated. The leaves are always firm and thick. C. rigidus has thinner, often cuneate-obcordate, mostly denticulate leaves, and blue or purplish flowers. Good fruit is desirable. Some specimens suggest hybridization with C. prostratus. C. verrucosus is very similar to rigidus in its foliage char- acters, but with slender twigs, alternate leaves and white flowers. The figure of rigidus in Bot. Mex. Bound. pl. 9, is evidently this species. Its fruit is mostly 4 to 6 mm. in diameter, with small dorsal horns, or none. he ay CALIFORNIA ACADEMY OF SCIENCES. C. macrocarpus approaches broad-leaved forms of cuneatus, and especially entire-leaved forms of crassifolius, in its foli- age, but its habit is said to be arborescent, its twigs are slender, and its leaves opposite. Its fruit is very similar to that of the entire-leaved crassifolius, but with prominent dorsal horns. SOUTH AMERICAN NEMATOGNATHI, 119 PRELIMINARY NOTES ON SOUTH AMERICAN NEMATOGNATHI. I. BY CARL H. EIGENMANN AND ROSA SMITH EIGENMANN, The present paper contains a list of the species of Hypoph- thalmidz, South American Siluride, Argiide and Callichthy- ide, and of the genus Plecostomus (Loricariidz), with the localities of the specimens contained in the Museum of - Comparative Zodlogy at Cambridge, Massachusetts, and descriptions of the following new forms: 1, Pimelodina nasus; 2, Luciopimelodus; 3, Pseudopimel- odus acanthochirus; 4, Rhamdia obesa; 5, h. Poeyi; 6, Rham- della (sub-genus); 7, Rhamdia tenella; 8, Rh. eriarcha; 9, R. parryt (Mexico); 10, Pimelodella; 11, P. pectinifer; 12, Sciadeoides (sub-genus) ; 13, Duopalatinus; 14, Steindachneria; 15, S. amblyurus; 16, Pseudoplatystoma fasciatum inter- medium; 17, Tachisurus lentiginosus; 18, T. jordani; 19, T. longicephalus; 20, T. agassizi; 21, T. gulosus; 22, Ageneiosus atronasus; 23, Auchenipterus fordicei; 24, Felichthys flaves- cens; 25, Trachycorystes porosus; 26, T. analis; 27, Hassar; 28, Doras wranoscopus; 29, D. spinosissimus; 30, Decapogon; 31, Corydoras hastatus; 32, Plecostomus limosus; 33, P. com- mersonu scabriceps; 34, P. seminudus; 35, P. macrops; 36, P. vermicularis. ‘The specimens in the Museum were collected by Professor Louis Agassiz and his assistants during the Thayer Expedi- tion in Brazil, and by Professor ‘Agassiz and Dr. Steindach- ner during the Hassler Expedition around South America. The names of collectors of new forms are given with the descriptions. A fuller account of all the species mentioned will be published later. 2p SER. Vou. I. (9) Issued July 18, 1888. 120 CALIFORNIA ACADEMY OF SCIENCES. HYPOPHTHALMIDA. I. HeELoGenes Giinther. 1. Helogenes marmoratus Ginther. II. HyporHTHaLmus Spix. Notophthalmus Hyrtle; Pseudohypophthalmus Bleeker. 2. Hypophthalmus edentatus Spix. Hypophthalmus marginatus, H. longifilis, and H. spixii Cuyv. & Val. Hypophthalmus edentulus Castelnau; Hypophthalmus jimbriatus Kner; Hypophthalmus perporosus Cope. Locality: Para. SILURID Zi. IT. Cattopuysus Miller & Troschel. Pimelotropis Gill; Pseudocallophysus Bleeker. 1, Callophysus macropterus (Lichtenstein). Pimelodus ctenodus Agassiz; Pimelodus insignis Schomburgk, in part: plate, but not description; Pimelotropis lateralis Gill. Localities: Obidos; Lake Jose Assu; Cameta; Rio Negro; Santarem; Tonantins; Lake Manacapuru. II. Prmetopina Steindachner. 2. Pimelodina flavipinnis Steindachner. 3. Pimelodina nasus sp. nov. Type, No. 7490. Length to base of caudal .34m. Para. Agassiz. This species differs from flavipinnis as follows: adipose fin 2 in the length (23 in flavipinnis); eye 8 in head (6); head not compressed behind the eyes (compressed); great- est width of head 13 in its length (8); maxillary barbels reach to the end of the adipose fin (beyond base of caudal); head everywhere strongly convex in a cross-section (almost flat between the eyes); depth above ventrals 44 (53-53); a dark humeral area, otherwise plain yellow (several rows of brown spots on the upper half of the body). Elongate fusiform, the back compressed, trenchant; SOUTH AMERICAN NEMATOGNATHI. BER width, above ventrals, 1? in the greatest depth. Head short, flat below, regularly convex above; profile very steep and almost straight. The thick skin obliterates the outlines of all the bones. Occipital process long and narrow, not reaching the dorsal plate. Hye small, its anterior margin equidistant from the snout and the free border of the opercle, its diameter 4 in the snout, 8 in the head, 24 in the interocular. Posterior nasal opening a transverse slit. Maxillary barbels reaching to or beyond the end of the adipose fin; mental barbels reaching beyond base of the pectorals, the post-mentals beyond their tips. Mouth entirely inferior, the snout projecting beyond it more than one diameter of the eye; lower jaw weak, the teeth scarcely evident, in a very narrow band; the teeth of upper jaw in a somewhat wider band. Gill-rakers 7-17. Distance of dorsal fin from tip of snout 33 in the length; first dorsal ray high and slender, scarcely spinous; the free margin of the fin straight. Adipose fin beginning imme- diately behind the dorsal, its base 2 in the length. Caudal fin deeply forked, the lobes of equal length, longer than the head. Fourth anal ray highest, higher than the fin’s base, 2 in the length of the head. Ventrals inserted little nearer anal fin than to tip of snout, 1} in the head. First pectoral ray scarcely spinous, slightly serrate on its inner margin, about as long as the head. Axil and humeral region blackish, otherwise uniform yellowish. Head 54; depth 53; depth of peduncle 13; Br. 9; D. 7; A. 12. III. Pirinampus Bleeker. 4, Pirinampus pirinampu (Spix). ? Pimelodus barbancho Humboldt; Pirinampus typus Bleeker. Locality: Cameta. 122 CALIFORNIA ACADEMY OF SCIENCES. IV. LuUcloPpIMELODUS gen. noy. Type, Pimelodus pati Valenciennes. First dorsal and pectoral rays not spinous, flexible and longer than any of the succeeding rays. Free margins of the dorsal and anal emarginate. Barbels scarcely mar- gined. Head depressed; snout elongate spatulate. anel continued to the base of the occipital process. cipital process a mere ridge. 5. Luciopimelodus pati (Valenciennes). Locality: Buenos Ayres. 6. Luciopimelodus platanus (Giother). VY. PsEUDOPIMELODUS Bleeker. Zungaro Bleeker; Batrachoglanis Gill; Lophiosilurus Stein- dachner. g Lophiosilurus. 7. Pseudopimelodus alexandri Steindachner. Pseudopimelodus agassizii. Steindachner. 4 Batrachoglans. 8. Pseudopimelodus parahybe Steindachner, Pseudopimelodus charus Steindachner (not C. & V.) Locality: Santa Cruz. 9. Pseudopimelodus raninus (Cuv. & Val.) 10. Pseudopimelodus pulcher Boulenger. § Pseudopimelodus. 11, Pseudopimelodus zungaro (Humboldt.) Pimelodus bufonius, charus and mangurus Cuv. & Val.; Zungaro hum- boldtii Bleeker. Locality: Goyaz. 12, Pseudopimelodus acanthochirus sp. nov. Type, No. 8133, one specimen, .063 m. Gurupa. Type, No. 8157, one specimen, .045m. Tajapuru. Type, No. 7732, two specimens, .093—.114 m. Teffé. Collected by Professor Louis Agassiz. Type, No. 7332, one specimen, .12m. Jutahy. SOUTH AMERICAN NEMATOGNATHI. 123 Depth, behind the humeral process, less than the width, tapering to the caudal peduncle. Head depressed, as wide as long. Frontal fontanel not continued behind the eye; an elongate occipital fontanel. Occipital process short, but meeting the longer dorsal plate. Large round pores on the head, most numerous on the snout and chin. Hye small, 24 in the snout, 8} in the head, 3 in the in- terorbital. Barbels all short, those of the maxillaries extending about to middle of pectorals. Mouth more than half as wide as length of head; intermaxillary band of teeth with a sharp, backward projecting angle; mandibulary band of teeth shallower than the intermaxillary band, about as deep as one diameter of the eye, scarcely widened in front. Humeral process nearly horizontal, strong, triangular, not reaching middle of pectoral fin. Distance of dorsal fin from tip of snout 23 in the length; the dorsal spine stout, toothed on its outer margin, 2 in head, the rays higher, nearly uniform. Distance of adipose fin from the dorsal more than the length of the dorsal fin; the adipose about as long as the dorsal fin. Caudal fin long, scarcely shorter than the head, rounded. Anal fin beginning and ending a little in advance of the anterior and posterior ends of the adipose; the tips of the anal rays reaching base of caudal fin when laid back. Ventrals inserted under the vertical from last dorsal ray, 14-1} in the head. Pectoral fin extending to the ventrals in young, much shorter in adult specimens, the spine 13 in head, little shorter than the longest ray, very much depressed, the teeth equally strong and long on both margins, those on the outer margin directed toward the tip, on the inner margin toward the base of the spine. Abdominal region sometimes covered with small papille; anterior portion of the lateral line with larger papille or 124 CALIFORNIA ACADEMY OF SCIENCES. short tentacles. Lateral line sometimes not extending to the caudal fin. Chocolate color; head faintly marked with lighter; a zig- zag light bar from gill-openings across the nape; dorsal fin dark brown, a short, transparent bar on the lower half of the last four dorsal rays; base and posterior third of caudal fin dark brown, the intermediate region transparent with dark mottlings; anal dark brown, margined with white, and a white bar on lower half of the last five rays; ventral and pectoral fins dark brown. Head 34; depth 5; Br. 8-9; D. I, 6; A. 10. VI. Ruamopra Bleeker. Pteronotus Swainson; Pimelonotus Gill; ? Notoglanis Giinther. DOUBTFUL SPECIES. I. Pimelodus velifer Humboldt; II. Pimelodus argen- tinus Humboldt; TI. Rhamdia laukidi Schomburgk; IY. Pimelodus grunniens Humboldt. § Nov. ? Occipital process and fontanel none. 13, Rhamdia breviceps (Kuer.) ¢ Rhamdia. 14, Rhamdia schomburgkii Bleeker. Pimelodus maculatus Schomburgk. 15. Rhamdia bathyurus Cope. 16. Rhamdia obesa sp. nov. Type, No. 7518. Length .26m. Teffé. L. Agassiz. Occipital process present; fontanel not continued be- hind the eye; jaws equal; intermaxillary band of teeth of about equal depth throughout; no dark lateral band; pores on the head not aggregated; dorsal I, 6; maxillary barbels extending beyond middle of the adipose fin in the adult. Vent midway between the bases of the caudal and pectoral fins. SOUTH AMERICAN NEMATOGNATHI. 125 Body short and deep, its greatest width less than its greatest depth. Head short, its width 1} in its length, - width, at the angle of the mouth, 2 in its length; the head flat above, covered with very thin skin, the surface of the bones deeply furrowed, striations radiating from the eyes and from the pit at the base of the occipital process. Font- anel a narrow fissure between the frontal bones. Occi- pital process very long, partly concealed under the skin, reaching more than half way to the dorsal spine to near the large, concealed, dorsal plate. Pores scattered over the head. Kye round, 24 in snout, 6} in head, 2 in the interorbital; its posterior margin in front of the middle of the head. Maxillary barbel (torn on one side) extending to tips of middle caudal rays; mental barbels reaching middle of pec- toral, post-mental to midway between tip of pectoral and base of ventral. Jaws equal, mouth comparatively small; teeth of the mandible as usual; the intermaxillary band of teeth shallowed and interrupted in the middle, its depth about 9 in its width, without backward projecting angles. Gill-rakers 3-++9. Distance of dorsal spine from tip of snout 2; in the leneth; dorsal fin shorter than high, the highest ray 1} in the head. Distance of adipose fin from the dorsal equal to one-third the length of the latter fin; adipose fin twice as long as the dorsal fin, 3 in the length. Caudal deeply cleft, the lobes longer than the head, about 3 in the length. Anal fin short and high, its margin narrowly convex, its highest ray 2 in the head. Ventrals inserted below the vertical from the last dorsal ray, 1f in head. Pectoral spine strong, terete, short hooks on its outer margin near tip and short teeth nearly the entire length of the inner margin; 14 in the head. 126 CALIFORNIA ACADEMY OF SCIENCUS. A large, dark humeral spot; fins dusky, the usual pale bar on the dorsal fin near its base. Head 3%; depth, below dorsal spine, 5, above first anal ray 43; depth of caudal peduncle 8; Br. 7; D. I, 6; A. 10. 17. Rhamdia sebz (Cuv. & Val.) Pimelodus stegelichii & musculus Miller & Troschel; Pimelodus holo- melas & millert Giinther. Localities: Tonantins; Gurupa; Rio Janeiro; Bahia; Xingu; Santa Cruz; Cudajas; Sao Matheos; Rio Doce; Serpa; Tabatinga; Goyaz; Para; Tefté; Surinam; Villa Bella. 17. (a,) Rhamdia sebe kneri (Steindachner.) 22 Pimelonotus vilsoni Gill. Localities: Tabatinga; Jutahy. 18, Rhamdia foina (Miller & Troschel.) 19, Rhamdia humilis (Ginther. ) 20. Rhamdia cinerascens (Ginther.) 21. Rhamdia pentlandi (Cuv. & Val.) 22. Rhamdia quelen (Quoy & Gaimaid.) Pimelodus sellonis Miller & Troschel ; ? Pimelodus bahianus Castelnau; Silurus sapipoca Natterer; Pimelodus wuchereri Giinther; Pimelodus queleni cuprea Steindachner; Pimelodus cuyabe Steindachner, Localities: Santa Clara; Rio Mucuri; Juiz de Fora; Campos; Rio Jequitinhonha; Mendez; Rio de Janeiro; Macacos; Sao Matheos; Rio Parahyba; Cannavierias; Rio Grande do Sul. 23. Rhamdia multiradiatus (Kner.) Pimelodus arekaima Schomburgk descr. not plate. 24, Rhamdia sapo (Valenciennes. ) Locality: Rio Grande do Sul. Emperov’s collection. 25. Rhamdia hilarii (Cuv. & Val.) Locality: Bon Jardin, on the Rio San Francisco. 26. Rhamdia wagneri (Giinther. ) Pimelodus cinerascens Kner & Steindachner (not Ginther); Rhamdia bransfordii Gill. Localities: Gorgona; Rio Chagres; Rio Obispo; Turbo, Atlantic Coast of Central America. 27. Rhamdia longicauda Boulenger. 28. Rhamdia dorsalis Gill. SOUTH AMERICAN NEMATOGNATHI. 7) 29. Rhamdia poeyi sp. nov. ‘Type, one specimen. Length .175 m. (to base of caudal fin); Goyaz; Senhor Honorio. Width of the body about equal to its depth; much com- pressed toward the caudal fin. Head short, flat above, the sides not very steep, the greatest width 1} in its length. Fontanel not continued behind the eyes. Occipital pro- cess very short, triangular, reaching two-sevenths the dis- tance from its base to the dorsal spine. A few large pores about the head. Kye 3 in snout, 7 in head, 1} in the interorbital. Maxillary barbels scarcely reaching the vertical from the dorsal spine; post-mental barbels about reaching to the gill opening, the mentals less than half as long. Branchiostegal membranes separate to below the anterior part of the eyes. Gill-rakers 3-+7. Distance of dorsal spine from tip of snout less than 3 in the length. Distance of adipose fin from the dorsal fin equals the length of the dorsal; adipose fin very low in front, its base about 3} in the length. Pectoral and ventral fins very short; pectoral spine short, terete, about half as long as the head, its outer margin smooth, the inner margin with strong teeth the entire length. Color brown, the sides and back densely covered with small brown dots and more minute black dots; a dark humeral spot, a narrow dark lateral band. Head 41; depth, above first anal ray, 7; depth of pe- duncie 94; Br. 6; D. I, 6; A. 18. 30, Rhamdia tenella sp. nov. Type, No. 7547. Length 31m. Cudajas. Width behind the head equal to the depth, tapering to the strongly compressed, caudal peduncle. Head broad and flat transversely, its greatest width 14 in its length; profile regularly and strongly convex; width at angle of mouth 2 in length of head. Fontanel a narrow slit be- 128 CALIFORNIA ACADEMY OF SCIENCES. tween the frontal bones. Occipital process extending a third of the space from its base to the dorsal spine. Hye, 3 in snout, 9 in head, 23 in the interorbital. Maxillary barbels reaching beyond the dorsal fin; mental barbels to base of pectorals, post-mentals to middle of pec- torals. Lower jaw shorter than the upper; intermaxillary band of teeth twice as deep at outer margins as at the middle; the greatest depth of the band 33 in its width. Gill membranes separate to below angle of mouth; gill- rakers slightly branched, 3-+-10. Distance of dorsal fin from tip of snout 3 in the length; dorsal rays of nearly uniform height, higher than the spine. Distance of the adipose from the dorsal fin 2 in the length of the dorsal fin, its distance from the caudal less than a diam- eter of the eye. Caudal deeply forked, the lower lobe broader, rounded, 14 in head. Anal longer than high, its longest ray 2} in head; tip of the anal fin not reaching the vertical from the end of the adipose by the length of the snout. Ventrals inserted behind the vertical from last dorsal ray, 2 in head. Pectoral spine very strong, scarcely shorter than the longest ray, recurved hooks on upper three-fourths of its outer margin; a deep groove almost the upper half of its inner margin, the basal third roughened; its length 1? in head. Color brownish; fins dusky; adipose margined with black. Head 44; depth 63; depth, above first anal ray, 8; depth of caudal peduncle 10; Br. 7-8; D. I, 6; A. 11. SOUTH AMERICAN NEMATOGNATHI. 129 § Rhamdella subg. nov. Type, Rhamdia eriarcha, noy. Fontanel continued to the occiput, a bridge across it be- hind the orbit. 31. Rhamdia microcephala (Reinhardt). $2. Rhamdia notata (Schomburgk). 33. Rhamdia eriarcha sp. nov. One specimen .26 m. in length. Dom Pedro. Rio Grande do Sul. Body elongate, slender, the width below the dorsal spine less than the depth. Head long, pointed, entirely covered with skin; the fontanel continued to base of occipital process, which reaches about half way to dorsal spine; the head flattish above, snout long and pointed; the width at angle of mouth 2 in length of head. Eye large, 2 in snout, 4} in head, its center about one- fourth nearer posterior margin of opercle than to tip of snout, the interorbital less than a diameter of the eye. Maxillary barbels reaching posterior edge of opercle; post-mental barbels not to gill-opening if drawn horizontally backward. Upper jaw produced; lips thick, strongly plicate; inter- maxillary band of teeth deep and narrow, the teeth minute; no teeth on vomer or palatines; teeth of the mandible similar to those of the upper jaw, in a narrow band which tapers backward. Gill-membranes separate to below center of eye; gill- rakers 1--5. Distance of dorsal fin from tip of snout 2§in the length; first ray of dorsal scarcely spine-like, smooth, its height, in- cluding filament, 13 in length of head; highest ray 1; in head, higher than the length of the fin. Adipose fin 25 in the length, distant from the dorsal fin as far as the base of the dorsal fin is long. Caudal forked to its base, the upper lobe greatly pro- 130 CALIFORNIA ACADEMY OF SCIENCES. duced, much longer than the head (broken off in the speci- men). Anal long and low, the posterior rays highest, 2} in head, the tips not reaching as far as the end of the adipose by more than one diameter of eye. Ventrals broad, reaching the anal, 1? in head. Pectoral spine strong, both edges roughened, 13 in head, the longest rays 1} in head. Color, chocolate above a black lateral band, much lighter below it; fins plain. Head 4?; depth 7; depth of caudal peduncle 14; D. I, 6; A. 15. 34, Rhamdia exsudans (Jenyns). 35.” Rhamdia jenynsii (Giinther). Pimelodus gracilis Jenyus (not Val.). Locality: Maldonado. *Nore.—The following two species from Mexico are in the collection of the Museum of Comparative Zodlogy. Rhamdia Parryi sp. nov. Types, No. 26292 a, Six specimens .093—.10 m. in length. Rio Zana- leneo, near Tonala, Chiapas, Mexico. Related to Rhamdia jenynsii Giinther. Head conical, its width 1} in its length, its depth at base of occipital pro- cess 14-13; top of head entirely covered with thin skin; fontanel extending from in front of the eyes nearly to the base of the occipital process, sepa- rated into two parts by a narrow bridge opposite the posterior orbital rim; occipital process very narrow and short, the exposed portion as long as wide, extending backward, imbedded in the skin, one-third the distance from its base to the dorsal spine. Dorsal plate rather large, triangular, entirely hid- den by skin, not articulating with the occipital process; space between an- terior and posterior nasal openings equal to the diameter of the eye which is 2 in anteorbital, 1} in interocular, 5 in head. Maxillary barbels short, extending to or beyond dorsal fin, mentals and post-mentals inserted in a nearly straight line, the former reaching edge of gill-opening, the post-mentals to base of pectoral. Lower jaw entirely included; width of mouth 22-3 in head; band of teeth of equal depth in both jaws, the intermaxillary band about 3 in head. Dorsal spine 12 in head, equidistant from snout and first anal ray, highest SOUTH AMERICAN NEMATOGNATHI. 163) 36. Rhamdia minuta Liitken. Localities: Rio Janeiro; Macacos. VII. PIMELODELLA gen. noy. Pseudorhamda Steindachner (not Bleeker). Type, Pimelodus cristatus. Head entirely covered with skin; occipital process nar- row, of the same width throughout, meeting the dorsal plate; fontanel prolonged backward to the occipital pro- cess with a bridge across it behind the eye. ray 12 in head, the dorsal fin about one-third higher than long; distance be- tween first dorsal and adipose fins greater than the length of the dorsal; adipose fin 3-33 in the length. Caudal lobes short and broad, the lower one very broadly rounded, the upper slightly longer and more pointed, 4-5 in the length. Anal rays short, more than two in head. Ventrals 1? in head. Pectoral spine depressed, short and stout, slightly more than one-half the length of the head, with numerous straight teeth on the posterior margin to near the tip, the longest about one-third the width of the spine; recurved hooks on the anterior margin. Pores on the chin and a few minute ones on head. Color light chocolate; « well-defined black band along the sides, continued through the eye and around the snout; numerous dark points scattered over the body; a purplish area at base of dorsal; sometimes narrow vertical bars along each side of the dorsal rays; caudal dusky. Head 44-43; depth 44-5; Br.6o0r7. D.I, 5-6; A, 12-14. Rhamdia petenensis Ginther. Pimelodus petenensis Giinther, Cat. Fish. Brit. Mus. V. 126, 1864 (Lake Peten); id, Fish. Central America 393, 1866. Three specimens: length .115—.125 m. Rio Zanaleneo, Chiapas, Mexico. Body below the dorsal spine scarcely deeper than wide, tapering to the caudal peduncle; head little wider than the body, its greatest width 1} to 13 in its length; flattened above; the profile steep and little convex, not de- curved at the snout; snout flat and broad, its width at the angle of the mouth 2-2} in the length of the head. Fontanel becoming very narrow behind the eye, continued to the base of the occipital process. Occipital process narrow, not reaching half way to the dorsal plate. Eye small, round; little nearer tip of snout than to opercular margin; 2-23 in snout, 2 in interorbital, 6-63 in head. Maxillary barbels reaching to end of dorsal or little beyond base of ven- 132 CALIFORNIA ACADEMY OF SCIENCES. 37. Pimelodella cristatus (Miller & Troschel). Pimelodus insignis Schomburgk; description, not plate; Pimelodus agasizit Steindachner; Pimelodus ophthalmicus Cope. Localities: San Goncallo; Avary; Villa Bella; Jutahy; Topajos; Rio Mucuri; Tabatinga; Hyavary; Coary. 