THE FRESH-WATER AMPHIPODA OF NORTH AMERICA.
By Ada L. Weckel,,Of Moline, Illhiok.
All the species of Amphipoda known to occur in the fresh watersof North America were described by Prof. Sidne}' 1. Smith in hisreport on the Crustacea of the Fresh AVaters of the United States, pub-lished in 1S74. Previous to that report practicalh^ no work had beendone on the fresh-water forms belonging to this group. Since thenmany new species have been described, and the present paper isintended to include descriptions of all of the species of Amphipodanow known to occur in the fresh waters of North America. Onlytwo species are here described for the first time, but as I have had theopportunity of examining specimens?in some cases the types?ofmost of the known forms, it was thought best to redescribe and figurethem.The genera and also the species, in several instances, were in a stateof confusion; consequently an extensive synon3'my of each speciesdescribed has been included. A bibliography has also been addedwhich lists the principal papers dealing with the amphipod fauna ofthe region covered by this paper.I am greatly indebted to Dr. S. J. Holmes, of the University ofWisconsin, under whose direction this work was conducted, for manycourtesies received during the preparation of this paper. My thanksare due also to Prof. S. A. Forbes for the loan of a large collection ofmaterial from the Illinois State laboratory of natural history. A largenumber of specimens were ol)tained from the United States NationalMuseum, and the valuable aid thus received is gratefully acknowledged.KEY TO FAMILIES AND GENERA.A. First antenntP with secondary flagelluin; iifth pencopods shorter than preceding;two rami of third uropods nearly equal; second gnathopods smaller thanfirst Lysianassid.e. Pontoporeia, p. 26.AA. First antennfe with secondary flagellum; fifth per;eopods longer than the pre-ceding; second gnathopods generally larger than the first Gammakid^eB. Telson cleft; uropods biramous.0. Inner ramus rudimentary; telson cleft not more than three-fourths distanceto base.Proceedings U. S. National Museum, Vol. XXXII?No. 1507. 25
26 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxn.D. Outer ramus of third uropods uniarticulate Eacrangonyx, p. 29DD. Outer ramus of third uropods V)iarticalate NIphargus, p. 36CO. Inner ramus not rudimentary, one-halftothree-fourthsas long as outer, whichis biartieulate; telson cleft to the base, or nearly so Gammarus, p. 38BB. Telson entire.C. Third uropods with rami.D. Third uropods uniramous; telson short and broad Crangmn/x, p. 49DD. Third uropods biramous. Inner ramus rudimentary, outer uniarticulate;last two segments of urosome coalescent Stygonectea, p. 51CC. Third uropods without rami Apocrangonyx, p. 53AAA. First antenufe without secondary flagellum; telson entire, short and broad;third uropods uniramous Orchestiid/e; Ilyalella, p. 54Family LYSIANASSID^E.Onl}" two species, Pontopore'ui lioyl Smith, and I\mtoj)ore!a jillcorv IsSmith, have been described as belonging to this family. 1 have nothad specimens of either, but will quote Professor Smith's descriptionsof them. Genus PONTOPOREIA Ki-0yer.KEY TO SPECIES.A. Antennpe not reaching to the posterior end of the abdomen; outer ramus of thirduropods not more than twice as long as the peduncle P. hoyi, p. 26AA. Antenna? extending to or beyond the tip of the abdomen; outer ramus of thethird uropods more than twice as long as the peduncle P. filicornh, p. 28PONTOPOREIA HOYI Smith.Gammarus brenstylis Stimpsok, MSS. (female).Pontoporeia affinis Smith, Am. Jour. Sci., 3rd. ser., II, 1871, p. 453; Prelim. Rep.on Dredgings in Lake Superior, 1871, p. 1022.Pontoporeia hoyi Smith, liep. U. S. Fish Com., 1872-73, (1874) p. 647, {)!. Ji, fig. 5.Professor Smith's description follows:On first examining specimens of this species, obtained in Lake Superior in 1871, Iregarded them as specifically identical with the Pontoporeia affinis of the Scandina-vian lakes and the Baltic. A subsequent and more minute comparison has, how-ever, revealed some differences, which are apparently constant. In the form andproportions of the segments of the thorax and abdomen, in the size and form of theeyes, in the minute details of antennulpe, antennae, and iriouth appendages, 1 candetect no differences by whi(;h it would be possible to distinguish specimens takenin Lake Superior from those sent from Lake Wetter, or from the beautiful figures ofthe Scandinavian species given by Sars.? In the first pair of legs, however, thepropodus in the American species is proportionately a very little shorter than in theEuropean, and the palmary margin is less oblique?that is, it is not so nearly parallelwith the posterior margin; the posterior margin is somewhat shorter, and furnisliedwith fewer hairs; and there are usually two small and slender spines on the palmarymargin near the tip of the closed dactylus, while in the European species there areno real spines upon the palmary margin, but only slender setiform hairs. In boththe European and American species, there is a very thin and narrow lamellar edge,
aHistoire Naturelle des Crustacea d'Eau Douce de Norvege, 1867, p. 82, pi. vii,figs. 10-25; pi. VIII, figs. 1-5.
