NOTES ON A COLLECTION OF FISHES MADE BY DR.EDGAR A. MEARNS FROM RIVERS TRIBUTARY TO THEGULF OF CALIFORNIA.By John Otterbein Snyder.Of Stanford University, California.
Wliile in charge of the biological section of the survey of theMexican boundary of the United States, from January, 1892, tiUSeptember, 1894, Dr. Edgar A. Mearns, major and surgeon, UnitedStates Army, made a collection of fishes from the various streamsencountered along the route. Those secured from the Gulf ofCalifornia drainage are of particular interest, for although considerablehas been written on the fishes of the region,^ there are yet manyquestions regarding their habits, relationships, and distribution whichare open to investigation.The collection made by Doctor Mearns contains specimens fromthe Colorado River and some of its tributaries, from the headwatersof the Yaqui near the international boundary, and also from theAltar and Sonoyta Rivers, each of which rims an independent courseto the Gulf of California. The following list will indicate the re-lationships of the streams examined:Colorado River.Gila River.Verde River.Oak Creek.Beaver Creek.Santa Cruz River.San Pedro River.
Sonoyta River.Altar River.Bear Creek.Yaqui River.Bavispe River.San Bernardino River.Cajon Bonito Creek.The faimas of the Sonoyta, Altar, and Yaqui Rivers are not yetweU enough known to justify any very definite statements as totheir relationships. It appears, however, that the fishes of theYaqui have been largely derived from the Rio Grande. It alsoseems safe to infer that the barrier of salt water which now separatesthe Colorado and Yaqui Rivers has been continuous with the exist-ence of the latter, thus preventing the Colorado fauna from extending
? Evermann and Rutter; Bull. U. S. Fish Comm., No. 14, 1894, pp. 473-486. Gilbert and Scofield;Proc. U. S. Nat. Mus., vol. 20, 1898, pp. 487-499. Meek, Seth Eugene; Field Columb. Mus. Publ. 93, zool.ser., vol. 5, Rio Yaqui System, p. XXXVHI. Ellis, Max M.; Univ. of Colorado Studies, vol. 11, No. 1,1914. Proceedings U. S. National Museum, Vol 49?No. 2125. 573
574 PROCEEDINGS OF THE NATIONAL MUSEUM. VOL. 49.to the Yaqui through a channel connection. At least two Coloradoforms have succeeded in reaching the larger streams of southernCalifornia, where they are now represented by Notolepidomyzon santor-anse and Richardsonius orcutti.One species, Notropis meamsi is described as new.Quotations are made from the field notes of Doctor Mearns.CATOSTOMUS INSIGNIS Baird and Girard.The species described by Kirsch ^ as Catostomus gila is synonjnnouswith C. insignis, no differences appearing when the types of both arecompared. On carefully reading Kirsch's paper it appears probablethat his collection included three specimens of C. insignis, which hedescribed as O. gila, and two examples of Notolepidomyzon clarki with
scales on the sides of moderate size, larger anteriorly than posteriorly * * *; theanterior scales are of circular outline * * * those of the back vary large.One of the latter he lists as C. vnsignis, and the other as C. clarki.There are four cotypes of G. insignis in the United States NationalMuseum. They have from 54 to 58 scales in the lateral series, 11 or12 between the lateral line and middle of back, 9 or 10 between thelateral Une and base of ventral, 26 to 30 between occiput and dorsalfin. There are 11 or 12 dorsal rays. In a cotype of C. gila the scalesare as follows: 56, 12, 11, 30.The following notes on the scales are taken from specimens collectedby Doctor Meams:Scales in lateral series 54Number of specimens 1Scales above lateral line 11Number of specimens 1Scales below lateral Une 10Number of specimens 10Scales before dorsal 26Number of specimens 3
55
NO. 2125. FISHES FROM THE GULF OF CALIFORNIA?SNYDER. 575The color of the ventral surface varies from pure white to a beautiful yellow. Theupper surface is dark, the line of division between dark and light being sharplydefined and located 3 to 5 scales below the lateral line.This is a large scaled sucker with elongate head, large lips, anddeep caudal peduncle. It has been reported from the lower parts ofthe Colorado system, its range evidently not extending up into theswift-flowing streams of the mountains.Doctor Mearns secured specimens from the Verde and Santa CruzRivers and from Beaver Creek near Fort Verde.CATOSTOMUS SONORENSIS Meek.Some small suckers from San Bernardino Creek are identified asC. sonorensis. These are somewhat more slender than that describedby Meek,^ and the scales and fin rays are similar to those of C. in-signis. No examples of the latter of similar size are available forcomparison.