38. Pimelodella wesselii (Steindachner). Localities: Cudajas; Para; Marajo; Rio Madeira; Rio Puty; San- tarem. 39, Pimelodella gracilis (Valenciennes). Locality: Goyaz. 40. Pimelodella pectinifer sp. nov. Type, No. 7508; length .19m., Campos. Hartt & Copeland. Body robust, compressed posteriorly; head conical, its greatest width 13 in its length, its depth at base of occipital process equal to its width. Profile straight from dorsal to eyes, thence decurved forward; head convex transversely. Lateral margins of occipital process concave; distance from the anterior to the posterior nostril 14 in eye. trals; mental barbels to base of pectozal, post-mental about to middle of pectorals. Mouth wide, terminal, the lower jaw slightly the shorter; teeth of the up- per jaw in a band of uniform depth, interrupted in the middle, its depth about 6 in its width. Gill membranes separated by a deep notch; gill-rakers 2-6. Distance of dorsal fin from snout 23-24 in the length; dorsal fin as long as high, its free margin rounded; highest dorsal ray 1?-2 in head; distance of the adipose from the dorsal fin equals the length of the dorsal; adipose fin 2$-3} in the length. Caudal divided to near its base, the lower lobe much the broader, rounded; the upper lobe pointed; the lobes about equal in length, 1} in head. Anal short and high, about as high as long, 2 in head. Ventrals inserted considerably behind the vertical from the last dorsal ray, about half as long as the head. Pectoral spine strong, recurved teeth along its anterior margin, strongest near the tip; much weaker teeth along its posterior margin to near its tip, its length 24-24 in head; the longest ray 13 in head. Body brownish, covered with numerous dark dots; lower part of cheeks and opercles similarly spotted; fins with small spots. Head 4-41; depth 54-64; depth of caudal peduncle 10; Br. 7; D.I, 6; A. 11-12. SOUTH AMERICAN NEMATOGNATHI. 135 Maxillary barbels not extending to the adipose fin; men- tal barbels reaching base, post-mentals to middle, of pec- toral fins. Width of intermaxillary band of teeth 4 in length of head. Eye 13 in snout, 4 in head, nearer posterior margin of opercle than to the tip of snout. Gill-membranes separate to below anterior margin of the eyes. Humeral process extending to middle of pectoral spine. Dorsal spine inserted nearer tip of snout than to origin of anal, its height 1} in head, the highest dorsal ray about equal to the length of the head. Adipose fin 4 in the length, its distance from the dorsal fin equals three-fifths of its own length. Caudal fin about 4 in the length. Anal fin convex, its highest ray 1? in head. Ventrals inserted behind the vertical from last dorsal ray, not reaching the anal by one-third their length. Pectoral spine 1 in length of head, the fin not reaching ventrals by one-half its length. Color uniform light brownish. Head 42; depth 53; Br. 6; D. I, 6; A. 12. 41, Pimelodella modestus (Giinther). 42. Pimelodella elongatus (Ginther). 43. Pimelodella lateristriga (Miller & Troschel). Localities: Santa Clara; Rio Mucuri; Rio Doce; Cannavierias; Sao Matheos; Mendez; Rio Trombetas; Obidos; Ica. 44, Pimelodella harttii (Steindachner), 45, Pimelodella buckleyi (Boulenger). Localities: Rio Parahyba; Macacos. 46, Pimelodella vittata (Kréyer). Localities: Rio San Francisco; Minas Geraes; Sao Matheos; Rio Jequitinhonha. . 47, Pimelodella chagresi (Steindachner). Locality: Obispo River. 48, Pimelodella brasiliensis (Steindachner). 134 CALIFORNIA ACADEMY OF SCIENCES. DOUBTFUL SPECIES OF Pimelodella orn or Pimelodus. I. Rhamdia cyanostigma Cope. II. Silurus quadrimacu- latus Bloch. VIII. Pimetopus Lacépéde. Pseudarioides & Pseudorhamdia Bleeker; Pseudorhamdia Liitken. 49, Pimelodus eques Miller & Troschel. Localities: Amazon near Fonteboa; Teffé; Obidos; Villa Bella; Jose Fernandez; Xingu; Tonantins; Hyutahy; Lake Hyanuary. 50. Pimelodus ornatus Kner. * Silurus megacephalus Natterer. Locality: Goyaz. . Pimelodus albicans (Cuv. & Val.) Arius albidus Val.; Arius moroti Val. fide Giinther. Locality: Rio Plata. 5 — 52, Pimelodus pictus Steindachner. Locality: Hyavary. 53. Pimelodus clarias (Bloch.) Pimelodus maculatus Lacépede; Pimelodus rigidus Spix; Pimelodus blochit Cuv. & Val.; Pimelodus arekaima Schomburgk (plate, not description); Mystus ascita Gronow; Pimelodus macronema Bleeker; Pseudariodes albicans Liitken; ? Pseudariodes pantherinus Liitken; Pseudorhamdi piscatrix Cope; Piramutana macrospila Ginther. Localities: I. var. (?) maculatus, Rio das Velhas; Uraguay; Jequit- inhonha; Rio Grande; Goyaz. II. vars. (?) macrospila, arekaim1, piscatrix, and noy., Para; Porto do Moz; Santarem; Obidos; Villa Bella; Coary; Teffé; Fonte Boa; Tabatinga; Hyavary; Cudajas; Rio Goneallo; Montalegre; Rio Reto; Rio Puty; Manacapuru; Tonantins; Lago Alexo; Cameta. III. var. (?) nov., Rio San Francisco. IV. var. (?) nov., Avary; Para; Ica; San Goneallo; Jutahy; Rio Puty. V. var. (?) macronema, Amazon. We have repeatedly examined and compared the 300 or more speci- mens in the collection, and can come to no other conclusion than that the many synonyms given here all refer to this extraordinarily variable species. 54. Pimelodus grosskopfii Steindachner. SOUTH AMERICAN NEMATOGNATHI. Las 55. Pimelodus labrosus Kroyer. 56, Pimelodus valenciennis Kroyer. 57, Pimelodus westermanni Reinhardt, 58. Pimelodus altipinnis Steindachner. Locality: Para. 59, Pimelodus fur Reinhardt. Pimelodus microstomus Steindachner. Localities: Rio das Velhas; Giquitiba; Rio San Francisco. IX. GEN. Nov. ? Head behind the eyes granulated; fontanel not extending behind the eyes; occipital process not reaching the dorsal plate. Barbels compressed, ribbon-like. 60. Pirinampus (?) agassizii Steindachuer. X. ConoruyncHos Bleeker. § Conorhynchos. 61. Conorhynchos conirostris (Cuv. & Val.) § Nov. a 62. Conorhynchos glaber Steindachuer. XI. Bacropsis Liitken. 63. Bagropsis reinhardti Liitken. XIT. Prramurana Bleeker. 64. Piramutana piramuta (Kner. ) XIII. PuAatTyNemMatTicHtTHys Bleeker. 65, Platynematichthys punctulatus (Kuer.) Bagrus nigripunctatus Kuer. 66. Platynematichthys araguayensis (Castelnau.) XIV. PHRACTOCEPHALUS Agassiz. 67. Phractocephalus hemiliopterus (Bloch & Schneider.) Phractocephalus bicolor Agassiz. Localities: Xingu; Coary; Teffé; Manacapuru ; Obidos; Lake Hy- anuary. 2D SER. Vou. I. (10) Issued July 18, 1888. 136 CALIFORNIA ACADEMY OF SCIENCES. XV. Scrapes Miller & Troschel. Leiarius and Sciadeichthys Bleeker. g Sciades. 68. Sciades pictus Miller & Troschel. Arius longibarbis Castelnau. Localities: Rio Negro; Villa Bella. + § Sciadeoides nov. Surface of the head and occipital process entirely covered with thick skin. Head little longer than broad. 69. Sciades marmoratus Gill. Locality: Tabatinga. XVI. Bracuypiatystoma Bleeker. Piratinga and Malacobagrus Bleeker. 70. Brachyplatystoma filamentosum (Lichtenstein. ) 71. Brachyplatystoma vaillanti (Cuv. & Val.) Platystoma affine Cuv. & Val). Platystoma mucosa Vaillant.=Platy- stoma verrucosum Boulenger. Localities: Tabatinga; Para; Porto do Moz; Avary; Rio Puty; Juiz de Fora on the Parahyba. 72. Brachyplatystoma reticulata (Kner.) 73. Brachyplatystoma rousseauxii (Castelnau.) Bagrus goliath Heckel. Locality: Para. XVII. DvopaaTinus gen. nov. Head narrow, little depressed, the width of the snout 1} in the greatest width of the head. Upper jaw little longer than the lower. ‘Teeth all alike, the vomerine patches far removed from the palatine patches. Adipose fin twice as long as anal. ; This genus is most nearly related to Brachyplatystoma in which the teeth of the inner series of the intermaxillaries are long, slender and depressible; of the outer series short, recurved and fixed. SOUTH AMERICAN NEMATOGNATHI. 137 74, Duopalatinus emarginatus (Cuv. & Val.) Localities: Rio San Francisco; Rio das Velhas. XVIII. SreEINDACHNERIA gen. nov. Type, Sletndachneria amblyurus nov. No teeth on the palate; teeth on the vomer in one or two patches; no dermal ossifications about the head; adipose fin much longer than anal. 75, Steindachneria amblyurus sp. nov. Platystoma parahybe Steindachner (in part.) Types, No. 7324, 2 specimens; .55 and .38 m. long. Jequitinhonha. Hartt & Copeland. . Body slender. Head much depressed, its width at the angle of the mouth is two-thirds of the greatest width of the head which equals its length. Kye 3 in snout, 83 in head, 33 in interocular. Maxillary barbels extending to the adipose fin. Teeth on the vomer in two patches. Dorsal spine slender, 23-3 in head, the rays 1} to 2. Base of the adipose fin almost twice as long as the base of the anal. Pectoral spine 2 in head; its outer margin roughened, its inner margin with sharp teeth. Caudal fin broadly rounded. Upper parts brownish, marked with round spots which are smallest on the head, sometimes confluent into vermicu- lations; adipose and caudal fins with spots; membrane of dorsal fin sometimes with a vertical series of spots; other fins dusky. Head 4; Depth 74; Br. 9 or 10; D. I, 6; A. 12. S. parahybe has the caudal deeply lobed, D. I, 7. 76. Steindachneria parahybe (Steindachner.) Locality: Rio Parahyba near Juiz de Fora. Types. XIX. Gen. noy. ? Palatine teeth present, forming, with the vomerine teeth, a narrow uninterrupted band. 77. Platystoma litkeni Steindachner. 138 CALIFORNIA ACADEMY OF SCIENCES. XX. HeEmisorvupim Bleeker. 78. Hemisorubim platyrhynchos (Cuv. & Val.) Localities: Manacapuru; Rio Negro; Montalegre; Rio Puty; Lago Alexo; Obidos; Coary; Tabatinga; Hyavary; Tonantins; Sao Paolo. XXI. PsevporLatystoma Bleeker. Hemiplatystoma Bleeker. 79, Pseudoplatystoma fasciatum (Linneus) (in part.) ? Platystoma truncatum Agassiz; Platystoma punctifer Castelnau. Localities: Obidos; Coary; Hyavary. 79 (a). Pseudoplatystoma fasciatum intermedium var. nov. Type, No. 7321, one specimen, .42 m. long. Obidos. Maxillary barbels not reaching to the dorsal, mental bar- bel not to the pectoral, post-mentals to middle of pectorals. Vomerine teeth narrowly joined. Lower parts white, upper parts brownish, with numerous narrow, dark wavy lines extending to the edge of the belly; dorsal, pectoral and ventral fins plain yellow; caudal with round brown spots, the anal with a very faint dusky spot atthe base of the last rays and an obscure streak on the tip of the rays; head with small round brown spots. 79 (b). Pseudoplatystoma fasciatum tigrinum (Cuv. & Val.) Locality: Obidos. 80. Pseudoplatystoma coruscans (Agassiz. ) Sorubim caparary Spix; Platystoma pardalis Valenciennes; Platys- toma punctatum Cuy. & Val.; Platystoma orbignianum Valencien- nes; Platystoma forschhammeri Reinhardt. Localities: Manaos; Rio das Velhas; Rio San Francisco, below the falls. XXII. Sorvsimm Spix. Platystoma Agassiz (name preoccupied in Diptera.) 81. Sorubim lima (Bloch & Schneider, ) Sorubim infraocularis Spix; Platystoma luceri Weyenbergh. Localities: Manacapurn; San Paolo; Fonteboa; Tabatinga; Hyavary; Teffé; Obidos; Maraiion; Ucayale; Iga; Lago do Manimo; Para; Rio Puty; Jutahy. SOUTH AMERICAN NEMATOGNATHI. 139 XXIII. Sorvupimicutuys Bleeker. 82. Sorubimichthys planiceps (Agassiz. ) Sorubim pirauaca Spix; Platystoma artedii Giinther; Sorubimichthys ortont Gill. Localities: Manacapuru; Hyavary; Teffé. 83. Sorubimichthys spatula (Agassiz.) Sorubim jandia Spix. 84, Sorubimichthys gigas (Giinther.) XXIV. Puatystomaticutuys Bleeker. 85. Platystomatichthys sturio (Knuer.) Localities: Para; Curuca; Rio Muria. DOUBTFUL SPECIES. Genus ? I. Bagrus flavicans Castelnau. II. Bagrus punctula- tus id. XXYV. Tacuisurus Lacépede. Bagrus, Galeichthys, Arius Cuv. & Val.; Arius Bleeker; Sciades Miller & Troschel; Sciades Bleeker 1863, not 1858; Ariodes Miller & Troschei; Hexanematichthys, Cephalocassis Netuma, Guiratinga, Sciadeichthys 1858, not 1863, Selenaspis, Hemiarius, Pseudarius Bleeker; Leptarius, Notarius, Gill; Cathorops Jordan & Gilbert. DOUBTFUL SPECIES. I. Arius nigricans Valenciennes; IL. Arius obesus Schom- burgk. g Galeichthys. 86, Tachisurus lentiginosus sp. nov. Types, No. 4970, two specimens, each .89 m. long. Panama. Stein- dachner. Body nearly terete anteriorly, becoming compressed backward, the width above the pectorals a little greater than the depth. Head flat, depressed, its depth at base of occi- pital process 1j in its greatest width, which is about 1} in its length. Occipital process somewhat roughened, about 140 CALIFORNIA ACADEMY OF SCIENCES. twice as long as its greatest width, its margins straight and oblique; the middle of the fontanel above the posterior part of the eye. Head everywhere covered with skin; sides of the head and opercle with vermiculating canals. Eye 23 in snout, 83 in head, 43 in interocular, 2} in in- terorbital. Snout somewhat pointed; upper jaw very little project- ing; lips thick, teeth all villiform, the intermaxillary band strongly curved. Vomerine teeth in two oval patches, joined to the larger patches of the palatines. Mandibulary bands separate in front; the outer margins, if continued forward, forming an angle atthe symphysis. Maxillary barbels reaching beyond base of pectorals; mental barbels reaching about? toward the gill opening, the post-mentals to the gill opening. Gill membrane forming a broad free margin across the isthmus. Gill rakers3-++4. Pectoral pore minute. Hum- eral process pointed behind. Distance of the dorsal from the snout 2} in the length. The dorsal spine covered with a membrane, its outer mar- gin granular; its height 13 in the head, the first soft ray 13 the length of the fin. Distance of adipose fin from the dorsal 33-4 in the length, the fin adnate, longer than the dorsal. Caudal lunate, the upper lobe longer, somewhat falcate, 434-5 in the length. Anal fin tivice as long as high, the highest ray 2-2} in the head. Ventrals short and broad, 1} in the head. Pectoral spine covered with a membrane, 11-1? in head. Light brown, becoming nearly white below, the sides freckled; fins reddish. Head 4-45; depth 5-6; D. I, 6; A. 22. 87. Tachisurus peruvianus (Liitken). Locality: Callao. SOUTH AMERICAN NEMATOGNATHI. 141 § Sciadeichthys. 88. Tachisurus flavescens (Cuv. & Val.) 89. Tachisurus temminkianus (Cuv. & Val.) 90. Tachisurus emphysetus (Miiller & Troschel). 91, Tachisurus brandti (Steindachner). ? Sciades troschelii Gill. Locality: Panama. 92. Tachisurus mesops (Cuv. & Val.) 93, Tachisurus proops \Cuv. w Val.) Locality: Pernambuco. 94, Tachisurus passany (Cuv. & Val.) 95. Tachisurus albicans (Cuv. & Val.) Bagrus valenciennei Castelnau; Arius valenciennesii Giinther. § Selenaspis. 96. Tachisurus herzbergii (Bloch). Pimelodus argenteus Licépede; Bagrus pemecus Cuy. & Val.; Bagrus calestinus Miller & Troschel; Bagrus mesops Kner (not C. & V.); Hexanemitichthys hymenorhinus Bleekcr. Localities: Para; Curuca; Bahia. 97. Tachisurus dowii (Gill). Arius alatus Steindachner. Locality: Panama. CO ie 98. Tachisurus luniscutis (Cuv. & Val.) Localities: Porto Alegre; Bahia; Nazareth, near Bahia; Rio Janeiro; Para; Porto Seguro; Sao Matheos; Cannavierias. 99, Tachisurus parkeri (Traill). Arius quadriscutis Bleeker. 2 Notarius. 100, Tachisurus grandicassis (Cuv. & Val.) Localities; Maranhao; Bahia. 100a, Tachisurus grandicassis parmocassis (Cuv. & Val.) Localities: Bahia; Sao Matheos; Maranhao. 100b, Tachisurus grandicassis stricticassis (Cuv. & Val.) Localities: Bahia; Maranhao. 142 CALIFORNIA ACADEMY OF SCIENCES. § Netwma. 101, Tachisurus dubius (Bleecker). 102. Tachisurus kessleri (Steindachner). Locality: Panama. 103. Tachisurus insculptus (Jordan & Gilbert). 104, Tachisurus planiceps (Steindachner). 105. Tachisurus platypogon (Giinther). Localities: Panama; Acapulco; ? Goyaz. 106. Tachisurus osculus (Jordan & Gilbert). Locality: Panama. 107. Tachisurus elatturus (Jordan & Gilbert). 108. Tachisurus barbus (Lacépéde). Pimelodus commersoni Lacépéde; Bagrus barbatus Quoy & Gaimard; Pimelodus versicolor Castelnau. Localities: Campos; Rio Doce; Rio Grande do Sul. @ Hexanematichthys. 109. Tachisurus seemani (Giinther). ?? Arius assimilis Jordan & Gilbert (not Giinther). Locality: Panama. 110, Tachisurus jordani sp. nov. ? Arius assimilis Jordan & Gilbert (not Giinther). Types, No. 4945. Two specimens, J’; length .256-.26 m. Panama. Rather robust, the width little less than the depth; cau- dal peduncle compressed. Head heavy, little broader than high, its height 13 in its length, its width 12-14, its width at angle of mouth 2-2}. Interorbital area flat and smooth; posterior portion of the head finely and sparsely granular; opercle and humeral process smooth. Occipital process about as long as broad, unusually sharply keeled. Font- anel extending to above posterior part of eye, continued as a deep groove to base of occipital process. Hye large, 1? in snout, 54 in head, 2 in interorbital, 22-3 in interocular. Maxillary barbels extending to the pectoral pore, post- mental at least to the gill-opening, mental about three-fifths as long as the post-mental barbels. SOUTH AMERICAN NEMATOGNATHI. 143 Snout blunt, decurved; upper jaw a little produced. Teeth all villiform, those on the vomer forming two small, separate, ovate patches which are contiguous to the twice or thrice as large palatine patches. Gill-membranes forming a fold across the isthmus. Gill- rakers 6+ 9. Pectoral pore large; vertical series of pores on the sides. Distance of dorsal spine from the snout 27-2 in the length; the spine of the dorsal and pectoral fins granular on the basal half of their outer margins; the inner margins serrate almost their entire length; the spines equally long 1} in head. Distance of adipose fin from the dorsal 3? in the length, the adipose more than half as long as the dorsal fin, its posterior margin free. Caudal deeply forked, the upper lobe longer, somewhat faleate, 37-4 in the length. Anal fin about as long as high, deeply emarginate, its highest ray 2? in the head. Ventral fins not reaching the anal, 2 in the head. Top of head and dorsal surface dark blue with metallic lustre, becoming silvery below; lower caudal lobe dusky; basal half of the inner surface of the paired fins black. Head 32-32; depth 53-53; D. I, 7; A. 18. “a 111. Tachisurus ceerulescens (Giinther). 112. Tachisurus guatemalensis (Ginther). 113. Tachisurus assimilis (Giinther). 114, Tachisurus surinamensis (Bleeker). 115, Tachisurus dasycehpalus ‘Giinther). 116, Tachisurus longicehpalus sp. nov. Type, No. 4972. One specimen, ,j'; length 29 m. Panama. Stein- dachner. Elongate, slender; greatest width little greater than depth. Head long and depressed, its greatest width 13 in its length, its greatest depth little more than half its length. Top of head obscurely granular, the granules almost en- tirely concealed by the skin; interorbital area flat and with 144 CALIFORNIA ACADEMY OF SCIENCES. four ridges which are obscurely granular, the inner two bordering the fontanel, the outer ridges curved in front, extending obliquely backward from near posterior nasal openings; occipital process as long as broad, its margins concave; fontanel continued as a deep groove to base of occipital process; opercle faintly striate; humeral process entirely covered with thick skin, not granular. Eye lateral, but well above angle of mouth, its diameter 13 in snout, 6 in head, 3 in interocular. Snout depressed, rounded in front. Maxillary barbels extending scarcely beyond base of pectorals; mentals not reaching gill-opening. Upper jaw little projecting; width of mouth 25 in head. Intermaxillary teeth long and slender, the depth of the band 44 in its width; vomerine and palatine teeth obtusely conical, the vomerine patches separate, contiguous to, but not confluent with the palatine patches. Gill-membranes not forming an angle where they meet, with a rather broad, free margin. Gill-rakers short and thick, 4+-5. Pectoral pore small; vertical series of pores on the sides. Distance of dorsal fin from snout 23 in the length, the spine 1? in head, its outer margin with granular teeth near its base, its inner margin with short teeth. Distance of adipose fin from the dorsal, 3} in the length. Adipose fin much longer than high, as long as the dorsal fin. Caudal fin forked, the upper lobe one-third longer than the lower, very nearly as long as the head, 33 in the length. Anal fin emarginate, scarcely longer than high, its height 21 in the head. Ventrals reaching almost to the anal, about 2 in head. Pectoral spine a little longer than the dorsal spine, 13 in the head; its outer edge roughened, inner edge with rather sharp teeth. Color: brown above, sides silvery; entire under surface SOUTH AMERICAN NEMATOGNATHI. 145 sprinkled with brown dots: a black median line on the back; fins dusky; barbels blackish. Head 323; depth 63; D. I, 7; A. 20. 117, Tachisurus rugispinis (Cuv. & Val.) Locality: Para. 118. Tachisurus phrygiatus (Cuv. & Val.) Arius dierperinki Bleeker. Locality: Maranhao. 119. Tachisurus grandoculis Steindachner. Locality: Rio Doce. g Tachisurus. 120 Tachisurus nuchalis (Giinther.) 121. Tachisurus agassizii sp. nov. Type, No. 7670, one specimen; length .235 m. Rio Grande do Sul. Senhor Albuquerque. Body compressed, especially backward, the depth greater than the width. Head narrowed forward, its greatest width 3 in its length, its greatest depth about 14; width at the mouth 2} in the length of the head. Top of head granular, the granulations forming striz nearly everywhere; occipital process as broad as long, with a median ridge, the margins concave. Middle of the fontanel over the posterior mar- gin of the pupil, continued backward as a triangular groove to near base of occipital process; interorbital area with the usual four smooth ridges; humeral process roughened, covered with skin; sides of the head with reticulating mu- cous canals. Eye 1} in snout, 54 in head, about 3 in interocular, 2 in interorbital space. Maxillary barbels about reaching gilli opening; mentals extending about half way to gill opening, the post mentals not much longer. Upper jaw projecting; teeth villiform in the jaws, sub- granular or bluntly conical on the palate, small; no graunu- lar teeth in the inner series of the mandible; the palatine patches smaller and more diverging than in spiaii. 146 CALIFORNIA ACADEMY OF SCIENCES. Gill-rakers united, joined to the. isthmus. Gill-rakers 8+ 14. . Pectoral pore moderate. Distance of dorsal spine from snout 23 in the length, the spine 13 in head, serrate on its posterior margin, granulate in front along basal half, becoming nearly smooth above. Space between dorsal and adipose fins 3 in the length, the adipose shorter than the dorsal fin, free posteriorly. Caudal forked, 4? in the length. Anal fin a little longer than high, its highest rays about half length of head. Ventral fins about 2 in the head. Pectoral spine as long as the dorsal spine, but stronger, serrate on its inner margin, granular along its outer margin. Dorsal surface brown; sides and ventral surface silvery; fins smutty. . Head 34; depth 5; D. I, 6; A. 19. 122. Tachisurus fissus (Cuv. & Val.) ? Arius puncticulatus Cuv. & Val. 123. Tachisurus melanopus (Giinther.) 124. Tachisurus spixii (Agassiz. ) Pimelodus albidus Spix; Arius arenatus Cuv. & Val.; Arius laticeps Ginther, Localities: Maranhao; Bahia; Rio Janeiro; Para; Santos in Sao Paulo; Abrolhos. 125. Tachisurus fiirthii (Steinda~huer.) Locality: Panama. 126. Tachisurus variolosus (Cuv. & Val.) 127. Tachisurus multiradiatus (Giinther.) Bagrus ? arioides Kner (not B. arioides C. & V.) § Cathorops. 