NO. 1507. FRESH- WA TER AMPHIPODA? WECKEL. 2 7extending nearlj' the whole length of the [)alinaiy margin. The dactylus is appar-ently a little longer and more slender in the European specie.s. The ohliquity of thepalmary margin, and its armature near the posterior angle, seem to be always char-acteristic of the American species. In the young specimens, however, there is oftenbut one spine, while in larger ones there are often three. In the third and fourthpairs of legs of the American species the dact3'lus is usually armed on the inside, alittle way from the tip, with two setiform haii'S, while in the European species thereis only one. Some young specimens of the American species, however, agree withthe European in having but one hair upou the dactylus, while large ones often havethree, and in the full-grown male from Lake Michigan, mentione<l farther on, thereare even "four.The most remarkable differences are in thepeculiai', elongated, j)apilliform append-ages upon the sternal portion of the thoracic segments. In the European species,Dr. G. 0. Sars describes and figures an elongated and slender process dependingfrom the middle of the sternum of several of the thoracic segments; and in the singlespecimen which I have examined, there are three of these processes, one each on thesecond, third, and fourth segments. Dr. Sars, who has studied the living animalsvery carefully, does not suggest what may be the use of the ajipendages, or whetherthey ever vary in number or position in different specimens. In specimens fromLake Superior there are usually seven of these appendages, one upon the second and
N'Toi'OREiA HoYi. ( Aftrr Smith.
)
two each upon the third, fi)urtli, and fifth segments. In form and size, these append-ages do not differ, except that in alcoholic specimens they seem to l)e a little longerin the American species. In some specimens of the American species, the append-age upon the second segment is wholly wanting, and in two specimens examinedcarefully there was only a single median one u{)on the third segment. In theabsence of all knowledge of the nature and use of these appendages, it seems uselessto speculate on their importance as distinctive characters. In a species of Ponto-poreia from the Gn\i of St. Lawrence, which I suppose to be the P. affinis of Kroyer,there are no such appendages on any of the thoracic segments. This fact, togetherwith the variation noticed in the specimens from Lake Superior, would seem to indi-cate that these appendages are not of so much importance as might at first besupposed.This species was found in great abundance in the dredgings in Lake Superior in1871, and occurred in every haul from 4 to 169 fathoms. It was also dredged byMr. J. W. Milner in Lake Superior in 1872, in 60 fathoms off Outer Island. It iscommon in the stomach of the whitetish from Lakes Superior and Michigan, andprobably also from the lower lakes. All the si^ecimens dredged in Lake Superiorwere taken in August and the early part of September, and none of the females werecarrying eggs during that time. Females carrying eggs were dredged by Dr. Stimp-son, in Lake Michigan, in 40 to 60 fathoms, off Racine, Wis., June 24, 1870, andwith them the adult male form with long antennulfe and antennte. This peculiar
28 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxii.form of the adult male, corresponding perfectly with the same form of the Europeanspecies figured and described by Dr. Sars, I have not been able to find among thenumerous specimens from Lake Superior. A single specimen of this form of themale was, however, sent to me by Dr. Stimpson under the manuscript name ofGammarus Hoyi, while two specimens of the female were sent as Gamrnarui< brevistylis.These are undoubtedly the same as the Gammarius Ilot/i and brerixtilKs mentioned,without descrii)tion, by Dr. P. R. Hoy. ?PONTOPOREIA FILICORNIS Smith.Gammarus fiUcornis Stimpson, MSS.Pontoporeiafilicornis Smith, Rep. IT. S. Fish Com., 1872-73 (1874), p. 649.Professor Smith's description follows:Male.?Outline of the head very much as in P. affinis. Eyes about as large as inthat species, slightly elongated, black. Peduncle of the antennula reaching nearlyto the distal end of the penultimate segment of the peduncle of the antenna, aboutas long as the head and the first segment of the thorax together; first segment largeand thickened; second half as long as the first; third slightly more than half as longas the second. Flagellum greatly elongated and very slender, reaching nearly to thetip of the abdomen, and composed of thirty-three segments, of which the proximal arelonger than broad, while they increase in length distally, until, near the tip, theyare many times longer than broad, and exceedingly slender. The upper side of theflagellum is nearly naked, only the alternate segments being furnished with twominute seta' near the distal extremity, while the under side of each segment is armeddistally with minute setae, and most of the segments with one or several clavate(olifactory) papilLe, and many of the segments have in addition a peculiar trans-parent, shallow, bell-shaped appendage, raised on a very slender peduncle. Sec-ondary flagellum reaching to the fourth segment of the primary, and composed offour segments, of which the terminal one is very short. Penultimate segment of thepeduncle of the antenna about as long as the first segment of the