^San Bernardino Creek near the international boundary.NOTOLEPmOMYZON CLARKI (Balrd and Girard).A comparative study of the species usually assembled in the genusPantosteus discloses the fact that they are separable into two well-defined, natural groups, one characterized by a thick cranium inwhich the parietals and frontals meet in a close, strong suture (fig. 1),and another by a relatively thin cranium where the parietals andfrontals are separated by a long, often very narrow, fontanelle whichapparently does not close even in very old individuals. For the firstgroup the name Notolepidomyzon ^ is available, while Pantosteusmust be retained for the second.Catostomus has a broad, usually rectangular fontanelle (fig. 3),and the lips are not armed with a horny sheath as in Notolepido-myzon and Pantosteus.Recent authors recognize three species of Pantosteus in the Coloradosystem, namely, P. clarM (Baird and Girard), P. arizonae Gilbert,and P. delpMnus (Cope).'*Specimens measuring about 240 mm. from Beaver Creek near FortVerde, are referable to P. clarki, as they possess the very large andsomewhat rounded scales of the anterior upper parts which character-ize that species. The scales in the first row above the lateral fine are
1 Field Columb. Mus. Publ. 93, Zool. Ser., vol. 5, p. 32, pi. 6.* The type oi Catostomus bemardini No. 174, U.S.N.M., a specimen measuring 193 mm., has 12 rays in thedorsal, 7 in anal, 74 scales in lateral series, 15 in series between lateral line and middle of back, 12 below lateralline, and 33 between occiput and dorsal fin.8 Fowler, Henry W., Proc. Acad. Nat. Sci., Phila., 1913, p. 47, Notolepidomyzon type arizonae. Cope andYarrow; Wheeler's Explor. west of the 100th merid., vol. 5, 1875 (1876), p. 673, Pantosteus type platyrhyn-chus. It will be noted that Notolepidomyzon was characterized by the size of the scales, wliile the typeof Pantosteus has a large fontanelle.
* Proc. Acad. Nat. Sci. Phila., 1854, p. 27, Catostomus clarki; U. S. and Mex. Bound. Surv. Ichthyol.,p. 38, pi. 22, Minomus clarki; Bull. 47, U. S. Nat. Mus., p. 170, Pantosteus arizonae; Proc. U. S. Nat. Mus.,vol. 20, p. 488, pi. 36; U. S. Geol. Surv. Wyoming, Foiuth Ami. Rept., 1870 (1871), p. 435, MinomusdelpMnus.
576 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.49.not much, if any, larger than those of the row immediately below.Those of the succeeding rows above the lateral line rapidly and some-what regularly grow larger and more rounded in outline, until on themedian part of the back they are much larger than any on the caudalpeduncle. The hps have a prominent horny sheath. The skuU isthick, the edges of the parietals and frontals imiting in a firm suture,thus completely obUterating the fontanelle.Of the other species, P. arizonae has a skull of similar structure,and very large scales above the lateral Une. A comparison of thetype of P. arizonae and cotypes of C. clarTci indicates that they arespecimens of the same species. The scales in the lateral series arefound to number from 62 to 75, the series between lateral line andback 7 to 9, between lateral line and ventrals 11 to 14, between occiputand dorsal fin 15 to 17.Specimensfrom Beaver Creek.Lateral line 69 68 70 66 66 70 62 64Above lateral line 98987979Below lateral line 10 10 10 11 10 10 11 9Beforedorsal 16 15 16 17 17 17 15 17Type and paratype P. arizonae.Lateral line 67 69Above lateral line 8 8Below lateral line 11 HBefore dorsal 16 17Cotypes of C. clarki.Lateral line 65 66Above lateral line 8 9Below lateral line 11 10Before dorsal. : 17 17There is no material difference in the form of the body, proportionsof various parts, fin rays, etc., as the following table of measurementswill indicate
:
NO. 2125. FISHES FROM THE GULF OF CALIFORNIA?SNYDER. 577The other recognized Colorado species has been considered synony-mous with P. hardus and P. virescens, described by Cope.^ The typesof P. delphinus and P. hardus are lost, and the locality from whichthese and P. virescens came is not known with certainty. Thedescriptions of all are equally brief. The type of P. virescens (No.16758, U.S.N.M.) remains. It is apparently of the same species asothers in the United States National Museum identified by lateinvestigators as P. delphinus. It measures 370 mm., being by farthe largest specimen of the genus which the writer has seen.