128. Tachisurus hypophthalmus (Steindachner ) 129. Tachisurus gulosus sp. nov. Two specimens, ¢; length .285—.29m. Panama. Steindachner. SOUTH AMERICAN NEMATOGNATHI. JAF Elongate, slender, the width scarcely less than the depth. Head broad, strongly depressed, the profile to tip of snout nearly straight; greatest width of head 13 in its length, its depth at base of occipital process about 2. Occipital pro- cess about as broad as long, obtusely keeled. Fontanel very narrow, its center above the posterior margin of the eye, continued backward as a groove. Top of head rugose posteriorly. Four bony ridges between the eyes, the inner two margining the fontanel, the outer extending obliquely backward from near the posterior nostrils; sides of head and snout with reticulating canals. Eye oblique, its center over the posterior end of the man- dible and about on a level with the rictus, its diameter 23 in snout, 8 in head, 5 in the interocular distance. Maxillary and post-mental barbels reaching to the middle of the pectoral fins or shorter; mental barbels to near base of pectoral. Jaws sub-equal, very thin; intermaxillary band of teeth very shallow in the middle; no teeth on the vomer; the space between the palatine patches equals 14 diameters of the eye, the teeth obtusely conical, in about three series in front, in a single series behind; mandibulary band of teeth little more shallow than the intermaxillary band. Gill membranes united and having a free margin. Gill- rakers 1} times as long as the eye, 12-130. Pectoral pore large; vertical series of pores on sides of the body conspicuous. Distance of dorsal fin from snout 24 in the length; dorsal spine rather slender, high, 14 in head; distance of adipose fin from the dorsal 33 in the length; adipose fin as high as as long, its posterior margin wholly free. Lower caudal lobe somewhat the broader, the upper lobe 12-13 in the head. Anal fin emarginate, the highest ray 22 in the head. Ventral fins truncate, 2?-24 in the head. Pectoral spine 13 in the head, its outer margin roughened; its Inner margin with recurved teeth. 148 CALIFORNIA ACADEMY OF SCIENCES. Color: bluish-gray above, silvery below; dorsal and caudal slightly dusky, other fins plain; barbels dusky. Head 34, depth 53-52; Br. 5; D. 1, 7; A. 23. This species is very closely related to Arius hypophthal- mus Steind. The differences are chiefly these: hypopthalmus. gulosus. Intermaxillary band of teeth six Intermaxillary band of teeth ten times as wide as deep. _ times as wide as deep. Palatine patches of teeth consider- Palatine patches of teeth 4 as wide ably deeper than the intermaxillary | as the intermaxillary band, four times band, twice as wide as deep, the dis- | as wide as deep, the distance between tance between them 4 (1 J. & G.) | them 1} diameters of the eye. diameters of the eye. Snout 3} in the head. Snout 32 in the head. 130. XXVI. GENIDENS Castelnau. Genidens genidens (Cuv & Val.) Genidens cuviert & granulosus Castlenau; Rhamdia luukidi Bleeker (in part). Localities: Porto Alegre; Santos; Rio Janeiro; Rio Sao Matheos; Rio Grande do Sul. XXVIII. Armuricutuys Baird & Girard. Pimelodus Bleeker, Silures de Suriname 65, 1864. 131, 132. 133 134, Ailurichthys panamensis Gill. Aelurichthys nuchalis Giinther. Locality: Panama. Ailurichthys bagre (Linneus). Galeichthys gronovii & eiduxii Cuv. & Val.; Bagrus micronemus Ran- zani; Mystus carolinensis Gronow; Pimelodus longifilis Mus. L. B. (fide Bleeker). Localities: Sao Matheos; Santos; Para; Curuca; Pernambuco; British Guiana; Bahia. Ailurichthys pinnimaculatus Stcindachner. Locality: Panama. Ailurichthys marinus (Mitchill). Silurus bagre (Bloch, not of Linnzus); Felichthys filamentosus Swain- son; Galeichthys parre Cuv. & Val.; Galeich'hys blochtt Cuv. & Val.; Galeichthys bahiensis Castelnau; Aelurichthys longispinis Ginther. Localities: Rio de Janeiro; Para; Bay of Balaxy; Mobile Bay; Per- nambuco; Victoria, Brazil. SOUTH AMERICAN NEMATOGNATHI. 149 XXVIII. Paraprptomystes Bleeker. Paradiplomystax Ginther. 135, Paradiplomystes coruscans (Lichtenstein). XXIX. Dzrpriomystss Bleeker. Diplomystax Ginther. 136. Diplomystes papillosus (Cuv. & Val.) 2? Silurus chilensis Linneus; Arius carcharias Leyboldt; A. villosus, squaius, micropterus and synodon Philippi. Locality: Rivers of Santiago. XXX. AGENEIOSUS Lacépede. Ceratorhynchus Agassiz; Hypothalmus Schomburgk; Pseu- dogeneiosus & Davalla Bleeker; Ageniosus Ginther. 137. Ageneiosus inermis (Bloch). g Ageneiosus. 138. Ageneiosus brevis Steindachner. Locality: Coary. 139. Ageneiosus atronasus sp. nov. One specimen; length .095 m. Exact locality unknown. Body as broad as deep under the dorsal spine. Head short, depressed, not much elevated behind the eyes, the profile nearly straight, much less steep than in brevis, the head covered with thin skin, the surface of the bones longi- tudinally ridged; greatest width of the head 14 in its length, width at rictus 1}; snout as in brevis, broadly rounded in front. Fontanel short, ending over center of eye, but continuing as a groove backward to the posterior margin of the eye. Occipital process scarcely as long as its basal width, codssified with the dorsal plate, its surface ridges somewhat granular. Kye 1 in snout, 34 in head, 2 in the interocular. Maxillary barbel bony at base, the bony portion being two-thirds of the whole length of the barbel, extending be- yond the rictus half the total length of the barbel, its lower edge veryminutely crenulate near its tip. 150 CALIFORNIA ACADEMY OF SCIENCES. Snout short, as in brevis, about 2 in the interocular width. Upper jaw everywhere narrowly projecting, the intermax- illary band of teeth about as in brevis. Lateral line zig-zag, with branches extending from the angles. Distance of dorsal fin from the snout 23 in the length; the dorsal spine 14 in the head, roughened in front; short teeth on its inner margin except near base. Distance be- tween the dorsal and adipose fins 2} in the length; adipose fin high and short. Caudal forked, the lobes pointed, 1 in the head. Origin of anal three-fifths nearer base of caudal than to the rictus. Ventrals reaching the anal, 1} in head. Pectoral spine 13 in head, otherwise like that of brevis. Color, purplish on the back; head dotted above and be- low with purple; upper lip dark purple, the lower lip with a very narrow purple margin; a rather obscure lateral band formed of dark dots; minute punctulations above the anal fin; all the fins more or less dotted with purplish, the caudal narrowly edged with that color. Head 33; depth, below dorsal 6, above anal 5; D. I, 6; A. 30. 140. Ageneiosus valenciennesi Bleeker. Locality: Rio Puty. 141. Ageneiosus armatus Lacépéde. 142, Ageneiosus ucayalensis Castelnau. Locality: Para. 143, Ageneiosus dentatus Kner. Ageneiosus purdalis Liitken. Localities: Teffé; Para; Cameta. 144, Ageneiosus porphyreus Cope. 145. Ageneiosus dawalla (Schomburgk). Ageneiosus inermis Cuy. & Val., notof Bloch; Ageneiosas sebe Giinther. § Pseudageneiosus. 146. Ageneiosus brevifilis Cuv. & Val. Locality: Serpa. . SOUTH AMERICAN NEMATOGNATHI. 151 147, Ageneiosus axillaris Giinther. XXXI. TETRANEMATICHTHYS Bleeker. 148. Tetranematichthys quadrifilis (Knuer.) XXXII. AUCHENIPTERUS Cuv. & Val. Euanemus Miller & Troschel. 149, Auchenipterus nuchalis (Spix.) Auchenipterus dentatus Cuv. & Val.; Huanemus colymbetes Miiller & Troschel. Locality: Villa Bella. ? 150. Auchenipterus fordicei sp. nov. Type, No. 7289, one specimen, length .12 m. Coary. L. Agassiz. Body short, deep, compressed. Head short; its width equals the distance from tip of snout to middle of opercle; its depth at base of occipital process is little less than its width; top of the head covered with very thin skin, the sur- face of the bones striate. Occipital process twice as broad as long, parabolic behind, joining the dorsal plate: dorsal plate deeply emarginate in front and behind, less so on sides, its length on a median line 23 in the diameter of the eye. Kye 1 in snout, 3§ in the head, 23 in the interocular space. Maxillary barbel reaching beyond base of ventrals, its osseous base not extending to center of eye; mental barbels reaching to the middle of the ventruls; post-mentals not to the middle of the pectorals. Snout rounded in front; jaws about equal, the width of the mouth at the rictus 13 in the width of the snout at the same place. Teeth as in ZF. nuchalis. Distance of dorsal fin from tip of snout 33 in the length; dorsal spine 1} in head, slender, its anterior margin smooth, its posterior margin with short teéth, the tip of the spine, when depressed, reaching the vertical from the end of the first third of the ventral fins. 2D SER. Vou. I. (11) Issued July 18, 1888. Ge! CALIFORNIA ACADEMY OF SCIENCES. Caudal fin 42 in the length. Origin of anal equidistant between snout and base of caudal. Ventral fins large, one-sixth longer than the head, the inner rays connected by a membrane, the tips of the longest rays extending to the 8th or 9th anal ray. Pectoral spine long and slender, 1 in head, curved, reach- ing a little beyond origin of ventrals, the outer edge of the spine smooth, the inner with sharp, recurved teeth. Pee- toral pore large. Color nearly plain, the back somewhat darker than be- low, the inner rays of the pectoral and ventral fins dusky; fins otherwise plain, light. Head 5; depth, at dorsal spine 5; greatest depth 43; D. I, 6; A. 46; V. 13-14. We have named this species for Mr. Morton W. Fordice, a student of American fishes. 151. Auchenipterus brachyurus (Cope.) XXXII. Epaprerus Cope. 152. Epapterus dispilurus Cope. Huanemus longipinnis Steindachner. Locality: Hyavary, types of longipinms. XXXIV. Fenicatuys Swainson. Pseudauchenipterus Bleeker. 153. Felichthys jequitinhonhe (Steindachner. ) Locality: Jequitinhonha. 154, Felichthys flavescens sp. nov. Type, No. 7362, one specimen, 9, .10m. Rio San Francisco. Hartt. Elongate compressed, tapering rapidly to a slender caudal peduncle; depth everywhere much greater than the width. Head short and deep, its greatest depth little more than its greatest width, which is 13 in the length of the head; width at the rictus 2; head covered with loose skin, the surface of the bones not evident; frontal bones not swollen nor honey- SOUTH AMERICAN NEMATOGNATHI. 1d3 comb-like; frontal fontanel open in front, extending back- ward to about the center of the eye. Occipital process 3 times as broad as long. Dorsal plate very widely forked in front, its width at the narrowest place more than twice its length on a median line; the posterior portion of the head strongly convex in a transverse section. Four pores on the snout, no conspicuous pores on sides or top of head. Eye longer than snout, 38} in the head, 1} in the inter- ocular. Maxillary barbel reaching to near tip of pectoral; mental barbel to base of pectoral, post-mental a little beyond its base. Snout rounded, upper jaw longer; teeth all villiform, ar- ranged as in jequitinhonhe. Gill-opening extending to base of pectoral. Humeral process covered with thin skin, somewhat roughened on its surface, extending a little beyond middle of pectoral. Pee- toral pore large. Lateral line undulating. Distance of dorsal spine from snout 3 in the length; dor- sal spine 14} in the head, its outer margin scarcely rough- ened, its inner margin slightly serrate; first soft ray 1? in the head. Distance between dorsal and adipose fins 24 in the length. Caudal deeply forked, the lobes pointed, 3? in the length. Anal emarginate, the first rays (female) not reaching to base of the last rays. Ventrals not reaching the anal, 1} in the head. Pectoral spine slightly shorter than the head, its outer edge and sides smooth, its inner margin strongly serrate. Back brown, the color composed of numerous brown dots; top of head and snout blotched with brown on a yellow ground; opercle and a triangular spot behind the eye yel- low; humeral region covered with numerous conspicuous brown dots; sides and ventral surface yellow; upper caudal lobe dusky, otherwise fins plain yellow, Eead 425 depthi ot; Dt 116; A. 20; V.-8;'P. I, 62%. 154 CALIFORNIA ACADEMY OF SCIENCES. 155. Felichthys affinis (Steindachner. ) Localities: Sao Matheos; Para. 156. Felichthys nodosus (Bloch.) Auchenipterus furcatus Cuy. & Val. Localities: Para; Bahia. XXXYV. AUCHENIPTERICHTHYS Bleeker. 157. Auchenipterichthys thoracatus (Kner). Localities: Coary; Hyavary. 158. Auchenipterichthys longimanus (Ginther.) Localities: Manes, Rio Madeira; Cameta. XXXVI. TracuycorysTEs Bleeker. 159. Trachycorystes glaber (Steindachner.) 160. Trachycorystes isacanthus (Cope.) 161. Trachycorystes insignis (Steindachner.) 162. Trachycorystes obscurus (Ginther. ) 163, Trachycorystes magdalene (Steindachner.) 164, Trachycorystes trachycorystes (Cuv. & Val.) Trachycorystes typus Bleeker. 165. Trachycorystes ceratophysus (Kner.) 166, Trachycorystes porosus sp. nov. Types, No. 7351, two specimens; length .13-.15 m. Brazil. Short and robust, little compressed; width of the head about equal to its length, its depth 1} in its length. Bones of the head coarsely granular. Fontanel oval. Snout and sides of the head with conspicuous pores. Three grooves on the occipital bone form a z-shaped figure, and are stud- ded with pores; other pores on top of the head. Eye circular, 2 in the snout, 7 in the head, 44 in the interocular. Maxillary and post-mental barbels extending a little beyond base of pectoral, mental barbels extending beyond insertion of post-mentals, measuring 12 in the length of the head. SOUTH AMERICAN NEMATOGNATHI. hae Lower jaw projecting; teeth fine, the intermaxillary band about 8 times as wide as deep. Humeral process extending obliquely upward, strongly granular, its lower margin serrate; reaching a little beyond middle of pectoral spine. Distance of dorsal spine from snout 3-33 in the length; dorsal spine slender, 13-2 in the head, its inner margin roughened, its anterior margin with a median series of diverging spines and two series of smaller teeth; the first ray higher than the spine. Space between dorsal and adipose fins 2} in the length; anterior margin of the adipose fin continuous with the profile of the back. Caudal obliquely truncate. Anal fin strongly convex behind, higher posteriorly than anteriorly. Pectoral spine strongly serrate, the outer teeth much longer than the inner oues, especially near the tip of the spine, the flattened sides strongly granular, the granules enlarged anteriorly, forming serre along either side of the long spine-like marginal teeth; the length of the spine 5-6 in the length. Lateral line somewhat undulatizg, with conspicuous pores. Dark reddish brown, with longitudinal interrupted bands; belly plain, lower side of head thickly dotted with darker; dorsal and anal fins spotted; caudal fin with irregu- lar dark cross bars; inner surface of ventral and pectoral fins more or less dusky. Head 4-43; depth about 33; D. I, 5; A. 25; V.6; P.I, 7. 167. Trachycorystes striatulus (Steindachner. ) Localities: Linhares, Rio Doce; Itabapuana; Campos; Sao Matheos; Gurupa. 168. Trachycorystes brevibarbus (Cope.) 169, Trachycorystes galeatus (Linnzus.) Aucheniplerus maculosus, inmaculatus & punctatus Cuv. & Val. Au- chenipterus iacustris Liitken. 156 CALIFORNIA ACADEMY OF SCIENCHS. Localities: Pernambuco; San Goncallo; Rio San Francisco, below the Falls; Tabatinga; Teffé; Rio Puty. ° 170. Trachycorystes robustus Giinther. 171, Trachycorystes analis sp. nov. Type, No. 7354, one specimen; length, .21 m. Arary ? Professor Agassiz. Width of the head equal to its length; width of the mouth 1? in length of head; the depth at the occipital process 14 in the length of the head. Fontanel oval, surrounded by bone. Mental barbels less than one-third the length of the head. Dorsal spine 1} in head, its anterior margin smooth except near the tip, its,posterior margin slightly serrate. Caudal broken. Anal rays crowded, gradually decreasing in height back- ward, the margin undulating. Pectoral spine very strong, 34 in the length. Dark brown; ventral surface dusky with numerous dark points; dorsal fin with dark spots most prominent near the free margin; tips of ventrals dusky, the base profusely dotted, the intermediate region almost plain yellowish; pectoral fin profusely spotted. Head 4; depth, 42; D. I, 6; A. 41; V. 6; P. I, 7. Otherwise as in maculosus. XXXVIT. CrnTRomocuius Kner. % Centromochlus. 172, 2? Arius oncina Schomburgk. 173. Centromochlus heckelii (Filippi. ) Centromochlus megalops Kner. Localities: Villa Bella; Obidos; Para; Tabatinga; Gurupa; Manaca- puru; Lago Alexo; Hyavary. 174, Centromochlus steindachneri Gill. 175, Centromochlus intermedius Steindachner. Localities: Tajapuru; Teffé; Jatuarana; Ica; Jutahy; Lago Alexo. 176 177, 178, 179, 180. 180c, SOUTH AMERICAN NEMATOGNATHI. Centromochlus perugie (Steiudachner.) Centromochlus aulopygius Kuer. Locality: Cudajas. Centromochlus albescens Reinhardt. Localities: Rio Parahyba; Rio Janeiro; Macacos. XXXVIIL. WeERtTHEIMERIA Steindachner. Wertheimeria maculata Steind. XXXIX. TRACHELYOPTERUS Cuv. & Val. Trachelyopterus coriaceus Cuv. & Val. Localities: Obidos; Porto do Moz. Trachelyopterus coriaceus maculosus var. nov. Type, No. 7337. Porto do Moz. Senhor Vinhas. 157 Body light brown, with longitudinal series of rusty brown spots. 181. 182, 183. 184, 185. 186. 187, XL. TRACHELYOPTERICHTYS Bleeker. Trachelyopterichthys teniatus Kuer Locality: Teffé XLI. Cerrtopsis Agassiz. % Hemicetopsis. Cetopsis candiru (Spix.) Localities: Tabatinga; Jutahy; Tonantins; Ica. Cetopsis plumbeus Steindachner. Z Cetopsis. Cetopsis ccecutiens (Lichtenstein. ) Locality: Gurupa. 2 Pseudocetopsis. Cetopsis gobioides Kner. Q SUBGEN. NOY. ? Cetopsis occidentalis Steindachner. Cetopsis ventralis Gill. 158 CALIFORNIA ACADEMY OF SCIENCES. XLII. AsreropHysus Kner. 188. Asterophysus batrachus Kner. XLII. Hemiporas Bleeker. Doras Bleeker (1863, not of 1858.) g Hemidoras. 189. Hemidoras nattereri (Steindachuer.) Localities: Jutahy; Coary; Teffé. 190. Hemidoras brevis (Kuer.) 191. Hemidoras fimbriatus ( Kner.) 192. Hemidoras punctatus (Kner.) 193. Hemidoras lipophthalmus (Kuer.) 194. Hemidoras accipenserinus ((tiinther. ) 195, Hemidoras stenopeltis (Kner.) Localities: Manaos; Rio Negro; Hyavary; Manacapuru; Teffé; Obidos; Tabatinga. 196. Hemidoras stiibelii (Steindachner.) 197. Hemidoras morei (Steindachner. ) 198. Hemidoras humeralis ( Kner.) 199, Hemidoras carinatus (Linneus.) Doras oxyrhynchus Val. @ Hassar* SUBGEN. NOV. Shields on the anterior half of the body rudimentary or none; snout long, conical. 200. Hemidoras orestes (Steindachuer.) Lovality: Huytahy. 201. Hemidoras affinis (Steindachner.) Locality: Rio Puty. XLIV. Oxyporas Kner. Pseudodoras & Rhinodoras Bleeker. « The Arowack name for the species of Doras. SOUTH AMERICAN NEMATOGNATHI, 159 Z Oxydoras. 202. Oxydoras niger (Valenciennes. ) Doras humboldti Agassiz; Corydoras edentatus Spix; Rhinoderaus prio- nomus Cope. Rhinodoras teffeanus Steindachner. Localities: Teffé (types of teffeanus?); Gurupa; Mancapuru; Coary; Obidos. 203. Oxydoras knerii Bleeker. g Rhinodoras. 204. Oxydoras d’orbigny Kroyer. 205, Oxydoras Amazonum (Steindachner. ) XLY. Doras Lacépede. Centrochir Agassiz; Lithodoras, Pterodoras, Platydoras, Acanthodoras, Astrodoras & Amblydoras Bleeker; Zathorax & Agamyxis Cope. §$ Lithodoras. 206. Doras dorsalis Cuv. Val. Doras papilionatus Filippi; Doras lithogaster Heckel. Locality: Para. § Doras. 207. Doras uranoscopus sp. nov. Type, No. 7235, one specimen .54m. Lake Hyanuary. Navez. Body rather heavy, depressed, depth below the dorsal spine ? as great as the width; caudal peduncle rather slen- der, wider than high. Width of the head 1} in its length; top of the head to near tip of snout, opercle, preopercle, suborbital and pre-nasal bones striate, the striz becoming broken up into granules in places, which, on the dorsal plate are spine-like, similar to those on the lateral scutes. Dor- sal plate not continued behind the anterior margin of the dorsal spine, with a downward directed process. Interor- bital area flattish; posterior portion of the head obtusely keeled. Fontanel club-shaped, not extending to the pos- terior margin of the eye; an elongate diamond-shaped depressed area behind it. 160 CALIFORNIA ACADEMY OF SCIENCES. Eye 33-4 in the snout, 8} in the head, 3 in tke inter- orbital; eye more superior than lateral. Maxillary barbels extending to posterior 4th of humeral process; post-mentals to opposite the insertion of the inner pectoral ray; mentals about 3 as long as the post-mentals. Snout pointed, its width at the rictus 2} in the head. Upper jaw projecting; teeth villiform, the intermaxillary band about six times as wide as deep. Gill opening extending to a point midway between upper angle of preopercle and the eye. Lateral scutes high, the third hook-bearing one the high- est, 13 in the length of the head, the scutes decreasing in height to the last, the median hooks increasing in size to the caudal peduncle, the exposed surface of those in front of the peduncle thickly set with small spines. Accessory rays of the caudal fin ossified, the anterior one forming a small plate. Exposed part of the skin verrucose. Humeral process narrow, tapering backward, reaching a little beyond the middle of the pectoral spine. Distance of the dorsal fin from the snout about 2? in the length. Dorsal and pectoral spines with the sides deeply furrowed, both margins serrate; the serration of the poste- rior margin of the dorsal spine reduced to a few spines. Adipose fin low, merging into the profile of the back ante- riorly. Caudal emarginate, 5 in the length. Anal high, rounded, first ray highest, 2 in the head. Ventral as high as the anal fin. Pectoral spine not quite reaching the ventral fins. Pec- toral pore minute. Fins spotted with brown. Head 4; depth 5; Lat. 1. 18; D. I, 6; A. (injured 2?) 6; WeG hee 9: 208. Doras maculatus Valenciennes. ? Doras granulosus Val. Doras murica Natterer. Locality: Avary? SOUTH AMERICAN NEMATOGNATHI. 161 209. Doras longipinnis Steindachner. ? Doras crocodili Humboldt. 210. Doras albomaculatus Peters. 211. Doras helicophilus ‘iinther. 212. Doras dentatus Kner. 213. Doras costatus (Linneus). Localities: Rio Puty; Rio Preto; San Gongallo; Xingu Cascade; Obidos ; Gurupa; Teffé. 214. Doras armatulus Cuv. & Val. 215. Doras hancockii Cuv. & Val. 216. Doras brachiatus Cope. $ Acanthodoras. 217. Doras calderonensis Vaillant. Doras depressus Steindachner. 218. Doras cataphractus (Linneus.) Caiaphractus americanus Bloch & Schneider; Doras blochii Cuv. & Val.; ?Doras brunnescens Schomburgk; Doras polyramma & poly- gramma Heckel; Callichthys asper Gronow. 