peduncle of theantennula; ultimate segment slightly shorter; penultimate and antepenultimate seg-ments furnished with long, plumose hairs below and several fascicles of short, seti-form hairs above. Flagellum much longer than the flagellum of the antennula, veryslender, and composed of about fifty very elongated and somewhat flattened seg-ments, which have about the same proportions as in the flagellum of the antennula,and are furnished with the same kinds of appendages.Epimera of almost exactly the same proportions and form as in P. ajfinis, and thefirst four margined with plumose hairs in the same way. First pair of legs verynearly like those of P. affinh; the palmary margin even slightly more longitudinalthan in that species, continuous with the posterior margin, and armed with two smallobtuse spines near the tip of the closed dactylus in addition to the setiform hairf.Second pair of legs of the same form as in P. affinis, except that the palmary marginis slightly concave and a little oblique in a proximal direction; the posterior mar-gin furnished with fascicles of setiform hairs, as in that species, and armed close tothe palmary margin with three or four small obtuse spines. Third and fourth pairsof legs like those of P. affinis, except that the dactyli have each three setiform hairsnear the tip, being in this as in several other respects nearer P. Ifoyi. Fifth andsixth pairs of legs almost exactly as in P. affinis, except the posterior margin of thepropodus in the sixth pair is armed with three pairs of small spines. Seventh pair,of legs having a few small spines on the propodus, but otherwise as in P. affinis.Lateral margins of the first second, and third segments of the abdomen with theangles rounded; lateral margin of the third segment furnished with a line of severalsubmarginal, plumose setse near the anterior angle, and behind them armed with fivelarge and acute spines directed backward, of which four are in pairs near the middle of?Deep Water Fauna of Lake Michigan, Trans. Wis. Acad. Sci., 1870-72, p. 98-101.
NO. 1507. FRESH-WATER AMPHIPODA?WECKEL. 29theiiiaffjin, and one alone near the posterior angle; the posterior edges of the lateralexpansions of all three of the segments furnished with a few, widely separated, jVlu-mose hairs. Peduncles of the first and second pairs of caudal stylets reaching toabout the same point, a little beyond the extremity of the sixth segment of theabdomen; the outer rami slightly longer than the inner, and those of the second pairof stylets only a little shorter than those of the first. Rand of the posterior caudalstylets longer than in P. affiiiis; the outer rannis rather more than twice as long asthe peiluncle, narrow, and tapering to an obtuse tip, both edges furnished with longplumose hairs, and the outer edge with a sharp spine at tlie base of ea(;h hair. Tel-son slightly longer than broad, cleft half-way to the l)ase, and each lobe tipped withtwo short spinules and a plumose seta. There ai'e two of the peculiar papiiliformappendages on the sternum of the third, fourth, and fifth segments of the thoi'ax,as in P. Hoyi, but apparently none upon the second.
? Length from the front of the head to the tip of the telson, 6 mm.Of this species, I have seen l)ut a single specimen, which was dredged with thelast species in Lake Michigan in 40 to 60 fathoms, off Racine, l)y Pr. Stimpson, fromwhom it was received under the manuscript name of Gaminarusjilicornis.This species differs remarkably from all the heretofore-known species of Pontopo-re'mie, in the excessive elongation of the flagella of the antennula' and antennte, acharacter which might be regarded by some naturalists as of generic value. Thevery close agreement with P. affiiiis and IIoijl in all other parts of the animal, how-ever, seems to indicate a very close afiinity with those species, especially the latter;and as this one peculiarity is very likely only a sexual character of the old males ofthe species, I retain the species in the genus. The mouth-appendages seem to agreeperfectly with those of the species just mentioned. The singular armature of thelateral margins of the third segment of the abdomen is not peculiar to this species,but is almost exactly repeated in P. affinis, P. Hoyi, and the marine species, alreadymentioned, from the Gulf of St. Lawrence, and is prol)ably common to the genus,although it seems to have been overlooked till now.Family GAAIMARID^E.Genus EUCRANGONYX Stebbing.KEY TO SPECIES.A. Telson about as long as peduncle of third uropods, not deejily cleft, thirduropods short.B. Outer ramus of third uropods aliout twii-e the length of peduncle; innerramus very small, telson cleft one-third distance to base.C. Eyes small, elongated, with black pigment E. gracilis, p. 32CO. P^yes small without black pigment E. packardii, p. 35BB. Outer ramus of third uropods shorter than peduncle; inner ramus absent,peduncle produced distally on the inner side into a short, blunt process;telson cleft three-fourths distance to base. _ E. bifrirms, p. 33AA. Telson in the male elongated, equaling from one-fourth to one-third lengthof body, rounded distally; telson in female short, with a slight emargina-tion; third uropods with short outer ramus and rudimentary inner ramus,E. mncronatus, p. 29EUCRANGONYX MUCRONATUS (Forbes).Crangnmjx rnvcroiutluK Forhes, Bull. III. State Lab. Nat. Hist., No. 1, 1S7(), p. 6.