^ Thereis some reason to suspect that this is not synonymous with P. del-phinus (Cope), for that authority distinguishes the latter by its short,wide head and thick body. He also says that the scales are sub-equal. The type of P. virescens is relatively slender and the scalesof the anterior dorsal surface are much smaller than the others, a factwhich Cope records. However, as no more important reason appearsfor distinguishing between P. delphinus and P. virescens, they arehere regarded as synon3niious. The skuU of this species is relativelythin and the fontanelle constantly remains open.Unless the present treatment of these nominal fonns is at faultit is evident that there are but two species of Pantosteus {Pantosteusand Notolepidomyzon) in the Colorado system. Both are easily dis-tinguished from species of Catostomus by the structure of the lips, andthey differ markedly from each other in the size of the scales. Whenthe crania of both are compared, it is observed that the heavy, firmskuU of N. clarlci, with its closely fitting frontals and parietals, con-trasts strongly with the comparatively light, thin skull of P. del-phinus, the sharp-edged frontals and parietals of which are separatedby an elongate and narrow fontanelle.With these Colorado forms in mmd it will be of interest to examinesimilar catostomids of the Bonneville Basin, where three species havebeen described, Catostomus generosus Girard, Minomus platyrhynchusCope, and Minomus jarrovii Cope,^ only one of which has been gen-erally recognized. The cotypes of C. generosus and M. platyrhynchusare preserved in the Museum. The former represents a species witha short, robust body and a very thick skuU without fontanelle, hencebelongmg to Notolepidomyzon, while M. platyrhynchus is relativelyslender and has a thin skuU with an elongate, narrow fontanelle,*
1 U, S. Geol. Surv. Wyoming, Fourth Ann. Rept. for 1870 (1871), p. 436, Minomus bardus. Wheeler'sExplor. West of the 100th Merid., vol. 5, 1875 (1876), p. 675, P. virescens.2 Numerous specimens of Pantosteus measurihg from 230 to 395 mm., the largest of which weighed IJpounds, were lately collected by the writer while engaged in a survey of the Bonneville system underthe direction of the United States Bureau of Fisheries. In these the fontanelle is present.3 Proc. Acad. Nat. Sci. PhUa., 1856, p. 174. Proc. Amer. Philos. Soc. PhOa., 1874, pp. 134, 135.
* The fontanelle, covered or closed as usual with a tough membrane, must have escaped the attentionof Cope, as M. platyrhynchus is the first species mentioned and described under his diagnosis of the genusPantosteus, which he characterizes by "a complete union of the parietal bones which obliterates the fonta-nelle so universal among the suckers."81022??Proc.N.M.vol.49?15 37
578 PROCEEDiyOS of the XATIONAL museum. vol.49.a member of the genus Pantosteus. The cotypes of jan'ovii can notbe fomid. Cope remarks that this is a less elongate species thanplatyrhynchus, thus lending color to the supposition that his observa-tions were made on an example of generosus. It is also worth whileto note that the original description refers only to " several specimensfrom near Provo. Messrs. Yarrow and Henshaw." There can thenbe no mistake as to the type-locahty, and apparently no good reasonto assume that any cotypes came from the Rio Grande.^The Lahontan system has one species, P. lahontan Rutter,^ whichis related to P. platyrJiynclius, each resembling the other closely bothin cranial structure and external characteristics. One form, P.jordani ^ Evermann described from the upper part of the MissouriBasin, seems to be a representative of P. platyrhynchus, although thefontanelle is reduced to a very narrow slip. But one species seemsto be known from the Rio Grande, Oatostomus plebeius Baird andGirard.* It has the thick skull without fontaneUe characteristic ofNotolepidomy2on. Santa Ana River, a coastal stream of southernCalifornia, contains one species santa-anae, a member of the samegenus. Its describer was mistaken in its relationships, having com-pared it with wrongly identified specimens.^The relationships here indicated, and which are beUeved to be inharmony with both the structure and geographic distribution of thespecies in question, may be indicated as foUows:Lips with a horny cutting edge; cranium thin; a narrow and usu-ally sUtlike fontaneUe present (fig. 2):Pantosteus?delphinus Colorado.platyrhynchus Bonneville.jordani Columbia.lahontan Lahontan.Lips with a horny cutting edge; craniimi thick; no fontanelle(fig. 1):Notolepidomyzon?clarki Colorado.santa-anae Santa Ana.generosus Bonneville.plebeius Rio Grande.The center of distribution of each genus is the Bonneville Basin.