219. Doras spinosissimus sp. ov. Type, No. 7222, one specimen .15m. Coary. Professor L. Agassiz. Short and thick, the width below the dorsal spine greater than the depth; caudal peduncle little deeper than wide. Head as broad as long, its depth 14 in its length: top of head, opercle, preopercle and suborbitals finely granular; the nasal bones with their free margin spinous. Top of head broad and flattish. Dorsal plate broad, without prominent downward or backward processes. Sutures of the skull marked by smooth lines. Fontanel reduced toa small oval opening surrounded by granulations. Kye 1? in the snout, 7 in the head, 3 in the interorbital. Maxillary and post-mental barbels reaching to the pec- toral pore; mental barbels 3 as long as the post-mentals. Snout broad, its width at the rictus 14 in the head. Jaws subequal; teeth minutely villiform, the intermaxillary band six times as wide as deep. 162 CALIFORNIA ACADEMY OF SCIENCES. Gill-opening continued forward to below the upper angle of the preopercle. Humeral process not reaching the tip of the pectoral spine by about an orbital diameter, its sur- face with short sharp spines, a series of which near the lower margin, is enlarged. Lateral scutes very high, coy- ering almost the entire sides, those above the first anal ray highest, 13 in the head, those on the caudal peduncle meet- ing the scutes of the other side above and below; each lat- eral scute posterior to the dorsal plate has a median hook and 5-14 smaller spines above and below it. Basal half of the caudal rays with about five series of small spines. Distance of the dorsal fin from the snout 2} in the length; dorsal spine 12 in the length of the head; its posterior margin smooth, its sides and anterior margin with many short spines, a smooth groove between the spines of the sides and front margin. Distance between dorsal and adi- pose fins 3} in the length; adipose fin oval, as long as the dorsal fin without the spine. Caudal rounded, two in the head. Anal rounded, the central rays the highest, as high as the length of the caudal. Ventrals not reaching the anal, 2 in the head. Pectoral spine strong, not reaching the ventral fins; its lower surface bluntly granular, its upper with short teeth like those on the humeral process, both margins finely serrate. Brown, marked with white; a white lateral band not as wide as the eye; a median series of white spots on the back; ventral surface and sides of the head irregularly spotted with white; top of the head with a median, inter- rupted light band; dorsal, pectoral and ventral fins spotted and marbled with brown and white; posterior margin of the adipose white; caudal and anal fins with undulating cross-bars of white and brown. Barbels annulated with brown and white. we A SOUTH AMERICAN NEMATOGNATHI. 163 iesdra =) depth oa; Tat. le 265 Dot, 5; A..12;.V., 6; Ph: 220. 221, 222, 223. 224, 225. 226. 227. 228. 229. 230. Doras marmoratus Reinhardt. Localities: Rio San Francisco; Rio das Velhas, 2 Amblydoras. Doras affinis Kner. Doras truncatus Bleeker. Doras weddellii Castelnau. Doras grypus Cope. Localities: Fonteboa; Teffé; Serpa; Porto do Moz; Silva, Saraca. §$ Centrochir. Doras crocodili Humboldt. § Agamyzxis. Doras castaneo-ventris Schomburgk. Doras pectinifrons Cope. $ Astrodoras. Doras asterifrons Heckel. Localities» Jutahy; Teffé; Porto do Moz; Serpa. Doras heckelii Kner. Localities: Jutahy; Tonantins; Teffé; Tabatinga. Doras monitor Cope. Doras nauticus Cope. XLVI. Puysopyxis Cope. Physopyxis lyra Cope. ARGIIDA. I. Arces Cuv. & Val. Brontes Cuv. & Val. 1. Arges sabalo Cuv. & Val. 2. Arges prenadilla Cuv. & Val. A. brachycephalus Giinther. Lake 164 CALIFORNIA ACADEMY OF SCIENCES. 3. Arges longifilis Steindachner. 4. Arges peruanus S‘einda:hner. II. Asrropierus Humboldt. 5. Astroblepus grixalvii Humboldt. III. Cycrorrtum Swainson. Stygogenes Giinther. 6. Cyclopium cyclopum (Humboldt.) Cyclopium humboldti Swainson; Stygogenes humboldti Giinther. Locality: Quito. 7. Cyclopium giintheri Boulenger. CALLICHTHYID &. I. Scieromystax Giinther. 1, Scleromystax barbatus (Quoy & Gaimard) II. Carapuractus Bloch. Callichthys Cuy. & Val. 2. Cataphractus callichthys (Linneus.) Callichthys tamoata Linnzeus; Callichthys asper Quoy & Gaimard; Caliichthys celatus & leviceps Cuv. & Val.; Callichthys loricatus Gronow; Callichthys kneri Gill; Callichthys affinis Giinther; Cai- lichthys hemiphractus Hensel. Localities: Rio de Janeiro; Pernambuco; Juiz de Fora; Bahia; Men- dez; Macacos; Porto Seguro; Surinam. 3. Cataphractus arcifer Hensel. III. Hoptosternum Gill. > Hoplosternum littorale Hancock. Callichthys subulatus & albidus Cuy. & Val.; Callichthys leevigatus Val. Hoplosternum stevardii Gill. Localities: Surinam; Gurupa; Para; Santarem; Tabatinga; Avary; Silva, Lake Saraca; Villa Bella; Porto do Moz; Lake Hyanuary; Ueranduba. & Hoplosternum thoracatus (Cuv. & Val.) Callichthys longifilis Cuy. & Val.; personatus Ranzani; exaratus d pic- tus Mill. & Trosch.; su’catus Kner; chiquitos Castelnau. SOUTH AMERICAN NEMATOGNATHI. 165 Localities: Curupira; Tabatinga; Cudajas; Gurupa; Teffé; Lake Hyanuary; Villa Bella; Para; Uerunduba; Santarem; Porto do Moz; Pernambuco; Obidos. 6. Hoplosternum melampterus (Cope.) IV. DEcApoGon gen. noy. Lower lip with a series of six barbels; ventral surface en- tirely mailed. 7. Decapogon adspersus Steindachner. Localities: Cudajas; Tabating:; Porto do Moz. V. Dianema Cope. 8. Dianema longibarbis Cope. VI. Brocuis Cope. Chenothorax Cope. g ? nov. 9. Brochis taiosh (Castelnau. ) 2 Chenothorax. 10. Brochis bicarinatus (Cope.) 11. Brochis semiscutatus (Cope.) $ Brochis. 12. Brochis dipterus Cope. 13. Brochis ceruleus Cope. VII. Coryporas Lacépéde. Hoplisoma Swainson; Hoplosoma Gill; Gasterodermus Cope. 14, Corydoras eques Steindachner. Locality: Cudajas. 15 Corydoras splendens (Castelnau.) 16, Corydoras elegans Steindaciner. Locality: Cudajis. 17. Corydoras nattereri Steindaciner Localities: Rio Doce; Rio Parahy b.. 166 CALIFORNIA ACADEMY OF SCIENCES. 18. Corydoras eneus (Gill.) 19 Corydoras armatus (Giinther. ) 20. Corydoras paleatus (Jenyns.) Corydoras marmoratus Steindachner; Cullichthys punctatus Val. and Cuv. & Val. Locality: Uraguay. 21. Corydoras punctatus (Bloch.) Corydoras geoffroy Lacepede; Corydoras ambiacus Cope. Locality: José Fernandez. 22. Corydoras trilineatus Cope. Corydoras agassizii Steindachuer. 23. Corydoras acutus Cope. 24. Corydoras amphibelus Cope. 25. Corydoras hastatus sp. nov. Types, No. 7747, two specimens .022 m. Villa Bella. Prof. L- Agassiz. Compressed, comparatively slender. Head as deep as long, its width 14 in its length; profile straight, steep; occipital process triangular; fontanel elongate, extending into the occipital bone; preorbital small. Eye large, 1} in the snout, 33 in the head, 2 in the inter- orbital. Snout little decurved; rictal barbels not extend- ing beyond the eye; lower lip terminating in two barbels. Coracoid processes striate, forming a ridge on the sides of the belly. Distance of the dorsal spine from the snout two in the length; the dorsal spine little shorter than the head. Caudal deeply forked, 2? in the length. Pectoral spine little longer than the dorsal spine, weakly serrate along both margins. Light brown; a jet black lateral band terminating at the base of the caudal in a large arrow-shaped spot, which is bordered posteriorly with white, itself narrowly mar- gined with blackish, the caudal dusky beyond; a jet black line extends on either side from a short distance (or) ~I SOUTH AMERICAN NEMATOGNATHI. | behind the ventrals to behind the anal fin; body and fins everywhere covered with minute black points. Head 34; depth 23; D. I, 7-8; A. 7-8; Lat. 1. 22. LORICARIID A. I. PrEcostomus* Gronowius. Hypostomus Lacépede; Cochliodon Heckel. % Cochliodon. . Plecostomus cochliodon (Kuer.) Cochliodon hypostomus Heckel MS. —y § Plecostomus. 2 Plecostomus spinosissimus Steindachner. o Plecostomus emarginatus Cuv. & Val. H. horridus Kner; Pl. scapularius Cope; Pl. tenuicauda Steindachner. Localities: Cudajas; Santarem; Manacapuru; Tonantins; Obidos; Fonteboa; Tabatinga; Hyavary; Sao Paolo; Goyaz. 4. Plecostomus limosus sp. nov. Types, No. 7869. Four specimens .25m. Rio Grande do Sul, from the Emperor’s collection. *The species of this genus are all closely allied, and the new species can best be described by the following key, which represents the relationships, as near as we can trace them: ODIs 6. A. 5. c. Jaws with 7 or 8 broad tipped teeth (Cochliodon). cochliodon 1 cc. Jaws with numerous fine teeth (Plecostomus). d. Head with three distinct ridges, snout pointed, head usually elevated, occipital bone ending in a pointed occipital process. e. Ventral surface covered with small plates. f. Free margins of the lateral plates with long spines. spinosissimus 2 ff. Free margins of the lateral plates with short spines or none. g. Post-humeral ridge becoming strongest on the tail, which is flat below. emarginatus 3 gg. Post-humeral ridge strongest in front, lower surface of tail similar to the upper surface. 2p SER. Vou. I. (12) Issued July 18, 1888. 168 CALIFORNIA ACADEMY OF SCIENCES. 5, Plecostomus carinatus Steindachner. 6. Plecostomus commersonii (Valenciennes. ) H. puncta'us Cuv. & Val.; H. subcarinatus Castelnau; Pl. spiniger Hensel. Localities: Rio de Janeiro; Santa Cruz; Macacos; Itabapuana. 6a. Plecostomus commersonii affinis Steindachner. Localities: Mendez; Santa Clara; Rio Mucuri; Rio de Janeiro; Rio Doce at Linhares; Rio Parahyba; Minas Geraes; San Antonio River. 6b. Plecostomus commersonii scabriceps var. nov. Types, No. 7894. Three specimens. .30-.35m. Sao Matheos. Hartt & Copeland. h. Tip of snout naked; lat. 1. 28-30; occipital bone margined by a single large nuchal plate i. Spots on the caudal less distinct than those on the dorsal; all the scutes with high keels. j. Lat. 1.28. Eye 4} in snout, 8 in head, 3 in in- terorbital; outer caudal rays little produced, middle caudal rays 14-1} in the head. Base of the dorsal fin equal to its distance from the caudal, or very little shorter limosus nov. 4. jj. Viat.1. 30. Eye 3} in snout; 6-65 in head. carinatus 5. zi. Spots on the caudal as distinct as those on the dorsal. k. Lower surface of the head covered with scutes. 1. Spots all small, 12 or more on one of the anterior scutes commersoni 6. ll. Spots large, not more than 4 on one of the anterior scutes. commersonii affinis 6a. kk. Lower surface of the head naked, except a tri- angular spot in front of the gill opening; spots as in affinis; occipital process short. commersonii scabriceps nov. 6b. hh. Tip of snout granular, occasionally naked in ple- costomus. m. Lateral plates all more or less strongly keeled. plecostomus 7. mm. Lateral plates, scarcely, if at all, keeled. Lat. 1. 27-28. - n. Head with vermiculating light and dark lines; ventral surfaces with short curved bars. _ vaillanti 8. SOUTH AMERICAN NEMATOGNATHI.’ 169 7. Plecostomus plecostomus (Linnzus. ) Al. guacuri Lacépede; ZL. flava Shaw; H. veres Cuv. & Val.; Pl. bi- cirrhosus Gronow.; Pl. brasiliensis Bl. Localities: Silva, Lake Saraca; Para; Hyavary; Coary; Rio Puty. 8. Plecostomus vaillanti Steiudachner. Localities: Rio Puty; Rio Preto; San Gongallo. 9, Plecostomus villarsi Liitken. 10. 1a 12, 13, 14, Plecostomus virescens Cope. Plecostomus biseriatus Cope. Plecostomus seminudus nov. Type: A single specimen. Locality: Brazil? Plecostomus anne Steindachner. Plecostomus pantherinus (Kner. ) nn. Head with round spots; ventral surface plain. villarsii 9. lil. Ventral surface more or less naked. o. Lower surface of head naked. virescens 10. oo. A quadrate naked space between the ventrals. biseriatus 11. ooo. Belly with scutes on the sides, and sometimes a narrow strip between ven- trals. p. Lat. 1. 27. Upper lateral plates with strong median keel; head with strong occipital and weaker temporal keels; nuchal plates bicarinate. A band be- tween the pectorals, sides of the belly and a narrow median area between the ventrals covered with scutes. Dorsal surface, sides and fins spotted with brown; ventral surface plain. seminudus nov. 12. pp. Lat. 1. 30; lateral plates not keeled. anne 13. oooo. Belly entirely naked. pantherinus 14. dd. Head without distinct ridges or keels; snout broad, rounded. q. Belly covered with granular plates. r. Tip of snout naked. LR s. Dorsal crossed by six or seven black zig-zag stripes. cordove 15. 170 CALIFORNIA ACADEMY OF SCIENCES. 15. Plecostomus cordove Giinther. 16. Plecostomus lima Reinhardt MS. 17. Plecostomus macrops sp. nov. Type No. 7888. One specimen, .28 m. Rio das Velhas. Allen & St John. 18. Plecostomus francisci Liitken. 19. Plecostomus alatus (Castelnau.) ss. Dorsal spotted or uniform dusky. t. Superciliary edge not raised, interorbital convex; eye 35 in snout, 63 in head, 2 in interobital; post-temporal with a long descending process which forms the entire posterior margin of the orbit. lima 16. it. Superciliary edge greatly raised, interobital area con- cave; eye 3 in the snout, 54 in the head, 13 in the in- terorbital; post-frontals without a descending process; dorsal uniform dusky. macrops nov. 17. rr. Tip of snout granular; eye 63 in head. Srancisci 18. qq. Belly partially or wholly naked. u. Each of the larger lateral plates with two or three yel- low spots; eye 2 in the interorbital. alatus 19. uu. Sides of the body and the fins with longish yellowish spots; head as wide as long. aurogutatius 20. uuu. Sides of the body usually with round dusky spots; fins similarly spotted or plain. v. Occipital bone bordered by two or three nuchal plates. w. Caudal plain. litkeni 21. ww. Caudal with wavy cross bars formed by dusky spots. vermicularis nov. 22, www. Caudal with two series of spots between each two rays. brevicauda 23. vv. Occipital bordered by a single nuchal plate. xz. A single series of large round spots between two dorsal rays. robinii 24. ax. Two series of spots between two dorsal rays; upper caudal lobe little shorter than the lower. wuchereri 25. xxx. Spots on the caudal much smaller than those on the dorsal; caudal very obliquely truncate. johnit 26. 66. A.68. (WNeoplecostomus) oy. A. microps 27, yy. A.8& granosus 28, SOUTH AMERICAN NEMATOGNATHI. 171 20. Plecostomus auroguttatus (Kner.) HI. asperatus Castelnau. 21. Plecostomus liitkenii Steindachner. Localities: San Antonio River; Campos. 22. Plecostomus vermicularis sp. nov. Types, Nos. 7814; 7848; 7849; 7850; 7851; 7857. Thirty-seven speci- mens. .08—.30m. Rio Parahyba; Rio Janeiro; Mendez; Macacos; Goyaz. This species is closely related to Pl. lutkeni Steind. The ventral surface in the adult is almost entirely cov- ered with granules, except a large naked area at base of ventrals; the anus is surrounded with granules. In the young, the belly is almost wholly naked, there being only a few granules in front of the anus, on the sides of the belly and on the throat. The spots on the head frequently coalesce into vermicu- lations; all the fins are profusely spotted with dark, the spots on the dorsal, pectoral and ventral fins sometimes coalescing into cross bars, those on the caudal forming reg- ular transverse series. Upper surface of the body with obscure markings, ventral surface plain. Hye 3-4 in the snout, 53-8 in the head, 2-23 in the inter- orbital. Lat. 1. 26-27. Head 3-83. 23. Plecostomus brevicauda Giinther 24. Plecostomus robinii Cuv. & Val. Pl. une Steindachner. Locality: Rio Una. 25. Plecostomus wuchereri Giinther. Locality: Rio Pedra. 26. Plecostomus johnii Steindachner. Localities: Rio Preto; Rio Puty. NEOPLECOSTOMUS SUBGEN. NOV. Type Plecostomus microps Steindachner. 27. Plecostomus microps Steindachner. Localities: Juiz de Fora, Rio Parahyba; Goyaz. 28, Plecostomus granosus (Cuyv. & Val.) ? Plecostomus watwata Hancock. 172 CALIFORNIA ACADEMY OF SCIENCES. REMARK.—Since writing the above, we have studied the Siluride Opisthoptere of Ginther (= Trichomicteride). The genus Heptapterus cannot be retained in the 7 richomicteride, It is very closely related to Rhamdella, which should be raised to generic rank. In Heptapterus the modified ante- rior vertebre have simple lateral processes beneath which the air-bladder lies free in the abdominal cavity, as in the Siluride proper. In Trichomicteride the anterior vertebre are coalesced, but the lateral processes are greatly compli- cated, forming a closed capsule in which is concealed the pair of small round air-bladders. Leaving out of consider- ation the position of the dorsal in its relation to the ventral fins, Hepapterus and also the genus Nannoglanis Boulenger possess all the characters which define the group Pimelodine Ginther. That the relative position of the dorsal and ven- tral fins is of no great importance may be gathered from a comparison of the following plates: Liitken, Rio das Velhas Flodens Fiske, pl. III, figs. 6—7; Steindachner, Flussf. Siidam. ITI, pl. V, fig. 1, and Boulenger, Proc. Zoél. Soc., 1887, pl. X XI, fig. 3, representing the genera Pimelodella, Heptapterus and Nannoglanis. After Rhamdella should therefore be inserted - = Via. Hepraprervus Bleeker. 36a. Heptapterus mustelinus Valenciennes. - +, Locality: Maldonado. 36b. ‘Hepta terus surinamensis Bleekor. 36c. Heptapterus collettii Steindachner. VIb. “Naxnoctanis Boulenger. 36d. Nannoglanis fasciatus Boulenger. GEOLOGY OF BAJA CALIFORNIA. bia \ NOTES ON THE GEOLOGY OF BAJA CALIFORNIA, MEXICO. BY W. LINDGREN, U. 8. GEOL. SURVEY. [The notes here presented are based on observations taken during a few weeks’ visit in January, 1888, to the vicinity of Todos Santos Bay and the mountains east of it. They are necessarily somewhat fragmentary, espe- cially as inclement rainy weather and snow in the mountains greatly inter- fered with the work. This paper will shortly be followed by another devoted to the micropetrography of the region here described. ] But little is at present known of the geology of the Cali- fornian Peninsula or Baja California. Almost the only trustworthy source is Prof. Gabb’s description, published in the Appendix to Whitney’s ‘‘Geology of California,” Vol. I. This paper is founded on observations during a trip through the entire peninsula from La Paz to San Diego, undertaken in 1867, with Mr. J. Ross Browne, Sr., in order to examine certain land grants in the central part of the territory. Some notes relating to the peninsula are found in the re- ports of the Mexican Boundary Survey and in Prof. Blake’s geological notes accompanying the Pacific Railroad Reports. The only paper relating to the palaeontology of the territo- ry is Dr. C. A. White’s ‘‘ On New Cretaceous Fossils from California,” Bull. No. 22, U. S. Geol. Survey publications. GENERAL TOPOGRAPHY. The peninsula of Lower or Baja California is nearly 700 miles long, and on an average 60 miles wide; it extends from lat. 23° to 32° 30’. Generally speaking, it is occupied by a ‘“‘one-sided” mountain range, sloping gradually and gently towards the Pacific, abruptly towards the Gulf of California, or, in the north, to the low deserts adjoining the gulf. Following Prof. Gabb, we may, topographically, divide the territory into three parts: 2p Ser. Vou. I. Issued September 25, 1888. 174 CALIFORNIA ACADEMY OF SCIENCES. 1. The extreme south, from La Paz to Cape St. Lucas. This small division apparently differs in structure from the rest, and is said to be occupied by a transverse H.-W. chain, not exceeding 5,000 feet in elevation. 2. In the 300 miles of distance from La Paz to Santa Ger- trudis the divide of the peninsula is placed very near the eastern coast, and its elevation does not exceed 4,000 feet; frequently it is not more than 3,000 feet. The eastern slope is abrupt and steep, while the western is occupied by gently inclined and smooth table-lands or mesas, separated by nar- row, rocky cafions. 3. The northern division, from Santa Gertrudis to the boundary line, is more varied in its topographic features. The table-lands disappear, and a mountain chain rising in the middle of the peninsula fills with its branches the whole western half; the eastern chain becomes lower and soon sinks under the sands of the desert plains adjoining the eulf. Rising rapidly, the western chain reaches an elevation of 10,000 to 11,000 feet in the range of San Pedro de Martis, about 120 miles south of the boundary line. From here it sinks again, and runs with a maximum elevation of from 5,000 to 6,000 feet up to Alta California. In the section referred to and described in these notes, from Todos Santos Bay eastward to the mouth of the Colorado River, a distance of about 100 miles, the summit of the range runs at a distance of 60 miles from the Pacific, and divides the peninsula into two distinct parts— the Desert and the Pen- insular Cordillera. This division applies for a distance of at least 100 miles south of the boundary line. The topographic features of the long western slope of the Cordillera are somewhat diversified, but on the whole they may be divided into three sections: 1. The coast range, or the first orographic block, rising gradually from the sea to an elevation of 3,000 feet in a dis- GEOLOGY OF BAJA CALIFORNIA. V7s tance of 20 to 30 miles. Surmounting this are several mi- nor ranges and sharp peaks attaining an elevation of 3,000 to 4,000 feet. Consulting Plate II, it will be seen that on the whole these latter form an interrupted series running from the boundary to below Santo Tomas at a distance of 10 to 15 miles from the coast. A rapid descent leads from the divide of the coast range to— 2. The interior valleys, an interrupted series of depres- sions in the middle of the chain at an eleyation of from 1,800 to 2,000 feet. 3. The second orographic block, rising rapidly from the valleys and continuing as an almost level plateau with a gentle slope up to the peninsular divide, and the abrupt, almost precipitous descent to the desert. The elevation of this remarkable plateau is from 4,000 to 5,000 feet. The desert forms a sandy plain, traversed by a few lower ranges, running northwesterly and southeasterly, and has in general a very inconsiderable elevation. Part of this plain is doubtless depressed below the level of the sea. GENERAL GEOLOGY. According to Prof. Gabb, the characteristics of the three principal divisions of the peninsula are as follows: The ex- treme south is occupied by a granitic range with mica- schists exposed on the northern side. The second division is pre-eminently that of the mesa sandstones. For a distance of over 300 miles the largest part of the peninsula is formed by these. Heavy beds of generally non-fossiliferous, coarse sandstones slope with a gentle dip westward from the divide towards the Pacific. Towards the east the strata, during the rapid descent to the Gulf, as a rule, preserve the same dip. It appears as if the sandstones had been cut off by a gigantic fault. A section across the peninsula would be somewhat as shown in Fig. 1. 