?
O. P. H.w, Am. Nat., XVI., 1882, p. 241.Enrrangonyx mueronains Stebbing, Trans. Linn. Soc. London, 2nd ser., VII,1899, p. 423.Bacirums mueronains W. P. Hay, Proc. U. S. Nat. Mus., XXV, 1902, p. 429.
30 PROCEEDINGS OF THE NATIONAL MUSEUM. VOL. XXXII.Blind. Head .sliohtly long-er than the first thoracic segment, con-cave on its anterior margin at the l)ases of the first antennjB and con-vex l)etween them. First antennae in the male half to two-thirds aslong as the body; in the female about half as long as the body; fiagel-lum four or five times as long as the peduncle and consisting in themale of about thirty-five segments, all but the first seven or eightof which are provided with slender olfactory clubs; secondary flagellumslightly longer than the first segment of the primary one (usuallyshorter in the female) and composed of two segments, the distal oneverv short. Second antenn;i' oidv a])out one-half as Ions.;' in th(> first
Fig. 2.?EUCRANGdNYX MUCRONATUS. ant. 1, FIRST antenna; CUlt. '2, SECOND antenna; gn., GNATHO-pod; T., telson; ur., uropod; gn. 1 9, gn. 2 $, and T. ? from a female; other figures from
pair; peduncle extending beyond that of the first pair; the two distalsegments elongated, about eijual in length; flagellum slightly shorterthan the peduncle and composed of about ten segments.In the male the carpus of the first gnathopods is broadly triangular,fully as l)road as the propodus; propodus about two-thirds as broad aslong; anterior and posti^rior margins only slightly convex, the poste-rior furnished with three small fascicles of long hairs; palm very o))-lique, slightly concave at the center, with a broad lamellar edge, fur-nished with two or three fascicles of long hairs, and about ten notched
NO. 1507. FRESH-WATEE AMPHIPODA? WECKEL. 31
spines, with usually two large spines near the center and three at thetip of the closed dactyl; dact}^ strongly curved, slightly longer thanthe palm. First gnathopods of the female with the propodus smaller,shorter, and broader distally than in the male.In the second gnathopods the carpus is slightl}' longer than l)road;much larger than in the tirst pair; propodus not much larger than inthe tirst pair and not so broad; the posterior and anterior marginsonly very slightly convex, the posterior furnished with more hairsthan in the tirst pair; palm longer than in the tirst pair, similarlyarmed, but having a few more spines; dactyl strongly curved, as long-as the palm. Propodus of the second gnathopods in the female not somuch elongated as in the male, but similarly armed.Coxal i)lates of the three posterior pera^opods oval, not large, withboth margins serrated.Abdominal segments rounded dorsally; the lateral margins and allthe angles broadly rounded.The tirst pair of uropods extend l)eyond the second pair, whichextend beyond the third; second pair not longer than the peduncle ofthe first; tliird pair short, not quite so long as the peduncle of thesecond pair; outer ramus ovate, truncate, about half the length of thepeduncle, provided distally and laterally with a few hairs; innerramus rudimentary, unarmed, a])Out one-third or one-fourth as longas the outer ramus.Telson in the male c} lindrical, very nuich elongated, equaling fromone-fourth to one-third the length of the l)ody, rounded distally, andfurnished with two clusters of long hairs. In the female the telson isshort, projecting beyond the tips of the third uropods, having a veryslight emargination, on either side of which there is a cluster of long,stout hairs.
_Length, 10-15 nuu.Collected in old wells at Irvington, Indiana; Champaign andNormal, Illinois.The telson in the female of this species is somewhat incised posteri-orly, l)ut less than is usual in this genus; in the male the telson isentire and greatly elongated, a feature which does not accord withthe definition of the genus as originally given by Stebbing. Theother characters of this species, however, are so much like those ofthe other members of this genus that it seems best to retain it in thegenus Eiicrangonyx^ to which it was assigned by Stebbing, ratherthan to place it in a new genus, Bactrurus^ as proposed by Hay.