1 Bull. U. S. Fish. Conun., No. 12, 1S92, p. 56. List of specimens of PantosUus now found in the U. S.National Museum.2 BuU. U. S. Fish Comm., Xo. 22, 1902, p. 1-16.
' Bull. U. S. Fish Comm., Xo. 12, 1892, p. 51. The writer follows Jordan and Evermann in identifyingP. colujnbianus Eigenmann (Amer. Xaturalist, February 4, 1893, p. 151) with P. jordani, not having suffi-cient material for a careful comparison. A few poorly preser\-ed specimens from the Columbia indicatethat this identification should be accepted provisionally, for the two may be distinct species. There isdanger of confusing this form with C. calostomus. The latter has a narrow, rectangular fontanelle.
? Proc. Acad. Nat. Sci. Phila., 1854. p. 28.
? Proc. U. S. Nat. ilus., vol. 34, 1908, p. 33.
NO. 2125. FISHES FROM THE GULF OF CALIFOR^^IASNYDER. 579One is justified in looking forward to the possible discovery of oneor even more undescribed forms, or at least to the extension of therange, of known species. It is quite within reason to presume thatN. generosus is represented in the upper Columbia Basin. Thestatus of platyrhynchus, jordani, and columhianus should be carefuUyinvestigated.The specimens here used for comparison were about 6 inches long,perhaps the average size of mature individuals. In very old exam-ples of Pantosteus the fontaneUe becomes linear. In the large typeof P. inrescens the parietaLs and frontals nearly meet.The smallest specimen of N. clarki examined measured 70 mm. inlength. No fontaneUe was present, but the parietals and frontalswere not very firmly united along the median line.Beaver Creek, Santa Cruz River.XYRAUCHEN TEXANUS Abbott-iThis species is remarkable for the high and narrow dorsal hump orkeel which rises abruptly from the rather low, flat head, and extendsto the dorsal fin. The bony structure of this keel is formed by thejunction of an enlarged occipital crest and a series of fused or closelyapposed and modified spines and interneurals. The anterior andstronger part of the crest is supported by the complex of vertebraeover which the chain of Weberian ossicles extends (fig. 6). Thiscomplex involves 3 developed vertebrae, no trace of another ap-pearing in the mature skeleton. The first of these is representedby a thin centrum without lateral processes, above which is thespine {S.l), narrow and knife-like, with a heavy, thick base, closelyapposed to the first interneural {%.!.), the base of the second verte-bral spine and the neural arch. The second vertebra is not anky-losed with either the first or third. Its neural arch is large, and thespine (iS. 2) above it is massive. With the second spine are firmlycoosified the broad, leaf-like interneurals between the first and fifthspines. This spine-interneural complex is the "trapezoidal inter-neural" of Lockington 2 and the ''large, interneural formed by athick central pdlar with anterior and posterior wings'' of Kirsch.^The third spine {S. 3) is broad and low, closely joined to the base ofthe second, or ankylosed with it. The fourth is slender, short, andsupplied with a bifid tip which clasps the lower edges of the lastfused interneurals. Five other interneurals (one of which is occa-sionally absent or partly developed) with thin, broad tips and narrow,pointed bases, together with the first interneural of the dorsal finfinally complete the bony crest.