1:76 CALIFORNIA ACADEMY OF SCIENCES. esta soode: Gulf of California. et anaes Zi) ZZ : SS Yi ZZ LZ tj ne The age of these sandstones has not been determined by Prof. Gabb. He suggests that they may belong to the Miocene period from the fact that certain fossil Ostraea, indicative of that period, were once sent to the U.S. Na- tional Museum from the southern part of the peninsula. The sandstones often contain large pebbles of volcanic rocks, suggesting their being deposited near the shore line of some continent, possibly once situated where now the eulf extends. In some parts of the division extensive met- amorphism prevails, but the exact character of the resulting metamorphics is not clearly indicated; the metamorphic action seems to be confined to a belt near the summit and along the eastern shore, and rather appears to be a regional metamorphism than one connected with granite or eruptive rocks. In many places recent dikes and lava flows cut and cover the sandstones. An especially remarkable recent volcanic belt crosses the peninsula from San Ignacio to the voleano of Las Tres Virgines. Post-pliocene deposits of small depth cover the mesa sandstone to an elevation of 400 to 500 feet. Some distance north of Santa Gertrudis the mesa sand- stones cease to form the major part of the peninsula, but continue to skirt the shore for along distance. The western mountain chain, which now becomes dominant, appears to be mainly composed of granitic rocks, covered in places by recent or older eruptive rocks. This granitic chain con- tinues northward till it crosses the boundary line and joins the San Jacinto Mountains in San Diego County. Prof. Gabb’s notes from the northern section are not so complete as those from the southern and central part, and of the region to be described here he has but few words to say. It would be highly desirable to obtain a geological section of GEOLOGY. OF BAJA CALIFORNIA. ”) 177 the peninsula across the range of San Pedro de Martis, or about 150 miles south of the boundary line. The geology of the surroundings of San Diego and the section from there to the Colorado River is known princi- pally from the notes of Prof. Blake while attached to the Pacific Railroad Surveys. <A copy of this profile will be found in Plate IT, fig. 2. The coast at San Diego is covered by deep post-pliocene strata (sand, clay, etc.), and these de- posits extend for about twenty-five miles inland, forming a very gently sloping mesa, at the eastern edge of which gran- itic rocks make their appearance; in some places there seems to be a narrow belt of porphyritic eruptives at the western edge of the granite.* The now more rapid ascent leads for thirty miles exclusively over granite to the summit, from which there is an abrupt descent to the Colorado Desert, although not quite so marked as further south. The distance from the base of the mountains to the Colorado River is about eighty miles. From near the summit the whole east- ern slope, according to Prof. Blake, is made up of meta- morphic slates of the age of which nothing definite is known. Although the whole long western slope is indicated as granite in the reports mentioned, still it is certain that the granite area contains one, if not several, enclosed masses of metamorphic slates. So, for instance, at Julian, not very far north of the boundary line.t Recent erup- tives are not mentioned, and it would appear, indeed, that they are very scarce in the western part of Southern California; off the coast there is, however, a volcanic belt with basaltic lavas along the islands. t Going south from San Diego, the level post - plio- cene beds change into a hilly coast, the higher mount- ains extending close to the sea. About forty miles southward recent volcanic flows, coming down to the sea in * See Mex. Bound. Survey, Geol. Report. + Report of the State Mineralogist of Cal., 1886. t Whitney, Geology of California, Vol. 1. 178 CALIFORNIA ACADEMY OF SCIENCES. abruptly ending tables 400 to 600 feet high, form a rocky coast line, which continues down to Cape St. Miguel. (See Plate II.) Beyond the coast higher, flat-topped mountains, probably also of volcanic origin, rise. At the cape just mentioned, the great bay of Todos Santos, or All Saints, opens, forming a wide sweep and terminating southward at the rough promontory of Punta Banda. (See Plate IT.) Mountains 2,000 to 3,000 feet high form a sort of am- phitheater around the bay, and the foothills come down close to the shore. One may distinguish several sep- arate chains or masses, one running E.S. E.—W.N. W., forms Punta Banda, ending westward in a sharp peak 1,200 feet high, and culminating in the 3,500 feet high mountain between the Maneadero and Santo Tomas. Another mass, extending N.-S., but cut in two by a lower pass east of Ensenada, has its culminating pointin a peak about ten miles N-N.E. of Ensenada. The rivers carrying but little water and mostly sinking in the sand before reaching the sea, run in deep canons to a few miles distance from the coast, when they debouch in large, level, alluvial valleys, contain- ing some arable land; such are Ensenada Valley and El Maneadero. At other places, the foothills, distant a mile or two from the sea, are skirted by a band of mesa land rising gently or in terraces from the sea to an elevation of 50 feet; so, for instance, at San Carlos, between Ensenada and El Maneadero. The region adjoining Todos Santos Bay has been examined in a somewhat more detailed way, and a geological sketch map of it prepared. (See Plate IV.) The maps of the International Colonization Company of Ensenada and surroundings, extended by my own observa- tions, have been used as a topographic basis. The different terranes on this special map will now be discussed separately. GEOLOGY OF BAJA CALIFORNIA. 179 DETAILED GEOLOGY. Post-PLIocENE StrRATA.—Although these deposits do not form a prominent and conspicuous part of the topography, they may be found in several places along the shore of Todos Santos Bay. The numerous oscillations of the shore line during post-pliocene time are equally plain in Lower Cali- fornia as along the coast north of it. According to Prof. Gabb these strata cover the mesa sandstones for hundreds of miles to the south up to an elevation of 500 to 600 feet above the sea, and consist of loose sands with shells, calca- reous marls, etc. At Todos Santos Bay the deposits are not extensive, but in places the old shore lines, indicating sev- eral pauses in the movement, may be observed. At Ense- nada there are no preserved strata, but the low mesa at San Carlos belongs to this division. In the low hills skirting the Sausal Valley remains of these strata occur resting on porphy- ritic rocks. There are white, loose sandstones with remains of shells; further up, one mile north of Sausal, a cliff about thirty feet high shows a succession of sandstones and yellow clay with coaly vegetable remains. On such seams the report of the occurrence of coal beds between Todos Santos and San Diego appears to be founded. Better than at any other point the ancient shore lines are shown on Punta Banda, as often well marked wave-built terraces. Two such shore lines are plainly seen running along the promontory, but on closer investigation, at least four distinct ones may be found. The lowest and plainest about 40 feet above the sea; above that another, not so distinct, at 180 feet; further up, indistinct remains of beaches, at 470 and 600 feet. The accumulations along the lowest beach line consist of sands and gravel of porphyritic rocks, together with numerous fragments of shells. TERTIARY StRATA.— These seem to be entirely absent, or have at least not yet been found along the upper coast of Lower California, an absence made more conspicuous by 180 CALIFORNIA ACADEMY OF SCIENCES. the enormous development of miocene strata in Southern California. It is probable that this region from the pres- ent shore line to the desert was not submerged during terti- ary times; observations tending to confirm this view will be recorded further on.* Cretaceous Strata.—The first evidence of the existence of these in Lower California was furnished by a few fossils sent to the national museum by Mr. C. R. Orcutt of San Diego, who found them at the southern shore of Todos Santos Bay. Dr. C. A. White has mentioned and described these fos- sils,+ arriving at the interesting conclusion that the beds in which they occur belong to a formation which has hitherto not been recognized, and which probably occupies a posi- tion between the Chico and Shasta groups. They are equivalent with another series previously found by Dr. G. F. Becker} along the shores of Mendocino County, and consisting of sandstones and shales, resting unconform- ably on metamorphic rocks, presumably to be referred to the Shasta group. It has been proposed to call the new series the Wallala group, from a locality near the northern occurrence. The fauna described by Dr. White comprises one new genus and five new species. Specially characteris- tic of the formation is the new genus Coralliochama (Chamidee), closely allied to the European genera of Caprima, Plagioptychus and Ichthyosarcolithes; further, a few small species of Cerithiwm, Trochus and Nerita. Consulting the sketch map of Todos Santos Bay (Pl. IV), it may be seen . *Miocene fossils have recently been found at Boleo, opposite Guaymas, jn the southern part of the peninsula. They occur in a series of gravels, clay and marls, interstratified with andesitic volcanic material. See E. Fuchs, Sur le gite de cuivre de Boléo. Bull. Soc. Géol. de France, XIV, 2, 79. 1886. | ~ + Bull. 22, U.S. G.S. On new Cretaceous fossils from California. Bull. 15, U.S. G.S. On the Mesozoic and Cenozoic Palzontology of California. ~ {Bull. No. 19, U. S. Geol. Survey. Notes on the stratigraphy of California. GEOLOGY OF BAJA CALIFORNIA. 181 that the exposures are not very extensive; they are, in fact, limited to a patch about three miles long east and west, and one mile north and south, on the northern side of Punta Banda. This promontory consists mainly of old porphyritic rocks, against which the cretaceous strata have been de- posited (See profile 2, Plate V); they form a perpendicular sea cliff about fifty feet high, and extending for about two and a half miles along the shore; in most places post-plo- cene gravel and sand cover the top of the cliff, but cretace- ous exposures may be found a little higher up, protruding from the later, covering beds. The strata consist of a series of yellow, coarse sand- stones, interstratified with heavy conglomerates of well rolled pebbles up to a foot in diameter, evidently petrographically identical with the massive rocks forming the peninsula and derived from them. The beds being very heavy, exact dips and strikes are difficult to obtain; they incline somewhat northward, dips from 10° to 20° being observed. In these heavy sandstones, generally void of fossils, there are at intervals strata two to four feet thick composed en- tirely of the calcareous shells of Coralliochama, accompanied by asmall specimen of Cerithium. The fauna is decidedly monotonous, and but a few new forms were found. There are strong reasons for believing that this isolated cretaceous mass is but the first of a series of similar strata farther south, and I only regret that I could not extend my trip in that direction. On Todos Santos Island, about twelve miles from the coast, there appears to be an exposure of sandstone similar to that of Punta Banda. At the har- bor of Santo Tomas, 40 miles south of Ensenada, a steep sand- stone cliff, containing many and large pebbles, is said to form the coast. Similar sandstones occur at Cape Colnett, about 70 miles south of Ensenada. From here it is not very far to where the mesa sandstones begin, and it appears to me quite probable that these also will eventually be found to belong to the cretaceous period. Gabb has followed tlie 182 CALIFORNIA ACADEMY OF SCIENCES. sandstones more carefully on the eastern than on the west- ern coast, and states expressly, regarding the former, that they continue to skirt the coast for a long distance north of the end of the second or middle division, where they occupy the whole peninsula. His description of the appearance of the sandstone corresponds very well indeed with the Todos Santos occurrence. It must be remembered that Gabb did not determine the age of the mesa sandstones, but only conjectured, based on a very slight clew, that they were Miocene. At the same time he describes the quite extensive metamorphism which has affected the strata in certain regions. This fact can not well be brought into ac- cord with such a comparatively recent age as that to which he assigns the rocks. At least nothing has been observed in Upper California, indicating that any extensive meta- morphism has occurred so recently.* Definite paleontological evidence, establishing the age of the mesa sandstones, would be very desirable, and help to clear many doubtful points in the geology of the penin- sula. PRE-CRETACEOUS Eruprives.—Gabb has already remarked that eruptive masses occupy much room in the northern part of the territory, and that they overlie the granite, but from his description it is not always apparent whether these eruptives are recent or older. From the southern end of Todos Santos Bay to the boundary, they occupy large areas, but are—at least in the section from Ensenada eastward— confined to the first coast range, not extending more than twenty miles west of the shore. On the sketch map massive eruptives occupy a large area, *The character of the miocene series recently found in the southern part of the peninsula at Boleo (see ante p. 180, foot-note), appears to be very dif- ferent from the mesa sandstones and from the Todos Santos occurrence; it also encloses volcanic materials of andesitic and basaltic character. It is probable that the shells referred to above (ostraea) came from a series equiv- alent to this and not from the mesa sandstone. GEOLOGY OF BAJA CALIFORNIA. 183 and form the rough ranges and peaks of the coast range al- ready mentioned in the introduction. So, for instance, north of Ensenada, where they reach quite down to the sea, and con- tinue towards the high peaks eastward; the hills east of San Carlos are also composed of eruptives; they become still more conspicuous in the high and abrupt range, south of the Maneadero, ending in Punta Banda and extending southward to Santo Tomas. These eruptive rocks have, on the whole, a porphyritic habit, and possess the appearance of considerable geologi- cal age; they are usually dense and hard, without accom- panying tufaceous masses, and contain a number of second- ary minerals, especially epidote; it is further evident that they have suffered considerable erosion. The petrographic character is rather varying and will be described more minutely in another paper. At Ensenada Point the rock has a dense, almost flinty ground mass, with small feldspar crystals; between Ensenada and Sausal gray or reddish porphyrites prevail; red and black porphyrites occur in the mountains N.N. EH. of Ensenada. East of San Carlos light yellowish-red quartziferous porphyries or porphyrites prevail, while normal brownish quartz porphyries may be found in the hills north of Sausal. The Punta Banda range is composed of at least two or three different eruptives: diorites of granular structure, hornblende porphyrites with large hornblende crystals, and finally a dense reddish porphyrite. Flow structure on a large scale is often apparent in these rocks; so, for instance, along the coast from Ensenada to Sausal, where a gray porphyrite flow about one hundred feet thick is seen to have covered a similar eruptive of a reddish color. In the hills east of San Carlos beautiful examples of this structure, on a large scale, may also be seen. The age of these massive rocks is quite satisfactorily determined as pre-cretaceous from the above mentioned 2p SER. VoL. I. (13) Issued September 25, 1888. 184 CALIFORNIA ACADEMY OF SCIENCES. cretaceous sandstone, which contains numerous pebbles of the same porphyrites. On the other hand, they are most likely younger than the highly compressed metamorphic slates of uncertain age in the interior. For the present the origin and source of these enormous flows of mesozic eruptives must remain uncertain. Along the section from Ensenada eastward there are no eruptives to be seen, and it is probable that they were poured out over the granite from vents near the coast. The presence of these effusive rocks in so large quanti- ties forms an important feature in the geology of the penin- sula, the more so as they appear to be absent, or at least little developed, in Alta California. Recent Ervprives.—At Sausal Beach a great number of pebbles of a black vesicular basaltic rock are found, evident- ly derived from the large recent volcanic flows, beginning at Cape San Miguel and extending for about twenty miles up the coast. The exact age of the flows cannot as yet be deter- mined, but judging from structure and form it must be quite recent There are no other neo-volcanic rocks in the limits of the sketch map, and probably none for a long distance south of the bay. One hundred miles further down the coast there is a small voleanic mountain at the bay of San Quentin; from there southward, the recent eruptives are apparently more common. Section From Topos Santos Bay HE. N. EH. To THE DEs- ERT AND THE CoLorapo River.—(See Plates II and III.) The profile represented in fig. 1, Plate II], begins at En- senada about fifty miles south of the boundary line and extends to the Colorado River. My own observations did not extend any farther than to the desert, and the eastern part of the section is drawn from the comprehensive view over the desert gained from the last escarpment of the mountains, aided by some notes of Professor Blake in the Pacific Railroad Reports, relating to the same territory. GEOLOGY OF BAJA CALIFORNIA. 185 It is seen at once that the section is in some respects quite similar to Blake’s above mentioned profile north of the boundary line, and that it is almost exclusively made up of granite and metamorphic slates, the former much predominating. THE FIRST OROGRAPHIC BLOCK OR THE COAST RANGE. Leaving the small quaternary valley of Ensenada, the sec- tion runs, gently ascending, on the divide between Ensena- da and Gallo Creek; to the right and Jeft rugged masses of porphyrites rise above the granite floor. Ata distance of ten miles from Ensenada the road descends in the Gallo Creek and continues along it up to the first divide. The granite is a hornblende granitite of coarse grain and white color, weathering into rounded blocks, and finally giving a sandy, barren soil. It is in composition and texture equivalent to the granite of the Sierra Nevada. The char- acter appears to remain very constant. Twelve miles from Ensenada, in the Gallo Canon, meta- morphic rocks are met with, extending for five miles, up to within two miles of the first divide. In passing thé contact, one gains the decided impression that the metamorphic overlies the granite; the former con- tinues up on the hills above the canon, while the bottom is yet in granite. This metamorphic series is at first glance not easy to recognize as such. Passing through the caiion rapidly one would be tempted to refer it to massive rocks. The first few miles run over chloritic and micaceous slates, mixed with other very indistinctly stratified metamorphic rocks. The chloritic schist runs N.W.-S.E. to N.-S. and stands nearly vertical. Farther on the stratification ceases entirely, and brown, irregular outcrops of dense to middle -grained dark-green dioritic and diabasic rocks prevail. Approaching the eastern edge of the area this habit becomes more apparent, and the rocks ad- 186 CALIFORNIA ACADEMY OF SCIENCES. joining the granite are coarse- to middle-grained diaba- ses. Before reaching the first divide granite appears again, but the metamorphic is seen to continue, covering the granite for some little distance north of the road. How far this metamorphic belt reaches is not certain, but the pro- bability is that it continues for quite a long distance south- ward. No quartz veins are seen in this zone, nor have any placers been reported from this neighborhood. Granite forms the summit of the first range, and rises in naked cliffs a few hundred feet above the pass (2,500 feet) to the north and south of the road; it continues forming the divide, northward, at least till opposite Real del Cas- tillo. From the summit of the pass the view is extensive and beautiful. Six or seven hundred feet below lies the largest of the interior valleys, Valle de San Rafael, a basin- like depression abont fifteen miles long north to south, and ten miles from east to west. On all sides it is surrounded by mountains, mostly bare, light-colored granitic ranges; the highest range, forming the main divide between the Pacific and the Gulf, rises directly eastward, and its summit forms a gently undulating sky line far away. The irregular series of depressions to which the name of ‘‘ the interior valleys” has been given, runs approximately north and south, and at a distance of about twenty or thirty miles from the coast. Northward it is represented by Valle de las Palmas and Guadaloupe; southward it continues in the valleys of Santa Clara, Santa Catarina, and Santa Cata- lina; all these are separated by more or less prominent trans- verse ranges, and are situated at somewhat differing eleva- tions, from 1,500 to 3,000 feet, but that they should be re- garded, all together, as a result of the same or similar geo- logical causes appears very probable. San Rafael Valley is topographically somewhat peculiar, as it is formed by two rivers, the San Carlos and the Guada- loupe, the former flowing through the southern, the latter GEOLOGY OF BAJA CALIFORNIA. 187 through the northern part of the valley, as a glance at the map (Plate IL) will show. The two basins are separated by a very low range of hills, dividing the valley in two sec- tions. Emerging from deep canons in the eastern range the rivers pursue a winding course through the valley to break through the western or coast range in equally deep and rocky gorges. To the east and west the valley is bordered by steep mountain barriers. An especially steep granite es- carpment runs along the western side from Real del Castillo southward. (See Plate II.) The road from Ensenada to the Real, which strikes the valley at its central east and west line, descends rapidly from the summit for the first few hun- dred feet over granite; the remaining two miles, till the level of the valley is reached, are occupied by dark, massive, more or less fine-grained dioritic rocks, which in all probability should be referred to the metamorphic series. As soon as we reach this last-mentioned terrane very nu- merous quartz veins, formerly absent or scarce, are noticed. When the valley is reached the road bends northward to- wards Real del Castillo, and runs over the quaternary accu- mulations of debris forming the surface of the valley. The low range of hills separating the northern from the southern part of the valley, consists partly of granite, partly of me- tamorphie slates. The little Mexican mining town of Real del Castillo lies on the Guadaloupe River at the point where it leaves the valley, and enters the rocky canons of the coast range. Some placer and even quartz gold-mining has been carried on here by the Mexicans for about twenty years. Fig. 1, Plate V, represents a profile across the northern end of the valley at the Real. It will be seen that immedi- ately below the steep granite escarpment to the west of the town metamorphic rocks begin, at first massive and dioritic, then decidedly slaty and composed mainly of chloritic schists, continuing for three miles eastward across the first low hills north of the valley. The strike is generally N.