32 PROCEEDINGS OF THE NA TIONAL MUSEUM. VOL. XXXII.EUCRANGONYX GRACILIS (Smith).Crangonyx gracilis Smith, Am. Jour. Sci., 3d ser., II, 1871, p. 45.'3; Prelim. Reporton Dredgings in Lake Superior, 1871, p. 1022; Rep. U. S. Fish Com.,1872-73, (1874) p. 654.?Fokbrs, Bull. 111. State Lab. Nat. Hist., No. 1, 1876,p. 6.?O. P. Hay, Am. Nat., XVI, 1882, p. 241.?W. P. Hay, Proc. U. S.Nat. Mus., XXV, 1902, p. 429.Eucrangonyx gracilis Stebbing, Tran.'^. Linn. Soc. London, 2d ser., VII, 1899,p. 423.Eyes small, elong-ated. First aiit(Mina' slig-ht!}' nioie than half aslong- as the body; peduncle with the tirst segment shorter than thesecond, which is about twice as long- as the third; flagelhim twice aslong' as the pedimcle and consisting' of al)Out thirty segments; second-ary flag'ellum about equal in length to the first segment of the primar}'
Fig. 3.
?
Eucrangonyx gracilis, Ann Akbor, Michigan, vu-p., maxilliped; for other inscrip-tions TO THIS AND SUCCEEDING FIGURES SEE FIG. 2. gn. '2'$, SECOND GNATHOPOD OF A SMALLfemale; g7i. 2" 9, second gnathopod of a large female. Maxilliped drawn to a largerSCALE than the OTHER PARTS.
one, composed of two segments, the terminal one slender and veryshort. Second antennte half as long as the first pair; ultimate andpenultimate segments of the peduncle nearlj^ equal; tiagelhnn some-what shorter than the peduncle and composed of eight to ten segments.Carpus of the first gnathopods in the male about as broad as long;the posterior margin produced but slightly and furnished with long-plumose hairs; propodus almost ([uadi-ate in outline, a little longerthan broad; palm convex, slightly ol>lique, and armed on each sidewith from ten to fifteen notched spines, which are e((uidistant fromeach other except at the posterior angle, where about five are crowdedtogether; each spine is furnished with a hair a short distance from thetip; the posterior margin slightly concave, with a few long hairs;
NO. 1507. FRESH- WA TER AMPHIPODA?WECKEL. 33dactyl stout, slightl,y curved, extendino- to tlie posterior margin ofthe palm.Carpus of tlie second gnathopods in the male like that of the tirstpair; propodus more elongated than in the first pair, lu'oadened dis-tally; palm oblique, somewhat arcuate, armed with fifteen or moi'espines similar to those of the first pair, but not crowded into a groupat the posterior angle; posterior margin of hand with five or sixgroups of hairs; dact} I strongly curved, as long as the palm. In thefemale the propodus of the second gnathopods is proportional!}^ moreelongated than in the male and does not increase in breadth distally.Third, fourth, and fifth perttopods with the coxal plates serrate, andwith small spines on both margins.Postero-lateral angles of the first three segments of the abdomenproduced, terminating in a small tooth.First pair of uropods projecting fiackward beyond the second andthird pairs, two rami equal and about two-thirds as long as the pedun-cle; second pair with rami equal, and equal in length to the peduncle;third pair short, reaching to the tip of the second pair, peduncleshort; outer ranuis about twice the length of the peduncle, innerramus ver}' small, slender, shorter than the peduncle, usuall}^ withoutspines or hairs. Telson slightly shorter than the peduncle of thethird uropods; about as broad as long, cleft about one-third the dis-tance to the base with two or three spines on the truncate extremitiesof the lobes.The specimens of this species differ greatly in size, varjnng in ma-ture specimens from T-IS mm., but the larger ones agree in all essen-tial features with the smaller.Dhtrihution.? Providence, Rhode Island; Ann Ai-))or and IsleRoyal, Michigan; Lake Huron; Lake Superior; Portage, Wisconsin;Champaign, Illinois; Irvington, Indiana; Delaware, Ohio; Nashville,Tennessee.This species can not belong to the genus Crangoni/x to which it wasassigned b}^ Smith. That genus as generally defined has the thirduropods uniramous and the telson entire. The new genus Eucran-gonyx^ in which this species has been placed bv Stebbing, is like Cran-gonyx in general, but with a small inner ramus to the third uropodsand with the telson emarginate.EUCRANGONYX BIFURCUS (O. P. Hay).Crangonyx bifurcus O. P. Hay, Am. Nat., XVI, 1882, pp. 145-146.Eyes oval, large, about twice as large as those of Eucrangonyxgracilis which this species resembles. First antennae not quite halfthe length of the body; third segment of the peduncle two-thirds aslong as the first or second segments, which are nearlj^ equal; fiagellumProc. N. M. vol. xxxii?07 3