1 Fowler finds that X. cypho of authors is C. texantu Abbott. Proc. Acad. Nat. Sci. Phila., 1913, p. 54.* Proc. Acad. Nat. Sci. Phfla., 18S0, p. 237.* Proc. U.S.Nat.Mus., vol. 11, 1888, p. 556, description of Xt/rauthen Eigenmann and Earsch, new genus.
580 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. 4?.An osseous crest not much unlike this is present in Car'jnodes andIctiohus, the elements being especially distinct in the latter (fig. 5).It is rather feebly represented in Catostomus, where the coosifiedpart of the crest involves the spines and interneurals anterior to andincluding the sixth vertebra. The fourtli, fifth, and sixth spinesonly are easily distinguished, their tips extending upward astride aknife-like blade formed by the ankylosed interneuriils.The skull of Xyrauchen differs greatly from that of Catostomus,as will be seen by a comparison of the illustrations (figs. 3 and 4).The fontanelle in Xyrauchen is broad anteriorly, narrow posteriorly,almost triangular in shape, the posterior part continued baclvwardand upward as a narrow trough on the base of the occipital crest.In Catostomus (insignis, commersoni, macrocJieilus, mniotiltus, ta-Jioensis) the foramen is broad, oblong, and more or less rectangular.This and other peculiar cranial chai'acteristics of Xyrauchen indi-cate no very close relationship between it and Catostomus.The doi-sal crest of the species is exaggerated in old age by theflattening of the head, the doi-sal contour of which becomes stronglyconcave in contrast to the convex skuUs of Catostomus, Caryiodes,and others.Doctor Mearns notes that the flesh of this species is excellent andof fine flavor.Specimens were collected in the Colorado at Yuma, where oneexample with a peculiar abnormal squamation appeared, the scalesbeing only about half the usual size, except in several very restrictedareas, where a few normal ones were found.Very large specimens, refusing to take a hook, were snared in deep holes amongtlie rocks. A line with several hooks attached was allowed to sink to the bottomand when a school moved over it the line was brought out with a sharp swish of thepole, a sucker usually being hooked.The entire venti-al sm-face is orange yellow; pectoral and ventral fins very slightlydusky, their upper surface marked basally with dusky olive; iris yellow, withgreenish-yellow granulation; lower third of head, including end of snout, orangeyellow, in strong contrast with the flesh-colored mouth; sides salmon color, granu-lated with dusky olive; dorsal surface, including top of head, dark olive brown;anterior ridge oii hump pale olive. Dorsal I'm olivaceous; anal orange yellow, indis-tinctly marked with dusky; caudal yellowish olive. In some examples the fms arerather rosy, and the scales of the sides are tipped with silvery. An occasional examplehas the dorsal siu-face dark olive brown, and there is a strong pinkish tint below thelateral line.Colorado River at Yuma; Gila River at Gila City.GILA ELEGANS Baird and Girard.There is one example in the collection which measures 15f inchesin length. The head is greatly flattened and a broad, rounded,nuchal hump is developed.The very narrow caudal peduncle, the depressed and elongatehead, and the dorsal hump of 6^. elegans caUs to mind the fact that
NO. 2125. FISIIE.^ FROM THE OULF OF CALIFORNIA?SNYDER. 581tliesG peculiarities are characteristic of other Colorado forms. Anunusually narrow caudal peduncle distinguishes 0. latipinnis andG. elegans, the depressed head is found in Ptychocheilus lucius, G.rohusta, and X. texanus, and the dorsal crest is conspicuous in X.texanus and G. rohusta. These characters become fully developedonly in old individuals.Gila River at Gila City; Colorado River, Yuma; Salton River,Colorado Desert, Lower California.GILA ROBUSTA Cope.In large individuals the back is greatly arched, the head flattened,and concave in outline. The maxillary appears longer, the tip ex-tending beyond anterior margin of eye, which it does not reach insmall specimens.G. seminuda ^ seems to be synonymous with this species. In theventral region the scales are small, thin, and not securely embedded.The observations of Ellis,^ "Specimens of this subspecies differfrom the typical G. rohusta in having no scales in the midventralportion of the body as far posterior as the base of the ventrals andno scales on the mid-dorsal region as far back as the middle or lastray of the dorsal fin," are not applicable to this form, as the cotypesof G. seminuda from the Rio Virgen have the entire ventral surfacescaled, where the scales have not been displaced by accident, inwhich case the pockets remain. ' The same is also true of the mid-dorsal region.Fishes of this species may be caught with baited hook, "the fleshof a bird or minnows of its own species answering equally well. Thestomach of one specimen contained a frog."An example measuring 9^ inches from the Verde was colored inlife as foUows