-S. 188 CALIFORNIA ACADEMY OF SCIENCES. or N.N. W. and 8.8. E., and the dip very steep, somewhat to the east; in comparatively few places, however, can either be well observed. At the above-mentioned distance from Real there is a hot spring, with a temperature of about 100° F., containing very little mineral salt but strongly impregnated with sulphuret- ted hydrogen. Hastward the hills of the easterly range rise, and on their slope the metamorphic series is traversed by a great number of dikes of granite, porphyry or similar rocks. This produces a peculiar effect, as the white dikes are visi- ble from a long distance on the hillsides, running parallel or sometimes intersecting each other like the meshes in a network. A short distance beyond this, granite appears to begin again. The whole slate series is filled with quartz veins, some of which contain gold; the larger of these veins at Real run about east and west. The granite is usually barren of gold. It is not known how far northward this slate area of Real del Castillo extends, but it is probable that it is entirely surrounded by granite, as extensive gran- itic areas are seen in the mountains to the north. There are several of these isolated slate areas in the northern part of the peninsula and in San Diego County. So, for instance, about fifteen miles east of Real, at the mining camp of Sara- goza, according to statements of reliable informants at Real. However, one cannot fail to be impressed by the enor- mous extent of the granite and the small areas occupied by metamorphic rocks. It seems evident that the slates are of but little depth and everywhere are resting, as detached fragments, one might say swimming, on the granite. Of the age of the metamorphic series, as found in or near San Rafael valley, nothing more definite can be said than that they are precretaceous, or at least older than the unaltered and nearly undisturbed rocks of the Wallala Group. On account of the extensive metamorphism it is not probable that fossils will ever be found in them. I do not, however, consider it improbable that they may be of the same age GEOLOGY OF BAJA CALIFORNIA. 189 and equivalent to the metamorphic slates of the gold-belt of the slope of the Sierra Nevada. THE SECOND OROGRAPHIC BLOCK, OR THE SIERRA MADRE. The profile (Plate IIT, fig. 1) crosses the alluvial deposits and small slate and granite knolls of San Rafael valley and continues over the second orographic block, the Sierra Madre, as it is called. Bare and white granite hills rise rather abruptly from the covering débris at their bases and the road winds upwards for five miles at a rather steep erade. The granite, of which enormous areas are seen to the north and the south, is identical in structure and composition with that of the coast range—it is a horn- blende-granitite of ight color and easily disintegrating. Arrived at an elevation of 3,500 feet and five miles from the valley, one is surprised to find oneself on a gently as- cending plateau, somewhat rolling and with small knolls here and there, but on the whole remarkably level; the eastern sky-line is formed by the plateau; north and south it extends for miles and miles covered with dense brush and yucca trees. At nine miles from the valley and 4,000 feet elevation, a small knoll about 100 feet high hes to the south of the road; it is entirely composed of brownish quartzite completely enclosed by the granite, a curious re- mainder probably of a formerly more extensive metamor- phic area. The plateau becomes still more level and for a distance of about fifteen miles up to the summit, ascends but 700 feet. Beginning at 4,000 feet and extending to the divide, a belt of scattered pines relieves the monotony of the landscape. Here and there are low ranges of hills or small isolated peaks with the yellowish-white granite weath- ering in rounded forms. Such a range, only a few hundred feet high above the plateau and interrupted by passes at the level of the latter, forms the peninsular divide at an ele- vation of about 5,000 feet. The road from Real del Castillo to Campo Nacional, a small mining camp near the summit, 190 CALIFORNIA ACADEMY OF SCIENCES. descends a little, ten miles from San Rafael valley, into the wide depression in the plateau in which the San Carlos river flows, and follows this up to the summit, a distance of about nine miles. The river here runs but little lower than the general surface of the plateau, which necessitates very rapid fall in the lower part of the course near San Rafael valley. The upper part for nine miles from the summit has a very inconsiderable fall; the water-course runs through a series of swamps and little lakes, and heads in a somewhat larger lake. There are no signs of glacial action. Itis not possible to see just where the divide is, so level is the gen- eral surface of the country near the summit. For a few miles eastward this character is preserved and then signs of a more active erosion begin to appear. The pine forest of the plateau disappears, the water-courses are worn down into gorges and canons walled in by white, abrupt, granitic ridges; the country slopes rapidly. It is a wild landscape of bare granite hills. Finally, climbing one of the ridges, the spectator suddenly and unexpectedly finds a view ex- tended before him which in grandeur and sublimity is sur- passed by but few places on the continent. He stands at the edge of a gigantic escarpment, descending about 3,000 feet in about five miles; naked granite cliffs, separated by steep ravines and afew canons more deeply cut into the rock form the face of this escarpment; at its base the Colo- rado Desert spreads out, a dazzling white plain at the end of which, fifty miles away, gleams the tide water of the Colorado River at the Gulf of California. Again, far be- yond this, may be seen the blue mountains of Arizona and Sonora. The Colorado River, above the debouchure, is cut off from view by a lower desert range, continuing N.N.W. and run- ning east of the great Salt Lake Magnata, a blue sheet of water on the white desert plain. The great escarpment, facing the desert, continues north- ward and southward; its general trend, however, is some- GEOLOGY OF BAJA CALIFORNIA. 191 what west of north. Far to the south it is seen capped with flat-topped tables. As far as can be seen the granite extends; it forms the whole descent except near the desert where a narrow belt of a reddish rock is seen, connecting southward with dark. basaltic tables covering the desert at the foot of the escarpment. Many little ridges and isolated, low volcanic hills dot the desert in a southeasterly direc- tion. if there is any metamorphic slate at all at the base of the range, it can be but a very narrow belt. The extent of the remarkable plateau forming the summit of the range, must be very great. To judge from statements of reliable persons this enormous granite mesa extends from near the boundary to a point fifty miles south of the profile near the old mission of Santa Catarina. At this place, Mr. R. Stevens, who has examined that country for a proposed railroad, informs me the gently sloping mesa, composed of eranite, reaches 5,000 feet at the summit, while the passes are eroded deeper, and are at an elevation of about 4,000 feet. An abrupt escarpment, equally steep as at the place crossed by the profile, leads down to the desert plain, here some- what higher, having at the base of the clitf an eleva- tion of 2,000 feet. South of Santa Catarina, a deep valley, ‘‘Valle de la Trinidad,” has cut far into the mesa; south of this the topography and geology are but very little known. We would thus have a continuous predominantly granitic plateau bordering on the desert, twenty to forty miles broad and extending about 80 miles southward from the boundary. The descent is more or less gradual on the western side, while towards the east there is an exceedingly steep slope to the desert. On no map, as far as I am aware, has this remarkable topographic feature been properly or even approximately indicated. On the general map (Plate II) an attempt has been made to represent the great eastern es- carpment. The map is based on the larger map of the coast line of Baja-California, as surveyed by the ‘‘ Narragansett’ in 1873; the interior topography is derived from my own 192 CALIFORNIA ACADEMY OF SCIENCES. observations and some data taken from a Mexican grant map of the ‘‘Rancho Santa Catarina.” It is more of a sketch than an accurate map, but it is at least approximately correct. The granite of the plateau is petrographically very uni- form. It is throughout a coarse-grained hornblende gran- itite, or a hornblende-biotite granite, with but small local variations. Before closing the description of the profile a few obser- vations must be recorded which throw an additional light on some geological problems connected with the plateau. Near the summit, about four miles east of the dividing range of hills, lies at an elevation of 4,750 feet, the little mining camp, Campo Nacional; it is directly on the line of the protile. It is situated in the pine forest of the level pla- teau,although in its immediate vicinity the canons and ravines of the eastern slope begin to cut into the granite. Stand- ing on a small elevation nothing but granite can be seen as far as the eye reaches, in ali directions. Within a radius of a few miles the gulches leading down to the desert have been worked and yielded a considerable quantity of gold. The latter is coarse and well worn, and the gulches are filled with well rounded smooth pebbles of white quartz, or a dark quartzite. This, in itself, is remarkable, as there certainly are no met- amorphic rocks anywhere in the vicinity. It was soon found that all these gulches led up toa small flat-topped hill about 200 feet above the plateau, called the Black Hill. This hill is about one-half mile long, east and west, and one-eighth to one-quarter mile wide; it is made up of a well-packed mass of auriferous metamorphic gravel in very smooth boulders, often six inches in diameter. Its depth is uncertain; I was told that once a shaft had been sunk 18 feet without find- ing bedrock, and I should think the mass would be about 50 feet thick. The mining has been entirely confined to the ravines leading down from the hill. GEOLOGY OF BAJA CALIFORNIA. 193 That this patch of auriferous gravel has been formed by an ancient river of considerable importance is certain, but the most interesting questions are, whence did it come and to where did it flow, and where are the metamorphic rocks that furnished the material for the boulders? I saw no granite pebbles; if there are any they are very few or have been destroyed by weathering. It will be observed that the eranite disintegrates very easily, and that the present rivers running through San Rafael valley carry mostly sand and but little pebbles. It was not possible to determine the direction of the fall as the bed rock is not exposed, but it is probable that the general course of the river was east and west. If so, where did it head if flowing west, or where did it go to if flowing east? The great escarpment of the range is but few miles distant. I was informed that at Camp Juarez, about 20 miles north of Campo Nacional, there are similar deposits; the gold from the two places is certainly identical. I was also told that still further north, but further down the west slope there is a similar hill near Vallecitos. Finally Mr. R. Stevens informs me that the plateau near Santa Catarina is capped with a reddish ‘‘cement,” which may be a volcanic formation or a gravel deposit. Nothing definite can be said of the age of these auriferous gravels except that they certainly date from a time when the drainage and the whole topography of the peninsula were very different from what they are now. They may be tertiary and equivalent to the auriferous gravels of the Sierra Nevada. At any rate, they show conclusively that the present plateau is part of a very old continental area. Of the desert and the desert ranges near the Colorado river I have but little to say as I did not visit them. I have sketched in the great Salt Lake and the ranges beyond, from the view obtained from the great escarpment. These mountains of the desert are about 2,000 feet high, and apparently predominantly composed of granite on 194 CALIFORNIA ACADEMY OF SCIENCES. which in places black, evidently recent, volcanic rocks rest. The range bordering the Salt Lake is probably entirely volcanic. The desert between the foot of the escarpment and the range just mentioned above, is 20 miles wide and slopes gently from both sides to a maximum depression in the middle. The elevation of this desert at the foot of the Sierra is probably about 1,000 feet, and the center lies pre- sumably considerably below the level of the sea. The same depression connects northward with the New River Basin, and at its northern end Prof. Blake in the Pacific Rail- road Reports estimates it to be 500 feet below the surface of the sea. Southward it connects with the delta of the Col- orado River bya gap in the desert range, a little south of east from Campo Nacional. Near that point it must, of course, be higher, or the river would flood the desert. South of this gap the desert ranges continue, and the ayver- age elevation of the desert itself increases. CONCLUSIONS. Prof. Blake’s section from San Diego to the desert, re- ferred to above (see ante p. 177), presents certain similarities and certain differences when compared with the one de- scribed in these notes. Both sections have a gentle west and a steep east slope, although this peculiarity is much more accentuated in the southern profile. In both, the principal mass is composed of an enormous eranitic plateau with minor areas of highly metamorphosed and compressed slates. Im the former these are main- ly confined to the eastern, while in the latter they are intercalated in the western slope. The large masses of mesozoic eruptives at the Pacific Coast are not represented in the northern section. The granite composing the main area appears everywhere to be identical—a white horn- blende granitite similar to that of the Sierra Nevada of Cal- , ifornia. GEOLOGY OF BAJA CALIFORNIA. 195 Younger eruptives are absent in the northern section un- til the Colorado River is nearly reached, while near the southern section they occupy considerable areas at the foot of the great escarpment. It is most probable that the present metamorphic areas are but small remnants compared with those that once coy- ered the granite; the exposures along the section, such as for instance the small quartzite knoll on the plateau, or the metamorphic gravel at Campo Nacional, go far to prove that the slates everywhere rest as shallow crusts on the granite. Considering the structural relations one cannot but be- lieve that the great eastern escarpment is formed by an enormous fault, along which a throw of several thousand feet has taken place, and the fresh character of the fracture goes to prove that the dislocation is comparatively recent. The occurrence of auriferous river gravel, indicating that a stream of no small dimensions once flowed near the very summit of the range, speaks very strongly for the ex- istence of such a fault; indeed it almost proves it. A part of the course of this old channel must evidently have been cut off by the fault. Regarding the San Rafael valley, it should be noticed that it is bordered both east and west by a rather steep escarpment, and that a line carried from the summit of the first range up to the beginning of the plateau of the Sierra Madre, would give a nearly unbroken slope for the whole peninsular range. It would, therefore, seem very plausible to consider the valley as a sunken area between two north and south faults (Graben Suess). The fact that in the sec- tion, metamorphic rocks are found near the summit of the first range, and then again at the level of the valley at the foot of the steep descent, speaks strongly in favor of this suggestion. At last, considering the structure of the whole range, one cannot fail to be impressed with its great dif- ference from the Coast Ranges of California and its 196 CALIFORNIA ACADEMY OF SCIENCES. ereat general similarity with the Sierra Nevada. There is the same gentle western and steep eastern slope—there is the same mainly granitic mass with smaller areas of aurifer- ous metamorphic slate; there are finally the auriferous gravels and the recent volcanic outbursts at the eastern foot. When, going one step further, we consider that the eastern fault of the Peninsular Sierra, lies nearly in the direct S.S. E. continuation of the eastern fault of the Sierra Nevada, the intimate connection of the two ranges becomes still more apparent. NEW SUBSPECIES OF SONG SPARROW. 197 DESCRIPTION OF A NEW SUBSPECIES OF SONG SPARROW FROM LOWER CALIFORNIA, MEXICO, BY WALTER E. BRYANT. Melospiza fasciata rivularis subsp. nov. Brown’s Sone SPAarrow.* Subsp. char. —Similar to Melospiza fasciata fallax, but darker and averaging larger, with longer and less robust bill. Adult in somewhat worn breeding plumage. (Type No. 3,024, collection of Walter E. Bryant from Comondu, Lower California, March 11th, 1888.) Above, dark burnt umber more pronounced on crown and postocular stripes. Medi- an line of crown and sides of neck, dark ashy gray. Super- ciliary stripe, ashy, lighter anteriorly. Edging of dorsal feathers, sides and flanks, grayish brown; dorsal feathers narrowly black centrally. Upper surface of tail, dark um- ber, the central ones darker along the shaft. Tertiaries, darker than tail (above) with ashy edgings. Greater wing coverts reddish brown on outer half; the inner half dark as tertiaries; some of the coverts show a blackish subterminal mark along the shaft. Under parts, grayish white. Breast, sides and crissum, heavily streaked with reddish umber. Wing, 71 mm.; tail feathers, 75 mm.; bill from nostril, 10 mm.; tarsus, 22 mm.; middle toe, 18.5 mm. Adult? in somewhat worn breeding plumage. (Type No. 3,041, collection of W. E. B. from Comondu, March 12th, 1888.) Not readily distinguished from male in color- ation, but averaging smaller in size. Wing, 64.5 mm.; tail * Dedicated to Mr. Herbert Brown of Tucson, Arizona. 2p SER. VOL. I. Issued September 29, 1888. 198 CALIFORNIA ACADEMY OF SCIENCES. feathers, 68 mm.; bill from nostril, 10 mm.; tarsus, 21 mm.; middle toe, 17 mm. In former lists of Lower Californian birds no mention is made of finding any variety of song sparrow on the peninsu- la. Therefore, when I first found these birds, tolerably common along the water-course of Comondu canon, more than a passing interest was taken in them, and a series of thirteen specimens was collected. These were found to be uniformly different from a pair of I. f. fallax, which I had received some time previously from Mr. Herbert Brown, and to him I sent a typical example of the Lower Califor- nian bird for comparison with additional specimens of fallax. He found it to vary from anything which he had, and kindly sent mea series of fifteen song sparrows, col- lected at Tucson, Arizona. Ten of these are unmistakably fallax; the others are referable to iontana, although prob- ably not typical of that form, no specimen of which is at hand. Melospiza fasciata rivularis seems to be intermediate be- tween montana and fallax, approaching more closely the latter in coloration. The differential characters of these two extremes, as given by Mr. Henshaw,* are as follows: Melospiza fasciata montana. — Above, umber brown, with margins of feathers gray, giving a strong grayish aspect to the plumage; back streaked with blackish brown. M. fasciata fallaz.—Above, chiefly bright reddish brown; back streaked with a darker shade of the same; streaks be- low, reddish, not black; size, smaller. The Lower Californian sparrows are not in as fresh plu- mage as I could wish, but the differences are sufficiently defined to be noticed ata glance. Whether the modified shape and greater average length of bill will prove constant in a larger series of this new form remains to be determined. Nothing peculiar in the habits or song of this bird was *Auk, I, 224. July, 1884. NEW SUBSPECIES OF SONG SPARROW. 199 noticed that differed from other west coast song sparrows. They were not found at any locality except Comondu, which is well watered and admirably suited to their needs. Asearly as the 12th of March, the females had commenced to set, but no nests were found. A table of the measurements taken is here appended. But little dependence can be placed upon the tail measure- ments of either form, nearly all being in somewhat worn condition at the tips. DIMENSIONS OF MELOSPIZA FASCIATA RIVULARIS, FROM COMONDU, Saree 9 = haga ae ea Ml 5 = 4 + ) = = a = B 5 yi ~ oa as <1 =] = 2 sf, 5 oO 3 77) oO i ° e-) : 2 = : aR x a 2 Be, tee ce 3 : n : ® 3 ® a a 2 : : Z mm mm mm mm mm 3024 4 March 11, <-.-./\) 7.5 75. 10. 22. 18.5 Type. 3045 3 Gu als} 68.5 70 5 9.5 22 18.5 3052 4 “Oy a3 71 72). 9.5 23 19 3058 é COP aba 72 73.5 10 23 17.5 3108 4 SODA teenies 71 72 9.5 21 18 3157 4 EGON “SK sactaal) We Mileseanpos Olas 22. 17.5 3158 3 ce Me venooe 72. 74, 10. 21.5 18. Typical. 3205 4 ce AG ae eas 71 74, 9.5 22 19 Average.. ..- 70.9 73 9.7 22 18.2 3029 (e) March 12....... 67. 67 9.5 21 17 3041 2 SNL Os crcvarey ste 64.5 68. 10. 22. ilies Type. 3042 fe) HT a D.Ce ero 68. Kite 9.5 21. Ries Typical. 3156 Q Atprills 10°. 15<101- 64. 66 95 21 18 3176 fe) ne De cresters 67 73 9.5 20 17.5 | Average...... 66.1 69 9.6 21 (83 (14) 200 “ON S,10}09TL09 DIMENSIONS OF MELOSPIZA FASCIATA FALLAX, FROM TUCSON. - xag +O) +O) +0) CALIFORNIA ACADEMY OF SCIENCES. HERBERT BRowN, Collector. = eo = ee a = E ; : 2 S 2 ® = : : = 5 S : : Fs 8 : E 5 st : = mm. mm. mm mm. mm, Jan, 17,1886....| 68. 13.5 3. 20. 16. Feb. 22, 1886....] 65. 69. 8.5 20. 15.5 cg CS 68. 72). 8.5 1. 15.5 Apr.11,1886....] 65. 68.5 8.5 21.5 15. “© 18,1886....| 67. 69. 9. 20.5 15. BOS CRE A 66. 70.5 9. a1 16. May 2, 1886....| 66. 73.5 9. 21, 15.5 Average...... 66.4 70.9 8.6 20.7 15.5 Jan. 2, 1887....| 60.! 66. 8. 20. 16. «O5, « ve 64. 71.5 8.5 215 16. Cie SC) IGG 1 64. 66. 8. 20. 15.5 Apr.18,1886....| 64 70. 9. 20.5 15.5 Average...... 63 68 4 84 20.5 15.7 FLORA OF THE SANTA BARBARA ISLANDS. 201 FLORA OF THE SANTA BARBARA ISLANDS. BY T. 8. BRANDEGEE. I. ADDITIONS TO THE FLORA OF SANTA CRUZ ISLAND. The following list contains the names of plants of Santa Cruz not specifically mentioned in Prof. E. L. Greene’s Catalogue of the Flowering Plants and Ferns of the Island of Santa Cruz, published in Bulletin No. IT of the Proceed- ings of the California Academy. A few species of Prof. E. L. Greene’s Catalogue having been collected late in the season could not be determined specifically and are un- doubtedly in the present list. The collection upon which this list is based, was made between the twenty-sixth of March and the first of May, at a time when the evanescent annuals and early flowering perennials were in full bloom. Most of the shrubs of the list are not common .on the island, and many of them were seen in but one locality or ina single canon. The Quercus lobata was small, so that Populus Fremonti is the only tree to be added to the flora, and only a half dozen specimens of it are growing ina southwest canon. Continued search will add many species to the flora of this island, and not until every cafon has been thoroughly explored, can a complete list of its plants be given. I amvery much indebted to Mr. Justinian Caire for permission to visit the island, and for hospitality while making the collection. BERBERIS PINNATA, Lagasca. DeELpHINniuM Parryl!. Gray, Bot. Gazette, XII, 50. ERYSIMUM ASPERUM, DC. Not common. BRASSICA CAMPESTRIS, L. 2p SER. VOL. I. Issued October 11, 1888. 