34 PROCEEDINGS OF THE NATIONAL MUSEUM. VOL. XXXII.composed of about twenty-five elongated segments; 8econdai\y flagel-lum slightl}' shorter than the first segment of the primar}^ flagelhmi,consisting of two segments, the distal one short; entire antennae fur-nished with comparatively few short hairs. Second antennas slightlymore than half as long as the first pair; two distal segments of thepeduncle elongated, nearh^ equal in length, the third segment short;flagellum not so long as the two distal segments of the peduncle, com-posed of about six elongated segments; antennas furnished with a fewhairs and a number of large olfactory clubs.Carpus of the first gnathopods in the male short and broadly trian-gular, nearly as broad as the propodus; propodus slightly longer than
Fig. 4.?EUCRANGONYX BIFURCUS, MALE. MaCOX, MISSISSIPPI.broad; lateral margins couA^ex, the posterior provided with four orfive small fascicles of hairs; palm oblique, armed on each side withabout twelve spines and a few hairs; dactyl strongly bent near thebase, as long as the palm. Second gnathopods with the carpus andpropodus YQYY similar to those of the first pair, but more elongated;the palm more oblique, armed with about fifteen notched spines;posterior margins furnished with six small fascicles of hairs.Last three pairs of pera?opods with the coxal plates oval, bothmargins serrate and armed with small spines.Abdominal segments rounded dorsally; postero-lateral angles of thefirst three segments produced backward into a tooth.
NO. 1507. FRESH-WATER AMPHIPODA?WECKEL. 35First uropods projecting backward beyond the second pair, and thesecond beyond the third; third pair consisting of a peduncle and asingle small ramus; peduncle more than twice as long as broad, fur-nished with a few spines, and produced distally on the inner side intoa short, blunt process furnished at the tip with a small spine; ramusabout two-thirds the length of the peduncle, provided with a fewspines. Telson cleft three-fourths the distance to the base, projectingbeyond the third uropods, lobes diverging toward the tip, each armeddistally with three or four spines.Length, 9 mm.Type locality.?Macon, Mississippi,This species differs from the other members of the genus Eucran-gonyx in the absence of the rudimentary inner ramus of the thirduropods. In all other characters, however, it so closely resembles thespecies of this genus that it seems best to retain it in Eucrantjonyxrather than to place it in a new genus.EUCRANGONYX PACKARDII (Smith).Crangonyx packardii Smith, 3Ieiii. Nat. Acad. Sci., I\', 1888, pp. 34-36, pi. v,fig?. 1-11.The following description, with iigures, is taken from MemoirsNational Academy of Sciences, where the species is described byProf. S. I. Smith as Crangonyx pachirdiLThis species iH so closely allied to Crangonyx gracilix that it might readily be mis-taken for it Avere it not for the peculiar structure of the eyes. The eyes of C. gracilisare composed of a few facets, and are abundantly supplied with black pigment. Inall the specimens of C. packardii which I have seen the eyes are observable withdifficulty, the black pigment being wholly wanting. The specimens received atfirst were very badly preserved, and I thought the absence of the pigment mightbe due to this fact; but subsequent examination of more perfect specimens showsthat this can not be the case, and that the eyes are in life undoubtedly w" holly with-out black pigment. The eyes are scarcely, if at all, observable in the ordinary alco-holic specimens, but when rendered translucent by immersion in glycerin thestructure of the facets is distinctly observable, as shown in fig. 5. As observed byDoctor Packard, the flagella of the antennuhE of C. gradiis are a little shorter,and usually contain four or five segments less than C. gracilis, but this is an uncer-tain character, and some specimens of C. gracilis from Lake Superior actually haveonly one or more segments than the subterranean species. In the antennte there areno constant differences. There are some very slight differences in the first and sec-ond pairs of legs; especially in the