:
Dorsal aspect dark, dotted with gold and silver; sides with obsolete blackish bands,most distinct toward the tail; below, silvery white; iris, golden; peritoneum, black.A smaller specimen was lighter colored and had a gold band extending from thesuperior extremity of the opercular opening to the base of the tail. In many speci-mens the ventral siu-face ranges in color from pure white to rosy salmon.The species is locally known as bony-tail or Verde trout, and when taken from coolwater the flesh has an excellent flavor.Verde River at Fort Verde; Gila River at Gila City.PTYCHOCHEILUS LUCIUS Glrard.Colorado River at Yuma; Gila River.RICHARDSONIUS GIBBOSUS (Baird and Girard).In determining the identity of specimens collected by DoctorMcarns it became necessary to examine the types of Gila gihhosa
1 Wheeler's Explor West of the 100th Merid., vol. 5, 1875 (1876), p. 666, pi. 31, figs. 1, la.2 Fishes of Colorado, Univ. of Colorado Studies, 1914, p. 57, Oila rohusta seminuda.
582 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.49.Baird and Girard/ Tigoma intermedia Girard,^ and Gila nigra Cope.^It was found that they represent the same species.The type of G. gihhosa, No. 223, U.S.N.M., is from the Rio SantaCruz, a tributary of the Gila. The following measurements of thespecimen are recorded in hundredths of the length to base ofcaudal fin. Length, 170 mm.; length head, 0.285; depth caudalpeduncle, 0.105; length caudal peduncle, 0.20; length snout, 0.08;diameter eye, 0.055; interorbital width, 0.09; depth head, 0.19;snout to occiput, 0.20; to dorsal, 0.58; to ventral, 0.54; length base ofdorsal, 0.18; base of anal, 0.09; height dorsal, 0.16; anal, 0.145;length pectoral, 0.16; length ventral, 0.12; number dorsal rays, 8;anal, 8; scales in lateral series, 74; between lateral line and middle ofback, 19; between occiput and dorsal fin, 36.One of the cotypes of T. intermedia, No. 232, U.S.N.M., has 76scales in the lateral series, 20 above the lateral line, and 39 betweenocciput and dorsal fin. There are about 8 small, short gillrakers.The cotypes of G. nigra, No. 16987, U.S.N.M., are from Ash Creek,Arizona. Specimens from San Carlos are also recorded by Cope, butthey should not be regarded as cotypes * as the original descriptionis based on larger examples, "7i inches," those from San Carlosbeing much smaller.Specimens from Olear Oieek al)oiit 20 miles above its confluence with the Verdewere examined about July 2. The back was dark olive green, the sides golden yel-low; sides of head brassy yellow granulated with black; ventral region yellow; irisyellow. Fins orange, the area around their bases vermilion.Santa Cruz River, Verde River near Fort Verde.A few fishes from Bear Creek, a tributary of Altar River, are pro-visional identified as R. gihhosus. They appear lo difl^er in formand scale characters from members of the genus taken in the Coloradoand in the Yaqui. Material for a careful comparison is not available,however. RICHARDSONIUS PURPUREUS (Girard).This species has a robust body, large head, and very short snout.The gillrakers are very short, scarcely evident in some specimens.The scales in the lateral series number from 55 to 60; above lateralline, 15 or 16; between lateral line and ventral, 8 or 9; betweenocciput and dorsal, 34 to 36. Dorsal rays, 8; anal rays, 8.San Bernardino River.NOTROPIS MEARNSI, new species.Five specimens of a Notropis, which apparently belongs to anunknown form, were collected m the Yaqui Basin. The species
? Troc. Acad. Nat. Sci. Phila., 1854, p. 28; U. S. and Mex. Bound. Surv. Ichthyol., p. 64, pi. 33, figs. 1-4.? I'roc. Acad. Nat. Sci. Phila., 1856, p. 206.3 Wheeler's Explor. West of the 100th Merid., vol. 5, 1876, p. 663, pi. 30, flg. 3.Bull. 47, U. S. Nat. Mas., p. 235, synonymy of L. niger.