202 CALIFORNIA ACADEMY OF SCIENCES. ARABIS ARCUATA, Gray. ARABIS PERFOLIATA, Lam. SILENE CONOIDEA, L. Determined by Dr, Sereno Watson, who says it has been found in the southern part of the State. It grows in the Santa Inez mountains, and is also found near Point Sur, below Monterey. Probably introduced from Europe, al- though it seems to inhabit the mountains, and not to be found about fields and dwellings. ARENARIA Dovauasti, Torr. & Gray. CALANDRINIA BREWERI, Watson. CALANDRINIA MARITIMA, Nutt. SIDALCEA MALVZFLORA, Gray. MALVASTRUM EXILE, Gray. Decumbent, or in exposed situations, smaller and erect. ERODIUM MACROPHYLLUM, Hook. & Arn. GERANIUM CAROLINIANUM, L. OxaLis WRIGHTH, Gray. This species czspitose from a stout woody caudex, is common about Santa Barbara and along the Coast Range as far north as to San Francisco. The branches often root at the nodes, especially in damp locations and more north- ern habitats. ASTRAGALUS NIGRESCENS, Nutt. HOSACKIA GRANDIFLORA, Benth. Lupinus concinnus, Agardh. PICKERINGIA MONTANA, Nutt. FLORA OF THE SANTA BARBARA ISLANDS. 203 SPIRHA DISCOLOR, Pursh. ALCHEMILLA ARVENSIS, Scopoli. SAXIFRAGA REFLEXA, Hook. TELLIMA CYMBALARIA, Gray. RIBES SANGUINEUM, Pursh., var. MALVACEUM, Gray. Ripes Menztiesi1, Pursh. | GODETIA QUADRIVULNERA, Spach. PEUCEDANUM CARUIFOLIUM, Torr. & Gray. Gauium Nutrauuu, Gray. STYLOCLINE GNAPHALIOIDES, Nutt. MaDIA DISSITIFLORA, Torr. & Gray. B@RIA GRACILIS, Gray. MATRICARIA DISCOIDEA, DC. SENECIO VULGARIS, L. MiorosEriIs LInDLEYI, Gray. MICROSERIS ELEGANS, Greene. MICROSERIS ANOMALA, Watson, Proc. Am. Acad., X XII, 475. MALACOTHRIX COULTERI, Gray. The flowers are light yellow. With it a plant having the same stout habit, large heads and general appearance, but with narrower involucral bracts, pappus wholly deciduous and receptacle naked, which seems to be a form of I. inde- cora and squalida, Greene. Ma LacoTHriIx CLEVELANDI, Gray. SPECULARIA BIFLORA, Gray. 204 CALIFORNIA ACADEMY OF SCIENCES. Gia Nevinu, Gray. GILIA ANDROSACEA, Steud. GILIA DIANTHOIDES, Endl. NEMOPHILA AURITA, Lindl. NEMOPHILA PARVIFLORA, Doug] PHACELIA VISCIDA, Torr. PHACELIA DISTANS, Gray. PLAGIOBOTHRYS CANESCENS, Benth. CONVOLVULUS PENTAPETALOIDES, L. MIMULUS LATIFOLIUS, Gray. MIMULUS LUTEUS, L. CASTILLEIA PARVIFLORA, Bong. APHYLLON FASCICULATUM, Gray. AUDIBERTIA NIVEA, Benth. AUDIBERTIA STACHYOIDES, Benth. ScUTELLARIA TUBEROSA, Benth. SALICORNIA AMBIGUA, Michx. Poputus FREMoNTI. var. WISLIZENI, Watson. QuERCUS LOBATA, Née. ALLIUM HYALINUM, Curran. Common throughout the island. ALLIUM LACUNOSUM, Watson. These two specius of Alliwm were determined by Dr. Sere- no Watson. FLORA OF THE SANTA BARBARA ISLANDS. CaLocHorTus ALBus, Dougl. CALOCHORTUS VENUSTUS, Benth. Bropi#A MINOR, Watson. LuzuLa comosa, Meyer. CAREX GLOBOSA, Boot. PHALARIS INTERMEDIA, Bose. TRISETUM BARBATUM, Steud. KG@LERIA CRISTATA, Pers. STIPA SETIGERA, Pres]. STIPA EMINENS, Cav. STIPA VIRIDULA, Trin. FESTUCA TENELLA, Willd. FESTUCA MICROSTACHYA, Nutt. Bromus Hooxertanvs, Thurb. Bromus ciuiatus, L. Poa annua, L. Poa Howe, V. & S. Determined by Dr. Geo. Vasey. EquiseTuM TELMATEIA, Ehrh. GYMNOGRAMME TRIANGULARIS, Kaulf. CHEILANTHES MYRIOPHYLLA, Desy. ADIANTUM EMARGINATUM, Hook. SELAGINELLA RUPESTRIS, Spring. 205 206 CALIFORNIA ACADEMY OF SCIENCES. II. FLORA OF SANTA ROSA ISLAND. The collection of Santa Rosa plants herein enumerated was made during the first ten days of June. The early vegetation had already disappeared and the Island had as- sumed the dry appearance common to western California after the spring rains have passed. Only the eastern and northern parts were visited, and doubtless an examination of the whole island earlier in the season will add a large number of species to its flora. To the owner of the island — Wr. Alexander P. More, a member of this Academy—for opportunity to make the col- lection and for facilities given, I am very greatly obliged. CLEMATIS LIGUSTICIFOLIA, Nutt. On Santa Cruz Island the flowers of some plants are per- fect, and the vines become three inches in diameter. RaANuNcuLUS CALIFoRNICUS, Benth. DeELPHINUM Parryl, Gray. EscHSCHOLTZIA CALIFORNICA, Cham. ==L. glauca, Greene. Eschscholtzias are very abundant on both Santa Rosa and Santa Cruz Islands, and the plants not only differ in habit, color of foliage and of flowers, but the same plant sometimes undergoes an almost specific change in general appearance. ‘Tall, large plants on Santa Cruz Island, hay- ing very glaucous foliage and graceful habit, by the end of April had become stiff and branching, had assumed a red- dish brown color throughout, and the flowers had changed from light yellow to deep orange. The color of the plants seems to depend somewhat upon the moisture of the soil, as is shown by a gradation from the glaucous ones near the bottom of cafions to the red-brown ones of the dry sides. On the loose soil of steep slopes the same plant seems to persist as an annual and become ZL. peninsularis, Greene. FLORA OF THE SANTA BARBARA ISLANDS. 207 Some of the seaward slopes of Santa Rosa Island abound with Eschscholtzias varying in color from red-brown to ex- tremely glaucous. The flowers are of all shades of color between light yellow and deep orange, and vary much in s1ze. DENDROMECON RIGIDUM, Benth. =D. Harfordii, Kellogg, and D. flexile, Greene. This shrub on Santa Rosa varies from forms having oval leaves with smooth margins to those having lanceolate leaves with rough margins. Sheltered from the wind it becomes D. flecile, Greene, but when exposed the leaves are smaller, narrower and rougher. PLATYSTEMON CALIFORNICUS, Benth. CHEIRANTHUS ASPER, Cham. & Schlecht. Seeds slightly winged. A form in situations sheltered from the wind resembles Erysimwm asperum in habit. ERYSIMUM INSULARE, Greene. SISYMBRIUM REFLEXUM, Nutt. LEPIDIUM LASIOPHYLLUM, Nutt. CAPSELLA DiIvaRIcATA, Walp. ISOMERIS ARBOREA, Nutt. VIOLA PEDUNCULATA, Torr. & Gray. HELIANTHEMUM SCOPARIUM, Nutt. Prostrate-spreading when growing in situations exposed to the wind. FRANKENIA GRANDIFLORA, Cham, & Schlecht. SILENE LACINIATA, Cav. =WS. simulans, Greene. Very common on both Santa Cruz and Santa Rosa Islands. Plants vary from a few inches to four feet in height. 208 CALIFORNIA ACADEMY OF SCIENCES. * SILENE GALLIca, L. SAGINA OCCIDENTALIS, Watson. STELLARIA MEDIA, Smith. LEPIGONUM MACROTHECUM, Fisch. & Meyer. PENTACHNA RAMOSISSIMA, Hook. CLAYTONIA PERFOLIATA, Donn. A form of this species having linear leaves is not uncom- mon on Santa Cruz Island. One specimen collected has the flowers glomerate on the disk and several linear radical leaves with one broadly deltoid. SIDALCEA MALVEFLORA, Gray. MALVA BOREALIS, Wallm. Eropium cicuTarium, L’Her. CEANOTHUS CRASSIFOLIUS, Torr. CEANOTHUS ARBOREUS, Greene. Smaller than on Santa Cruz Island, and hardly more than a bush in appearance. Leaves nearly entire and smooth. An interesting form indicating its mainland representative. Ruus pDIversILoBA, Torr. & Gray. RUS INTEGRIFOLIA, Benth. & Hook. HOSACKIA STRIGOSA, Nutt. HOSsACKIA MARITIMA, Nutt. Hosackia GLABRA, Torr. =Syrmatiwm dendroideum, Greene. Some of its forms are exactly the mainland plants. Victa AmeERIcANA, Mihl. FLORA OF THE SANTA BARBARA ISLANDS. 209 Lupinus CHAmIssonis, Esch. Lupinus micRANTHUs, Dougl. = L. umbellatus, Greene. The plants are slightly more spicate than those of Santa Cruz Island, and the mainland plants are not ditferent from those of both Islands. TRIFOLIUM TRIDENTATUM, Lindl. ASTRAGALUS LEUCOPsIS, Torr. & Gray. ASTRAGALUS MIGUELENSIS, Greene. Very abundant in some localities. It is probably a vari- ety of A. lewcopsis. MELILOTUS PARVIFLORA, Desf. Prunus micironius, Walp., var. OCCIDENTALIS (Lyon). =P. occidentalis, Lyon. On Santa Rosa Island it is confined to the bottom of the cafions, and is much larger and more tree-like than on Santa Cruz Island On Santa Cruz it sends up several trunks from one root and resembles a big bush, but on Santa Rosa a single trunk ascends sometimes fifteen feet before it be- gins to branch. The leaves are generally more or less sin- uate-dentate and sometimes entire. In the Santa Inez Mountains, near Santa Barbara, the leaves sometimes have entire margins and it becomes six inches in diameter and attains a height of fifteen feet. A large specimen near San Francisco has a diameter of more than two feet. Rusus URSINUS, Cham. & Schlecht. ADENOSTOMA FASCICULATUM, Hook. & Arn. Spreading-prostrate in situations exposed to the wind. Rosa Catirornica, Cham. & Schlecht. HETEROMELES ARBUTIFOLIA, Roemer. 2.10 CALIFORNIA ACADEMY OF SCIENCES. LYONOTHAMNUS ASPLENIFOLIUS, Greene. Not common. The trees are small and often distorted by the wind. This species always forms small groves of a hun- dred more or less trees. It sends up several trunks from one crown, and the whole grove probably is connected un- derground by its roots. HEUCHERA PILOSISSIMA, Fisch. & Meyer. = H. maxima, Greene. Not uncommon throughout Santa Cruz Island and very abundant in the cafons of Santa Rosa. When growing amongst bushes on Santa Cruz the flowering stems reach a height of five feet. On sun-exposed rocks it is often not more than six inches high. TILLHA MINIMA, Miers. COTYLEDON LANCEOLATA, Watson. (ENOTHERA BISTORTA, Nutt. (ENOTHERA DENTATA, Cav. CENOTHERA CHEIRANTHIFOLIA, Horn. Many forms, one of which is /. nitida, Greene. ZAUSCHNERIA CALIFORNICA, Presl. GODETIA QUADRIVULNERA, Spach. CLARKIA ELEGANS, Dougl. MENTZELIA MIORANTHA, Torr. & Gray. ECHINOCYSTIS FABACEA, Naudin. The plants of both Santa Cruz and Santa Rosa Islands seem to be this species rather than HL. mucrocarpa. They mature either four or eight large seeds. OpunTIA ENGELMANNI, Salm., var. (?) LITTORALIS, Engelm. FLORA OF THE SANTA BARBARA ISLANDS. 211 MESEMBRIANTHEMUM CRYSTALLINUM, L. Very abundant. Daucus pusinLus, Mx. PEUCEDANUM CaRUIFOLIUM, Torr. & Gray. SampBucus GLAuca, Nutt. SYMPHORICARPOS MOLLIS, Nutt. LonIcERA HISPIDULA, Dougl., var. VACILLANS, Gray. GaLium APARINE, L. GaLium NUTTALLI, Gray. —G. Miguelense, Greene. Climbing high amongst bushes, or less luxuriant on rocky hillsides, or depressed prostrate when exposed to the wind, and then it becomes G. Miguelense. Berry shining white. GALIUM ANGUSTIFOLIUM, Nutt. GRINDELIA GLUTINOSA, Dunal. = G. latifolia, Kellogg. Referred to this species by Dr. Gray. Small specimens are the same as some from the mainland. The involucre is more or less squarrose and the akenes are auriculate - bor- dered. Pappus awns 2—3, and sparingly ciliolate-scabrous. Very balsamic-viscid during anthesis. BIGELOVIA VENETA, Gray. SotmpaGo CaLiForNica, Nutt. CORETHROGYNE FILAGINIFOLIA, Nutt. DIPLOSTEPHIUM CANUM, Gray. = Hazardia cana, serrata and detonsa, Greene. Leaves vary in outline from sharply serrate and crenate to almost entire. ERIGERON FOLIOsuS, Nutt. ERIGERON GLAvuCUS, Ker. FIP CALIFORNIA ACADEMY OF SCIENCES. ERIGERON SANCTARUM, Watson, n. sp. in litt. Santa Inez Mountains, near Santa Barbara and Santa Rosa Island. AsTER FOLIACEUS, Lindl. BAcCCHARIS PILULARIS, DC. BaccHaris Dovenasi, DC. STYLOCLINE GNAPHALIOIDES, Nutt. Fiiuaco CaLirornica, Nutt. GNAPHALIUM PURPUREUM, L. GNAPHALIUM DECURRENS, Ives, var. CaLIFORNICUM, Gray. GNAPHALIUM SPRENGELIL, Hook. & Arn. LEPTOSYNE GIGANTEA, Kellogg. Mania sativa, Molina. Hemizonia Fascicutata, Torr. & Gray. HEMIZONIA PANICULATA, Gray. LAYIA PLATYGLOSSA, Gray. VENEGASIA CARPESIOIDES, DC. B#RIA GRACILIS, Gray. Bamria PaLMERI, var. CLEMENTINA, Gray. The specimens from Santa Rosa Island are more fleshy- thickened and the bracts more strongly carinate than those from Santa Cruz. Pappus of the disk flowers generally of four, sometimes of as many as seven pale; of the ray flow- ers mostly two, often one and sometimes none, similar to the pappus of B. gracilis from the Islands and Santa Bar- bara. FLORA OF THE SANTA BARBARA ISLANDS. Biles ERIOPHYLLUM STHCHADIFOLIUM, Lag. CHENACTIS TENUIFOLIA, Nutt. AMBLYOPAPPUS PUSILLUS, Hook. & Arn. ANTHEMIS Cortu.La, L. AcHILLEA MiLLEFouium, L. ARTEMISIA CALIFORNICA, Less. Artemisia Lupovictana, Nutt. CNICUS OCCIDENTALIS, Gray. = C. lilacinus, Greene. This species on Santa Cruz Island has the involucral bracts strongly incurved; on Santa Rosa, either straight or strongly incurved; in the Santa Inez Mountains, sometimes strongly incurved, commonly slightly incurved and often straight. The Santa Rosa Island plants having strongly in- curved bracts to the first large flowers, often have straight ones to the smaller heads that are produced later in the season. CENTAUREA MELITENSIS, L. PEREZIA MICROCEPHALA, Gray. STEPHANOMERIA VIRGATA, Benth. Annual, from six inches high upward; leaves more or less resinous dotted; pappus white, fragile and easily sep- arating from the akene; including S. tomentosa, Greene, and probably S. elata, Nutt. STEPHANOMERIA EXIGUA, Nutt. Annual, pappus bristles more persistent, more or less dilated at base and united into phalanges, and often with sete intermixed. MICROSERIS LINEARIFOLIA, Torr. & Gray. Plants which would be referred to I. macrocheta, Gray, 214 CALIFORNIA ACADEMY OF SCIENCES. were collected and are evidently a form of MM. linearifvlia. Robust plants with strong fruiting heads of M. linearifolia sometimes send up from near the base scapes bearing heads with the pappus of M. macrocheta. The fruit of the J. ma- crochceeta form seems always to be undeveloped. Mrcroseris LINDLEYI, Gray. MaLacoTHRiIx INCANA, Torr. & Gray. MALACOTHRIX SAXATILIS, Torr. & Gray. HIeRACIUM ARGUTUM, Nutt. TROXIMON GRANDIFLORUM, Gray. SoNCHUS OLERACEUS, L. SoONCHUS ASPER, Fuchs. ARCTOSTAPHYLOS TOMENTOSA, Dougl. ARCTOSTAPHYLOS PUNGENS, HBK. =A. insulavis, Greene. ARCTOSTAPHYLOS DIVERSIFOLIA, Parry. DoDECATHEON HENDERSONI, Gray. Erytar®a Dovueuasil, Gray. GILIA ATRACTYLOIDES, Steud. GILIA ANDROSACEA, Steud. Git1a NEVINI, Gray. ELLISIA CHRSANTHEMIFOLIA, Benth. PHACELIA DISTANS, Benth. = PP. scabrella, Greene. Calyx lobes rarely incised. Leaves not as finely and compoundly dissected as they are in the common forms. PHACELIA VIScIDA, Torr. PHACELIA RAMOSISSIMA, Dougl. or FLORA OF THE SANTA BARBARA ISLANDS. nal KRYNITZKIA LEIOCARPA, Fisch. & Meyer. AMSINCKIA INTERMEDIA, Fisch. & Meyer. CONVOLVULUS MACROSTEGIUS, Greene. DICHONDRA ARGENTEA, Willd. (?) Soranum NieruM, L., var. Douauastt, Gray. SoLtanumM XANTI, var. WALLACEI, Gray. LINARIA CANADENSIS, Dum. ANTIRRHINUM NUTTALLIANUM, Benth. CoLLINsIA BICOLOR, Benth. PENTSTEMON CORDIFOLIUS, Benth. MimuLus aLurinosus, Wendl. = MM. puniceus, Steud., Diplacus arachnoidens and parviflorus, Greene. Abundant on both Santa Cruz and Santa Rosa Islands: very variable, with all the forms and colors running into each other. MimuLus Lurteus, L. MIMULUS NASUTUS, Greene. CASTILLEIA PARVIFLORA, Bong. CASTILLEIA HOLOLEUCA, Greene. Answers to the description of C. foliolosa, Hook. & Arn., in the Synoptical Flora, but differs from the Santa Inez forms in the character of its pubescence. Its calyx is near- ly equally cleft and the bracts are usually red. ORTHOCARPUS PURPURASCENS, Benth. APHYLLON TUBEROSUM, Gray. SPHACELE CALYCINA, Benth. = S. /ragrans, Greene. SALVIA COLUMBARIZ, Benth. 2p SER Voz. I. (15) Issued October 11, 1888. 216 CALIFORNIA ACADEMY OF SCIENCES. AUDIBERTIA STACHYOIDES, Benth. var. REVOLUTA. Margins of the leaves revolute, giving to the bush a pe- culiar appearance. Sracuys BULLATA, Benth. = 8S. acuminata, Greene. PLANTAGO ParTaGonica, Jacq. ERIOGONUM NuDUM, Dougl. = LH. grande and rubescens, Greene. The rose-colored variety is the form of Santa Rosa Isl- and, and differs only in color from my Santa Cruz specimens, and both agree with mainland forms. ERIOGONUM ARBORESCENS, Greene. RuUMEX SALICIFOLIUS, Weinm. PoOLYGONUM AVICULARE, L. CHORIZANTHE STATICOIDES, Benth. PTEROSTEGIA DRYMARIOIDES, Wisch. & Meyer. ABRONIA UMBELLATA, Lam. CHENOPODIUM AMBROSIOIDES, L. CHENOPODIUM ALBUM, L. ATRIPLEX CALIFORNICA, Moquin SumpA TORREYANA, Watson. SALICORNIA AMBIGUA, Michx. PARIETARIA DEBILIS, Forst. SaLIx LHVIGATA, Bebb. PopuLus TRICHOCARPA, ‘Torr. QuERCUS DuMOSA, Nutt. QueERcus LoBaTA, Née. FLORA OF THE SANTA BARBARA ISLANDS. yA QUERCUS TOMENTELLA, Engelm. Common in cafions on the east and north sides. Larger than on Santa Cruz Island. QUERCUS AGRIFOLIA, Liebm. PINUS INSIGNIS, Dougl., var. BINATA, Engelm. Pinus Torreyana, Parry. The Forestry Report of the Tenth Census gives a station near San Diego as the only certain habitat of this pine and adds that it is doubtfully reported from Lower California and one of the islands off Santa Barbara. About one hun- dred trees are growing on the bluffs of the eastern shore. HABENARIA ELEGANS, Bolander. SISYRINCHIUM BELLUM, Watson. ALLIUM LACUNOSUM, Watson. BLOOMERIA AUREA, Kellogg. Bropima CAPiITaTa, Benth. = B. insularis, Greene. Plants no different from the island forms and equally as tall and broad leaved flourish in favorable situations of the Santa Inez Mountains. Lintum Humepouptitit, Roezl. & Leicht. CaLocHortus ALBUS, Doug]. ZYGADENUS Fremont, Torr. PuHynbLospapix Torreyi, Watson. Luzuia comosa, Meyer. JUNCUS BUFONIUS, L. JUNCUS PATENS, Meyer. Juncus Batricus, Dethard. ScIRPUS PUNGENS, Vahl. 218 CALIFORNIA ACADEMY OF SCIENCES. Carex Dovauast, Boott. Porypocgon MonspPe.iensis, Desf. STIPA SETIGERA, Pres]. AGROSTIS VERTICILLATA, Vill. AGRosTIs SCOULERI, Trin. Determined by Dr. George Vasey. K@LERIA CRISTATA, Pers. DisTIcHiis spicata, L. Bromus Hooxerianvs, Thurb. ELYMUS CONDENSATUS, Presl. AGROPYRUM REPENS, Beauv. HorpEvuM muRINvUM, L. Festuca microstacuys, Nutt. Frsruca Myurus, L. Potypopium CaLirornicum, Kaulf. PELLHA ANDROMEDEFOLIA, Fée. CHEILANTHES MYRIOPHYLLA, Desv. GYMNOGRAMME TRIANGULARIS, Kaulf. PrERIS AQUILINA, L. ADIANTUM EMARGINATUM, Hook. ASPIDIUM RIGIDUM, Swartz. II. COMPARISONS BETWEEN THE FLORAS OF SANTA CRUZ AND SANTA ROSA ISLANDS AND THE SANTA INEZ MOUNTAINS. Prof, E. L. Greene in his ‘‘ Studies in the Botany of Cali- fornia and parts Adjacent,” published by the California FLORA OF THE SANTA BARBARA ISLANDS. 219 Academy of Sciences, Vol. 2, No. 7, brought to notice the occurrence on the island of Santa Cruz of several trees, either peculiar to the CaJifornian insular flora or endemic upon the island itself, and to add representatives of these trees to the Jesup Wood Collection of New York, Prof. C. S. Sargent asked me to go to the island and endeavor to obtain them. This opportunity to study the insular flora was gladly em- braced, and six weeks were spent in making a botanical col- lection from all parts of the island; then to add to my knowl- edge obtained from Santa Cruz and to enjoy the pleasure of exploring an island almost wholly unknown botanically, a visit was made to the adjoining island of Santa Rosa. At different times, collections of plants with especial reference to the insular floras were made in the Santa Inez Mountains representing the vegetation of the ocean slope of this coast range within a dozen miles or more of Santa Barbara. The general trend of the mainland coast from Santa Bar- bara to Point Conception is westerly, and about twenty-five miles to the south and nearly parallel to it, lie the islands of Santa Cruz, Santa Rosa and San Miguel. Santa Cruz, the largest in area and most easterly, is long and narrow in shape, rugged and mountainous in profile, with its surface much broken by deep, rocky canons. Santa Rosa, nearly as large as Santa Cruz, occupies a central position, distant from the other two islands about ten miles; is more nearly circular, or perhaps, rectangular in outline, having a width of about twelve miles; the central hills or peaks reach an altitude of about 1200 feet, not more than half that of the Santa Cruz mountains; its surface slopes from the highest points to the shore, and is somewhat broken by cafions and gulches. San Miguel, the most westerly, is a small island of low elevation. The situation and topography of the islands in connection with the prevailing winds, determine the variety of their 220 CALIFORNIA ACADEMY OF SCIENCES. flora and also somewhat the form and: habit of the species. The distorted trees, the prostrate spreading bushes and the encroaching sand dunes show plainly that the wind most affecting the vegetation comes from the north and northwest. Santa Cruz, on account of its easterly location, receives more protection from the Santa Inez Mountains against the north winds than Santa Rosa, and the high mountains of the island itself afford additional shelter to the plants and trees of its valleys and canons. Its large area, its sheltered position and its physical characteristics account for the large number of species growing upon Santa Cruz Island. San Miguel, the westerly member of this group of islands, is almost wholly unprotected by the mainland coast, and re- ceives the full force of the northwest winds. Santa Rosa, situated to the windward of Santa Cruz, and consequently receiving less protection from the northern coast, with its topography not so mountainous or so varied, possesses a flora smaller in number of species, and as it partakes somewhat of the conditions of both the adjoining islands its flora is a mixture of that of those two islands; the eastern side approaching nearly to the conditions of Santa Cruz, its flora is mainly of species growing upon that island, while the western side approaching San Miguel in situation and physical characteristics resembles it in gen- eral botanical features. The shrubby vegetation and trees of Santa Rosa, are nearly all confined to the sheltered hill- sides and canons of the eastern portion of the island. Many of the plants of the islands vary in their form and habit more or less from those of the same species growing upon the neighboring mainland, and those of one island of- ten differ much from those of another, and even upon the same island some species present a width of variation, the extreme limits of which appear distinct. The species of Santa Rosa, as would be expected from the conditions of their existence, vary not only from those of the mainland FLORA OF THE SANTA BARBARA ISLANDS. PA and Santa Cruz but even extremely upon the island itself. These local variations upon the islands are caused mainly by more or less exposure to the strong and almost incessant northerly winds. A fine example of this class of variation is afforded by the Dendromecon of Santa Rosa, which, ac- cording to its location, assumes the mainland form with narrow, rough-edged, or the insular form of Santa Cruz, with broad, smooth-margined leaves. The variation of a plant of one island from its form upon another island is always interesting, often puzzling, and from the lack of connecting forms is, in many instances, considered specific. The two species of the insular genus Lyonothamnus, one from the island of Santa Catalina, and the other from Santa Cruz and Santa Rosa, are so closely related, that, as far as known, their specific distinctness rests only upon the form of the leaf, and the closeness of their relationship is still further strengthened by the fern- like divisions of LZ. asplenifolius appearing upon many of the leaves of L. floribundus. The two species of Lyonothamnus will probably be con- sidered distinct; and the great extent of water separating the islands renders the existence of connecting forms doubt- ful. But Ceanothus arboreus of Santa Cruz and its representa- tive upon Santa Rosa, are examples of different forms of dif- ferent islands that can hardly be considered specifically distinct, although the specimens are apparently as much so as those of many received species of this genus. Between these examples of extreme variation and the larger number of plants which retain upon the different islands a similar form, numerous instances of more or less deviation might be given. Insular variation, or the variation of the island species from their mainland ‘representatives, is more complicated, but seems mainly either to be the result of an insular cli- mate or a separation of their respective habitats for a long Aya CALIFORNIA ACADEMY OF SCIENCES. interval of time, and sometimes of a’combination of both causes. The intervening miles of restless ocean rendering the ex- istence of connecting forms between a species of the main- land and its insular form improbable, tends to give recog- nized specific rank to forms which on the mainland would hardly be considered good varieties and accounts for many of the so called endemic species of islands. The direct effects of an insular climate upon vegetation, with least intervention from other forces, are best shown upon Santa Cruz and in the protected canons of Santa Rosa. No general rule will apply to all plants, but different species and genera seem to be variously influenced by the climatic conditions. As by far the larger part of the island flora is composed of the maritime and Coast Range plants, species that delight in ocean breezes, spray and fogs, it is not surprising that many of them, in the midst of such conditions, should often become very luxuriant. Of course differ- ence of size is not a specific difference, but often this insular luxuriance obscures the close relationship, which, on that account, is sometimes best shown by insular de- pauperate specimens, as in the case of Brodicea capitata, the Santa Rosa Grindelia, and other species. An interesting fact in connection with insular variation, and one that is not easily accounted for, is the tendency that mainland shrubs and bushes show on the islands not to develop their lower limbs and branches and therefore become tree-like. The most conspicuous examples of these tree-like bushes are Prunus ilicifolius, Cercocarpus, Rhus, Rhamnus, Ceanothus. An insular form is not always a more luxuriant, but sometimes is what may be considered a degenerate form: a good illustration of this sort of variation is afforded by FLORA OF THE SANTA BARBARA ISLANDS. 