females, but not greater than usually exist in theindividuals of a single species in allied genera, and any large series of specimenswould undoubtedly show^ all the intermediate forms. In the third to the seventh pairof legs there is a constant difference in all the specimens examined, the spines beingmore numerous, longer, and more slender in C. gracilis. The spines upon the firstand second caudal stylets are a little shorter and more obtuse in C. puck-ardii thanthey usually are in C. gracilis; otherwise there is no difference in the caudal styletsand telson.These differences are all such as very naturally lead to the supposition that thissubterranean form has been derived from the C. gracilis at no very remote period,
36 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.. XXXII.although thia supposition may well be held in reserve until we have a more com-plete series of the subterranean species for comparison. C. gracilis occurs as farsouth as Grand Rapids, ^licliigan, whence we have received specimens from Mr. N.Coleman, and it very likely occurs in the same region as C. packardii.The figures are all from Professor Packard's original specimens, collected fromwells in Orleans, Indiana, by Dr. Moses Elrod. Only one of the.'-e has the bodyentire; this is a female, 5.5 mm. long, and from whicli figs. 5 to 8 were made. A
Fig. 5.?EUCRANGONV.V PACKARDII (After Smith.) ah, terminal portiox of the abdomenWITH THE UROPODA.large specimen, a female about 7.5 mm. long, unfortunately wanting most of theantennulee and antennse, collected from a well at New Albany, Indiana, by Dr. JohnSloan, was sent to Doctor Packard for examination.Genus NIPHARGUS W. P. Hay.NIPHARGUS ANTENNATUS (Packard).Crangonyx antennatus Fackard, Am. Nat., XV, 1881, p. 880.Eucrangonyx antennatus Stbbbihg, Trans. Linn. Soc. London, 2d ser., VII, 1899,p. 423.Niphargus antennatus W. P. Hay, Proc. U. S. Nat. Mus., XXV, 1902, p. 429.Eyes composed of a few slightly pigmented facets or wanting alto-gether, when present of no definite shape. First antennte more thanhalf as long as the body; first and second segments of the pedunclenearly equal in length, slightly longer than the third; flagellura aboutthree times as long as the peduncle, composed of about twentj^-fivesegments in the female; the number of segments in the flagellumvaries in the two sexes according to Hay's description, there being-more in the female than in the male; secondary flagellum longer thanthe basal segment of the primar}^ one, compo.sed of two segments, thedi.stal one very short; each segment furnished distall}' with a few longhairs. Second antennae not half so long as the first pair; flagellumshort,- not greatly exceeding the third segment of the peduncle andcomposed of about eight segments.
NO. 1507. FRESH- WA TER AMPHIPODA? WECKEL. 37In the female the first g-nathopods have thecarpu.s short, triangular,as broad as the propodus; propodus less than twice as long as In-oad;anterior margin slightly convex; posterior margin straight or nearlyso, furnislied with a few fascicles of long hairs; palm slightly obliqueand convex, furnished on each side with a few hairs, six or eightshort notched spines, and one long stout spine at the tip of the closeddactyl; dactyl stout and curved. In the male, from Hay's descrip-tion, the propodus is larger, st()ut(n% and more strongly armed thanin the female.Propodus of the second gnathopods of the female elongated, twiceas long as broad, narrowed distally; lateral margins slightly convex,the posterior one furnished with a few fascicles of long hairs; palm
m. 3Fig. 6.?NiPHARCius antennatus, female. Ntcka.jack Cave, Shei.lmound, Tennessee.
oblique, convex, armed as in the first pair, but with more numerousspines. In the male, according to Hay's description, the secondg-nathopods are larger than in the female and have the propodus pro-portionall>^ more developed.The abdominal segments are rounded above and without spines.The uropods project backward about an equal distance; the terminalpair are slightly longer than the peduncle of the second pair, with theouter ramus well developed, composed of two segments, the terminalone very short, and armed distally with a few spines, the first segmentfurnished on both margins with four or five spines; inner ramus short,rudimentary^ and furnished with a few hairs at the tip.