NO. 2125. FISHES FROM THE OULF OF CALIFORNIA?SNYDER. 583which is Darned in honor of its discoverer is characterized b}^ therather rounded and elongate body, slender caudal peduncle, posteriorposition of the dorsal fin, large scales, large eye, teeth 4-5, short ali-mentary canal, almost complete lateral line, and the absence of darkbars or spots.On comparison with other forms the exact relationships of thespecies do not appear. It is probably allied to some form indigenousto the Rio Grande.Description of Notropis meamsi, type No. 76163, U.S.N.M., fromSan Bernardino River, Mexico, near monument 77 of the interna-tional boundary; Dr. E. A. Meams, collector; October 6, 1893.Head 4.1 in length to base of caudal; depth, 3.9; depth caudalpeduncle, 10; scales lateral series, 40; between lateral line and middleof back, 8; between occiput and insertion of dorsal, 19; dorsal rays, 8;anal, 8.
NOTEOPIS MEARNSI.Snout rather pointed, the length about equal to diameter of eye.Cleft of mouth oblique, extending slightly beyond anterior border oforbit. Dorsal inserted behind a vertical through base of ventral.Caudal deeply notched, the lobes pointed; 4 or 5 upper and lowerrays not fully developed. Edges of dorsal, anal, and ventrals convex;pectorals obtusely pointed. Teeth 4 on the right side, 5 on the left;strong, pointed, with hooked tips; a jiarrow grinding surface present.Peritoneum silvery; a few dusky spots on its dorsal surface; liningof opercle silvery. Alimentary canal shorter than body, folded onceand back. (Viscera and teeth from paratype.) Pseudobranchiaevery large; 4 or 5 very short stubby gill rakers.Scales large and regular. Lateral line complete or nearly so;slightly decurved anteriorly.Color silvery, darker and with little luster on the dorsal surface ; abroad, lateral, silvery stripe present, which gradually narrows oncaudal peduncle; no bars or spots; a linear, black stripe, not parallelwith the lateral line, extends from below dorsal fin nearly to base ofcaudal.
584 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. 49.The paratypcs exhibit some variation, as illustrated in the annexedtable. The pores of the lateral line fail to develop on occasionalscales, especially on the caudal peduncle.Notropis meamsi.? Measurements of type and paratypes.Length of body mm.. 41 40 35Lengthhead 0.26 0.25 0.255Depth body 27 .28 .28Depth caudal peduncle 10 .10 .11Length caudal peduncle 21 .25Length snout 06 .07 .07Diameter eye 065 .065 .07Interorbital width 075 .085 .085Depth head 19 .18 .19Snout to occiput 21 .20 . 21
.
Snout to dorsal 55 .58 .55Snout to ventral 50 .50 .51Length base of dorsal 12 .13 .14Length base of anal 115 .13 .12Heightdorsal 20 .21 .21Height anal 17 .16 .19Length pectoral 19 .20 .22Length ventral 17 .14 .17Length caudal 26 .27Doisal rays 8 8 8Anal rays 8 9 9Scales lateral line 40 37 38Scales above lateral line 8 8 8Scales below lateral line 5 4 4Scales before dorsal 19 18 17AGOSIA CHRYSOGASTER Glrard.The teeth of 10 specimens were examined and found to have awell-developed grinding surface. They were sharp and distinctlyhooked at the tips. There were 4 on each side. The fins are unusuallyshort and rounded in some cases, the ventrals not reaching the analopening. Tubercles are present on the heads of the males taken inJanuary, and the females contain many eggs in various stages ofdevelopment, some very large, apparently ripe. The peritoneum isdark; almost black.Doctor Mearns observes of the color:Blackish above, greenish olive on the sides, and white below. Steel-blue reflectionson sides of head.Sonoyta River. PLAGOPTERUS ARGENTISSIMUS Cope.Three specimens of this interesting spined minnow bear rudi-mentary scales on the anterior dorsal parts of the body. On thethroat, just posterior to the lower lip, is a mass of papillae like thoseon the lips of Catostomus.GUa River, near Gila City.