223 Malacothrix Coulteri, which is, however, not a Coast Range plant. This plant, on the eastern end of Santa Cruz Isl- and, is found in its mainland form, and growing with it plants varying toward Jl. insulavis, which, in exposed situations, might easily become J. squalida or indecora. A species which can flourish in the Sierra Nevada moun- tains, withstand the extreme summer heat of the cen- tral valleys, and also grow amongst the fogs of the Coast Range and adjacent islands, must reflect somewhat in its characters the very diverse conditions under which it ex- ists. The extreme differences of climate that are found within a few miles distance make the flora of California a flora of variable species, a flora that should be studied in the field or with abundant material from many localities. The division of the year into wet and dry seasons in- duces a form of variation which it may be not out of place to notice, especially as it has some connection with insular plants; it is that many perennial species undoubt- edly also persist as annuals if they are able to mature their seed before the summevr’s drought kills them. This pecu- liarity is alluded to by Dr. Engelman in the Botanical Ga- zette, Vol. VI, 235, with reference to Hschscholtzia. The great extent of coast line of the islands, Santa Cruz having over fifty miles and Santa Rosa more than forty, compared with their area, affords a large habitat for a maritime flora, and plants most abundant in the vicinity of salt water, form a large proportion of the insular flora. This is more noticeable on the sloping hills of Santa Rosa than on the bold and rocky shores of Santa Cruz, and ac- counts for the seemingly great preponderance of snch sea- shore plants as Eschscholtzia, Cotyledon, Abronia, Mesembri- anthemum, and others. The insular floras should be compared with those of the neighboring shore and coast range rather than with that of 2p SER. VOL. I. (16) Issued October 11, 1888. 224 CALIFORNIA ACADEMY OF SCIENCES. California in general, and then the absence of certain genera and species does not appear so remarkable. The summit of the coast range seems to be a barrier that limits the habitat of many species, but there are a few from the interior not found in the Coast Range flora that unex- pectedly reappear upon the islands. The known flora of Santa Cruz and Santa Rosa Islands now numbers close upon four hundred species, that of Santa Cruz amounting to three hundred and eighty, and that of Santa Rosa to nearly two hundred species, with only twenty found upon Santa Rosa that have not been reported from Santa Cruz. If the insular endemic flora of these islands is supposed to number twenty species, some doubtful ones must be in- cluded; nine of these twenty endemic species inhabit also Santa Catalina and Guadalupe Islands, leaving eleven or less, peculiar to Santa Cruz, Santa Rosa and San Miguel. Of the remaining three hundred and eighty species, over three hundred and fifty-five grow about Santa Barbara and in the adjoining’ Santa Inez mountains, leaving twenty-five species still to be accounted for, which, with the present knowledge of their distribution may be considered as be- longing to the San Diego flora, or in a few instances, to the plants of the interior region. The following notes and observations concerning main- land and insular plants having, perhaps, sufficient value to be worthy of publication, are here appended: Devpuinium Parryi, Gray. Bot. Gazette, XII, 50. This seems to be the most common species of the Santa Inez mountains, and is abundant upon the islands. PLAtysTIGMA CALIFORNICUM, Benth. & Hook. =P. denti- culatum, Greene. FLORA OF THE SANTA BARBARA ISLANDS. Oy The island specimens are more denticulate than the more or less denticulate forms of the mainland. THYSANOCARPUS LACINIATUS, Nutt. 7. vamosus, Greene. CALANDRINIA BREWERI, Watson. Appears to be more abundant upon Santa Cruz Island than on the mainland. Rueamnus crocea, Nutt. =A. insularis, Kellogg. The large leaved island Rhamnus is abundant in the Santa Inez mountains at the higher elevations, but is always a bush less than six feet in height. A smaller leaved form erows at lower elevations. This species in some localities of the more northern por- tions of the State, with the habit of a bush, becomes as tall as upon Santa Cruz, where it is a fine example of a main- land bush becoming an insular tree. Ruvs ovata, Watson. All the species of us are larger and more tree-like upon the island than on the mainland, even R. diversiloba almost becomes a small tree. In some of the southwest cafions of Santa Cruz near the sea, ft. ovata, with a single trunk and compact rounded head, resembled in appearance small ap- ple trees, and in full bloom presented a fine appearance, very different from the bush of the summits of the mainland mountains. It is one of the very first to start forth anew from its roots after all vegetation has apparently been killed by fire. CERCOCARPUS PARVIFOLIUS, Nutt. ==C. betulefolius, Nutt. A mainland shrub that often is an insular tree. Gatium CaLirornicum, Hook. & Arn. =G. flaccidum, Greene. As noted in the Flora of California this species often has a pubescent ovary, and the pubescence sometimes persists upon the mature white fruit. 226 CALIFORNIA ACADEMY OF SCIENCES. AUDIBERTIA STACHYOIDES, Benth. Common about Santa Barbara and on Santa Cruz Island. The Santa Barbara specimens, perhaps, collected from a single bush are pistillate, with no trace of stamens. CALOCHORTUS ALBUS, Doug]. Is white flowered upon the mainland, and light-purple flowered upon Santa Cruz. BOTANICAL NOTES. 227 BOTANICAL NOTES. BY MARY K. CURRAN. T. Plants from Baja California. The following small but very interesting collection of plants was made by Mr. Walter E. Bryant, already well known as an ornithologist, who spent some of the early months of this year in the vicinity of Magdalena Bay, studying and collecting the mammals and birds of that region. Several plants have been omitted from the list, being generically undeterminable because imperfect; and many are specifically so, for the same reason, or for lack of the necessary literature. ARGEMONE Mexicana L. var. ALBA. SISYMBRIUM CANESCENS Nutt. SISYMBRIUM, sp. LEpPIDIUM NITIDUM Nutt. LyROcARPA CouLTERI Hook. & Harv. BISCUTELLA, sp. OLIGOMERIS SUBULATA Boiss. KRAMERIA PARVIFOLIA Benth. DrymariA VEATCHII.—Prostrate-spreading, minutely pu- bescent; leaves ovate, tapering into slender petioles of about the same length and together with the flowers, fascicled in the axils; pedicels slender, 2-8 lines long; sepals concave oblong- ovate with broad scarious margins; petals nearly twice the length of the sepals, 6-8-lobed, the outer lobes rather large, spatulate, longer than the rest, and extending down- ward as an undulate margin to the claw, the intermediate ones linear or lanceolate: seeds dark-brown, strongly curved. 2p Ser. Voz. I. (16) Issued Dec. 11, 1888. 228 CALIFORNIA ACADEMY OF SCIENCES. First collected by Dr. Veatch, on Cedros Island—his fragmentary specimen so much resembling Mollugo verticil- lata that it has been overlooked till Mr, Bryant’s specimens came to hand. Fovgurera Formosa HBK. Gossypium Davipsont Kell. MALVASTRUM, Sp. MALVASTRUM, sp. ANODA HASTATA Oay, SPHMRALOBA, sp. ABUTILON, sp. Hrpiscus, sp. MELOCHIA, sp. FaGonraA CantrorntcA Benth. Larrea Mexicana Moric. Staphylea geniculata Kell. Viscainoa geniculata Greene. Flowering specimens of this obscure plant have lately been collected by C. R. Orcutt (distributed under name Chitonia simplicifolia Watson), by Dr. Palmer, and now by Mr. Bry- ant. The description appended is drawn from this fuller material. Fruiting specimens seem to have been collected only by Dr. Veatch. Shrubby, low and branching, pubescent or nearly glab- rous: leaves alternate sessile or short-petioled, entire (ovate lanceolate to cuneate-obovate, obtuse or retuse) or 2-4 pinnate with entire uneven leaflets, after the manner of Larrea; stipules small, subulate, caducous: flowers her- maphrodite: peduncles rather short, stout, 1-flowered, opposite the leaves, deflexed: sepals 4-6, uneven, imbri- cate, deciduous: petals pale-yellow, nearly orbicular, 4-6 lines long; dise obscure; stamens 8; filaments naked, subu- BOTANICAL NOTES. 229 late; anthers versatile: ovary villous,somewhat stipitate, 3-(1- lobed, 3-6-celled; styles short, coherent, as many,and stigma as many lobed as there are carpels in the ovary; ovules ana- tropous, two in each cell, pendulous from the inner angle; micropylar extremity produced upward, as in the figure of Guaiacum, Gray’s Genera, but in less tubular shape (in the description of Viscwinoa mentioned as a ‘‘small hemis- pherical white strophiole)”; raphe ventral; fruit more or less pubescent, 3-6 lobed, #-1 inch long, tipped by the stout styles; carpels coriaceous, separating at maturity from each other, navicular (in dehiscence), opening along the inner angle, and remaining attached by slender threads to the upper part of the axis formed by the united placenta: seeds two in each carpel, remaining suspended from above the middle of the axis, ovoid, with a fleshy green or glau- cous testa and thin dark-brown tegmen; albumen corneous- cartilaginous, rimose; embryo nearly straight in the axis of the albumen almost as long as the seed; cotyledons nearly orbicular, foliaceous, or a little fleshy, curved at the apex, their edges directed to the raphe and to the axis of the fruit, radicle short, conical, superior. It will be observed that the description of the seed is almost exactly that of Guaiacwm by Dr. Gray in Gen. of Am. Plants—even the wording has been intentionally fol- lowed in order to still more emphasize their resemblance. Some curious errors which appear in the earlier descrip- tions are herein corrected. Dr. Kellogg, in the original notice, states that there are pellucid dots in the leaves; a careful examination of these makes it probable that he was misled by small gummy exudations. The errors of the sec- ond description (Viscainoa), which contains the following statements: ‘‘testa dull and dark-brown; embryo very small at the base of a copious hard-cartilaginous or almost cor- neous albumen; cotyledons rounded, somewhat convolutely enfolding the short blunt radicle””—are more vital and quite incomprehensible in the light of the author’s succeeding 230 CALIFORNIA ACADEMY OF SCIENCES. remark—‘‘Concerning this rare and curious shrub of the Lower Californian peninsula no new knowledge is forth- coming beyond what has been gained by a minute and thor- ough examination of the good fruiting specimens collected by Dr. Veatch, almost thirty years ago, and now preserved in the herbarium of the California Academy of Sciences, Pittonia I., 163.” Equally at variance with the facts are the deductions drawn from the merely external appearance of the capsules, on account of which it was relegated to the Eu- phorbiacez and placed near Simmondsia. Its true place is certainly in Zygophyllacez between Guaiacum and Chitonia, and probably a slight modification of the generic character will admit it into the latter genus. ZIZYPHUS, sp. CARDIOSPERMUM MOLLE HBK. CARDIOSPERMUM TORTUOSUM Benth. ? DaLEA, sp. DALEA, sp. DALEA, sp. DaALEA, sp. Hosackia, sp.—Near H. rigida, but annual. PHASEOLUS FILIFORMIS Benth. CALLIANDRIA, sp.—Near C. myriophylla. ACACIA, sp. Cassia, sp.—Near C. Lindheimeri. COTYLEDON, sp. (ENOTHERA CRASSIFOLIA Greene. (ENoTHERA.—Near CZ. speciosa. GAURA PARVIFLORA Dougl. BOTANICAL NOTES. 231 The discovery of a second species of the following genus renders necessary some modification of the generic charac- ter. In habit the species are very unlike, the first being a herbaceous annual, with widely separated leaves, the second a shrubby perennial, with small crowded ones. They agree, however, in two important particulars—the calyx throat is closed by a glandular disk, below which the pistil is adnate to the tube—at least it is not free in the cav- ity—and the ovary is developed within a leafy branch. Gongylocarpus Cham. & Schl. Herbs or shrubs. Leaves alternate. Flowers axillary, solitary. Base of the calyx-tube adnate to the branch, slender, 4-lobed, long-produced, closed at the throat by a fleshy glandular disc; limb 4-lobed. Petals 4, inserted with the stamens below the margin of the annular disc. Stamens 8, the alternate ones shorter. Anthers ovate; ovary 2-3-locular; style filiform; stigma capitate; ovule solitary in the loculus. Fruit woody, enclosed within the branch. Embryo white, filling the seed. G. RUBRICAULIS Cham. & Schl. G. FRUTESCENS.—Shrubby, diffusely branched, 1-3 feet high; bark shreddy; leaves alternate, often many-ranked by slow development of the axis, entire, fleshy, oblanceolate, tapering into a short petiole: calyx rose-colored; tube slen- der, about an inch long; sepals reflexed, linear-lanceolate, about as long as the petals; tips free in the bud: petals paler, obovate-oval, a half-inch or more in length: stamens shorter than the corolla; filaments filiform: ovary irregularly angled, 2-celled: fruit an aggregation of 3-20 ovaries enclosed within a terminal twig at first densely covered with leaves and flowers, which falling leave it marked only by the small protuberances to which the calyx-tubes were attached. Don CALIFORNIA ACADEMY OF SCIENCES. Tt has not been possible with the material at command to determine with certainty the shape of the embryo; it is more or less curved, somewhat irregularly so, the ovaries altering slightly by mutual pressure. The fruiting twigs look like oblong irregular galls, more or less constricted in parts, and seem to remain on the plant for a considerable time—perhaps more than a year. This very peculiar genus seems to be nearest to Jussiea and Ludwigia, the capsules of which, with alternate bracts upon the peduncle, or even on the ovary, show a distinct approach to it, and the structure of the glandular disc is very similar. EUCNIDE CorDATA Kell. MENTZELIA HISPIpA Willd. SYMPETALEIA AUREA Gray. Marah minima Kell. PASSIFLORA FETIDA L. HOovusTONIA ASPERULOIDES Gray. AGERATUM, sp. HorMEISTERIA FASCICULATA (Benth). FRANSERIA CHENOPODIFOLIA Benth. FRANSERIA Bryantu.—Suffrutescent, low and branching, nearly glabrous: leaves bipinnatifid, with linear lobes: é spike short 10-15 flowered: ? solitary, 2-3 flowered, usu- ally from the same axil as the é spike: fruiting involucre, as well as the weak inch-long subulate-terete, straight spines, smooth and shining white. A remarkable species from its very long, white spines, which give it a rather formidable appearance. HELIOPSIS, sp. VIGUIERA SUBINCISA? Benth. BOTANICAL NOTES. 233 BIDENS BIPINNATA Gray. LEPTOSYNE COREOCARPA Gray.—Of what I take to be this species there are three examples: one, much the larger plant, agrees very well with Dr. Gray’s description, to which may be added —akenes strongly scabrous on both surfaces, rays rather large, purple. In the two other plants the rays are apparently white, the akenes smooth on both surfaces, and the border not cleft entirely through. Perhaps this last form may be L. parthenoides, or dissecta—species which Dr. Gray seems to have suspected to be too near the above. PerityLeE Cauirornica Benth. & var. NuDA Gray.—Rays yellow; leaves varying greatly in size and dissection. PerityLe Frront Torr.—Rays white, conspicuous. DyYsoDIA ANTHEMIDIFOLIA Benth. PoropHyLLuM.—A shrubby species. NICOLLETIA Epwarpsit Gray. — Probably this species, though a much more robust plant than the one described, possibly perennial; rays yellow, becoming purplish with age; glands infrequent on the involucre and small on the leaves; scales of the pappus lacerate, the awn arising from a deep notch. PrCcTIS MULTISETA Benth.—Pappus of the disk usually awnless, and with merely the rudiment of a crown. TRIXIS ANGUSTIFOLIA D. C. MALACOTHRIX XANTI Gray. ASCLEPIAS SUBULATA Decaisne. PHACELIA PEDICELLATA Gray. NAMA HISPIDUM Gray. HELIOTROPIUM CURASSAVICUM L. HELIOTROPIUM, sp. 234 CALIFORNIA ACADEMY OF SCIENCES. HELIOTROPIUM, sp. KrynitzK1a, near K. angustifolia, Gray. KRYNITZEIA, sp. IpoMa, sp. CONVOLVULUS, sp. SoLANUM ELHAGNIFOLIUM Cav. SOLANUM, sp. PHYSALIS GLABRA Benth. PHYSALIS, sp, NICOTIANA, sp. Lyctum, sp., near L. Californicum. Lyctum, sp., near L. gracilipes. Many of our species of Lyciwm seem too closely related; a careful study of intermediate forms in the field would probably much reduce their number. ANTIRRHINUM CYATHIFERUM Benth. Dr. Gray, in his Synoptical Flora, casts a doubt upon the distinctness of his A. chytrospermum, in the following words: ‘*A, cyathiferum of Lower California appears to differ from the following in having linear-lanceolate sepals of only half the length of the tube of the corolla and a shallower cup to the seeds.” The specimens brought by Mr. Bryant from Mag- dalena Bay show great variation in the comparative length of the sepals and tube of the corolla, as well as in the cups of the seeds, so that it is probable a careful comparison will show them to be too closely related. No one familiar with Mohavea viscida Gray (Antirrhinwn confertifiorum Benth.) can look at these plants and fail to be struck by their close relationship, especially in habit, cap- sules and seeds. BOTANICAL NOTES. 235 STEMODIA DURANTIFOLIA (L.) CASTILLEIA, sp. MARTYNIA ALTHEEFOLIA Benth. RUELLIA, sp. RUELLIA, sp. BELOPERONE CALIFORNICA Benth. VERBENA, Sp. LiIpPia, sp. TEUCRIUM GLANDULOSUM Kell. Hyptis Emoryi Torr. STACHYS COCCINEA Jacq. PLANTAGO PaTAGONICA Jacq. ABRONIA MARITIMA Lam. ABRONIA GRACILIS Benth.—Probably a form of A. wmbel- lata. Mirasinis LavIs (Benth.) Oxybaphus levis Benth. IM. Cal- yornca Gray.—Mr. Bryant’s specimens, brought from Magdalena Bay, are nearly but not quite glabrous, the ine- quality of the involucral lobes variable and often not greater than is found in our Californian forms. ALLIONIA INCARNATA L. PTEROSTEGIA MACROPTERA Benth. Dr. Parry and Mr. Greene propose a new genus, Harfor- dia, for this species, P. galioides Greene, which he considers identical with it; and P. fruticosa Greene, from Cedros Is- land, because, as they state, it “differs from Pterostegia in its perennial habit, its axile excentric embryo, and from all known Eriogonez in its dicecious flowers.” That the above statements are, to a certain extent, erro- neous is shown by these specimens brought from the original locality. 236 CALIFORNIA ACADEMY OF SCIENCES. P. galioides, which Dr. Parry reduces to a synonym of P. macroptera, is in our herbarium specimens much more unlike that species than is P. fruticosa, which seems to dif- fer from it only inits more shrubby and compact habit and shorter nodes. The leaves in Mr. Bryant’s specimens are spatulate-obovate, but much broader than in P. fruticosa —in Mr. Greene’s specimens of P. galioides they are very narrow. The flowers, so far as known, as well as the akenes, are identical in all. It is possible that these plants are sometimes dicecious, but P. macroptera, at least in our spec- imens, has male and female flowers from the same axils— the former soon falling. The embryo agrees with the di- agnosis of Pterostegia in Benth. & Hook. Gen. Pl. (who place it in Koenigiez, not in Eriogoneze)—‘‘Kmbryo excen- tricus, curvus, cotyledonibus suborbiculatis, radicula long- iuscula accumbenti-ascendente.” As compared with that of P. drymarioides it is not quite so much curved, and its cotyledons are oval or oblong instead of suborbicular. Any distinctions founded upon variations in form or de- velopment of the bracts lose all force upon examination of our familiar California species. It seems not to be known that this species (P. drymarioides) fruits in two different forms on the same plant, although Hooker and Arnott as long ago as the publication of Bot. Beechy described the flowers as polygamons. The two flowers springing from the axils of opposite leaves are usually unequally developed, one matures in the ordinary variable wing-saccate shape, and the other either aborts and falls, or the bract is folded simply and closely round the akene, the meeting margins erose-denticulate, or, especially in the southern forms, the bract is so littledeveloped that the maturing akene projects one-half its length beyond it. AMARANTUS PALMERI Watson. ATRIPLEX BARCLAYANA Dietr. ? SumpA ToRREYANA Watson. BOTANICAL NOTES. PAS wf EUPHORBIA ERIANTHA Benth. EUPHORBIA MICROMERIA Boiss. EUPHORBIA POLYCARPA Benth. KUPHORBIA, sp. PHYLLANTHUS, sp. SIMMONDSIA CALIFoRNICA Nutt. FIcus, sp. ANEMOPSIS CALIFORNICA Hook. LORANTHUS, sp. MUHLENBERGIA DEBILIS Trin. BoOvUTELOUA POLYSTACHYA Benth. BovuvTELOVA OLIGOSTACHYA Torr. ' Cencurus Patmert Vasey. II. Papaveree of the Pacific Coast. As long as species exist and naturalists are found to study them, there will probably be differences of opinion as to their limits, and it is far easier and more gratifying to the personal ambition of a botanist to name and describe forms which shall thenceforth bear his name as an appendage, than to patiently study variations from many localities, and probably end by finding them all run together under older names. There is, however, a limit to the multiplication of species after the manner which the lamented Engelmann ealls ‘‘the too common mistake of adopting the characters of a single individual as those of the species,” which is soon found in the minute description necessary for the sep- aration of forms, leading continually to the more mi- nute description of other forms which will