NO. 2125. Fli^IIES FROM THE GULF OF CALIFORNIA?SNYDER. 585SALMO nUDEUS Gibbons.A specimen 8^ inclies long from Oak Creek appears to representfishes that have been introduced from some CaUfornia stream. It isa beautiful trout with a very deep body and small head. The adi-pose dorsal is short and unusually high. The dorsal half of the bodyis closely covered with spots, sharply outlined and perfectly roundon the head where they are about half as large as the pupil ; smallerand irregular in outline on the body, many of them somewhatX-shaped. The dorsal and caudal are closely covered with verysmall spots, the dorsal with a row of elongate ones along the base.Lower half of head and body, pectorals, ventrals, and anal immacu-late. No teeth occur on the tongue posterior to those of the glos-sohyal. There are 140 rows of scales in the lateral series.Specimens of this same trout with 116 to 120 scales in the lateralline and measuring about 15 inches are recorded by Doctor Meamsfrom Clear Creek.This creek runs in a canyon with walls over a thousand feet high. At the bottomalong the stream is a forest of Douglas spruce, willow, aspen, walnut, maple, box-elder, pine, and hackberry. There is a dense growth of hop and grape vines togetherwith shrubs, annuals and ferns. The canyon walls are of limestone above, sand-stone below.Color above, a dark vinaceous olive, becoming silvery below the lateral line andyellowish beneath and on the opercles. Dorsal yellowish olive, thickly spotted;paired fins reddish orange, obsciu'ely spotted with dusky. Back and sides denselyspotted with black, the spots on head, back, and adipose fin being rounded, those onsides and posterior parts irregularly X-shaped. The iris is golden, as are the operclesand preopercles in places. On August 13 a female full of large eggs and measuring19J inches was taken in Oak Creek. The color was similar to that described above.Peritoneum whitish or colorless. A smaller one, 13 inches long, differed only inbeing whitish below. Other trout from Oak Creek were pure white below betweenthe pectoral and ventral fins; the sides to or above the lateral line, greenish yellow;above lateral line the ground color was olivaceous; iris yellow; mottled with black;Bides of head brassy yellow; dorsal and caudal greenish yellow.The stomachs often contained quantities of sand and small pebbles. Wasps,grasshoppers, and grubs were frequently found there also. The largest trout caughtweighed somewhat more than 3 pounds. June is the best month for fishing, whenthe trout are foimd high up toward the soiu-ce of the brook.Oak Creek. POECILIA OCCIDENTALIS (Balrd and Girard).I ." Specimens collected by Herbert Brown in Santa Cruz River nearTucson, November 20, 1893, contained eyed embryos. In oneexample, measuring 2| inches, there were 15 such young, besidesmany embryos and eggs in various stages of development.Gila River near Gila City and Adonde Siding; Santa Cruz River;San Bernardino River; Cajon Bonito Creek.
586 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. 49.CYPRINODON MACUIARTOS Baird and Girard.Examples colleoted in the Sonoyta February 1 are thus describedby Doctor Mearns.The females of this species are beautifully banded with black on a ground of greenishoUve above and white below. The posterior portion of the opercle is bright steel-blue. The males are not sharply banded and the colors are plainer. They arevery tenacious of life, and ascend small streams, entering cold springs wheneverpossible.Gila River at Gila Gty and Adonde Siding; Sonoyta River;spring near Sonoyta River.EXPLANATION OF PLATES.Plate 76.Crania of catostomids.Fig. 1. Notolepidomyzon clarki X 2. Beaver Creek, Arizona,2. Pantosteus lahontan X 3. Humboldt River, Nevada.3. Catostomus insignis H 1. Colorado River, Arizona.4. Xyrauchen cypho X f . Rio Verde, Arizona.Plate 77.Osseous structure of the dorsal crests ofFig. 5. Ictiobus bubalus6. Xyrauchen texanus Colorado River, Arizona.
U. S. NATIONAL MUSEUM

U. S. NATIONAL MUSEUM PROCEEDINGS, VOL. 49 PL. 77
l.-j >
?-5.S. /: / 'v;?
>\
6Osteology of Fishes.For explanation of plate see page 586.
i