38 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxii.Telson emarginate, the emargination reaching to the middle ornearl}^ to the base; each lobe furnished at the truncate tip with threeor four stout spines.Length 10 mm.This species has been collected at various places within NickajackCav^e, Shellmound, Tennessee.The genus Niphargus^ to which this species has been assigned byHay, differs from Crangonyx in the telson, which in the latter genusis entire and in the former divided, and in the third uropods, which inCrangonyx are uniramous and in Nlphargus biramous with the outerramus biarticulate. It differs from Eucrangonyx^ in which genus ithas been placed by Stebbing, in the outer ramus of the third uropods,which in Encrangonyx is uniarticulate and in Nij)hargiis biarticulate.Genus GAMMARUS Fabricius.KEY TO SPECIES.A. Telson cleft to the base.B. Eyes present.C. Third uropods with basal segment of outer ramus elongated, terminal segmentshort, styliform; inner ramus half or more than half as long as the outer;both rami furnished with long plumose hairs.D. Inner ramus of third uropods at least three-fourths as long as the outer;nearly as broad as the basal portion of the outer ramus.PI Terminal segment of outer ramus of third uropods with long plumosehairs on lateral margins; secondary flagellum composed of two to foursegments G. limnxus, p. 42EE. Terminal segment of outer ramus of third uropods without longplumose hairs on the lateral margins; secondary flagellum composedof five or six segments G.fasciatus, p. 40DD. Inner ramus of third uropods only slightly more than half as long as outerramus; much more slender than the basal portion of the outerG. propinqims, p. 46CC. Third uropods with the basal segment of the outer ramus elongated, withfascicles of a few long hairs on lateral margins; terminal segment short;inner ramus rudimentary, shorter than the peduncle. . . G. ramellus, p. 38BB. Eyes absent G. accus, p. 47AA. Telson not cleft to the ))ase G. jmrpurascens, p. 45GAMMARUS RAMELLUS, new species.Eves small, reniform; their distance apart more than twice theirdiameter. First antennae not quite half so long as the bod}^; first andsecond segments of the peduncle about equal and nearly twice as longas the third; flagellum composed of twent}' to twenty-five segments;secondary flagellum slightly exceeding the terminal segment of thepeduncle in length and consisting of three or four segments. Secondantenniv more than half as long as the first pair; peduncle longer thanin the preceding pair; flagellum composed of eight to twelve segments.First gnathopods of the male with the carpus triangular, as broad aslong; propodus slightly longer than broad; lateral margins convex;
NO. 1507 FRESH-WATER AMPHIPODA? WECKEL. 39the posterior one furnished with three or four small fascicles of hairs;palm oblique, with two emarginations, and furnished with a few longhairs and a group of six or eight spines at the posterior angle; dactylstrongly curved, as long as the palm.Carpus in the second gnathopods more elongated than in the firstpair; propodus oblong; lateral margins only slightly curved, the pos-terior one furnished with six oi seven fascicles of long hairs; palmless oblique than in the tirst pair, furnished with a row of long hairsat the base of the lamellar edge and six or seven spines at the roundedposterior angle; dactyl not so strongly curved as in the tirst pair,longer than the palm.
Fig. 7.?Gammaki's eamellus, male. Point Arena, California, m., mandible, drawn to aLARGER SCALE THAN THE OTHER PARTS; Ob. d., DORSAL SIDE OF THE LAST SEGMENTS OF THEABDOMEN.Coxal plates of the last three peraeopods with the posterior marginsserrated and furnished with short spines.Postero-lateral angles of the first three abdominal segments pro-duced but little and terminating in a small spine. The last three or fourabdominal segments are furnished dorsally on each side with a fascicleof two or three spines and a few short hairs.First uropods projecting backward slightly beyond the peduncleof the third pair; third pair with the peduncle short, as broad as long,armed distally with a few hairs; outer ramus elongated, biarticulate,the basal segment long, tapering distally, armed laterally with fasci-cles of a few long hairs and one or two large spines; terminal segmentver}^ short, tapering distally and terminating in a tuft of long hairs;inner ramus rudimentary, slightly more than half as long as the
40 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxn.peduncle, furnished at the tip with a spine and a few hairs, and a fewshort hairs on the inner margin.Telson cleft to the base, the lobes rounded distall}^ and furnishedwith a few hairs and a few small spines.Length, 10 mm.Type locality.?Point Arena, California. Specimens from Portland,Oregon, were also examined. These were larger and stouter thanthose from California.Type.?C^i. No. 32841, U.S.N.M.GAMMARUS FASCIATUS Say.Gammarus fasciatus Say, Jour. Acad. Nat. Sci. Phila., I, 1817, p. 374.
?
De Kay,Zool. of New York, VI, 1844, p. 37.
?
Bate, Cat. Amphipodous CrustaceaBritish Mus., 1862, p. 210, pi. xxxvii, fig. 6.?Smith, Rep. U. S. Fish Com.,1872-73 (1874), p. 653.?Forbes, Bull. 111. State Lab. Nat. Hist., No. 1,1876, p. 6.E3-es reniform; their distance apart slightl}^ exceeding their diame-ter. First antennae with the first and second segments nearly equalin length and twice as long as the third; fiagellum twice as long as thepeduncle and composed of about thirt}^ segments; secondary iiagellumas long as the second segment of the peduncle and consisting of five orsix segments. Second antennae about two-thirds the length of thefirst; peduncle exceeding that of the first pair, with the two distalsegments elongated and nearly equal; fiagellum slightly shorter thanthe peduncle and composed of about fifteen segments.In the male the carpus of the first gnathopods is triangular, slightlylonger than broad; propodus not quite twice as long as broad, muchnarrowed distally; palm slightl}" concave, very oblique, continuouswith the posterior margin, with a narrow lamellar edge, armed witha few long hairs, near the middle of the inner side with a stout spine,at the base of which there is a fascicle of long hairs, and three orfour smaller spines near the tips of the closed dactyl; posterior marginwith four or five fascicles of long hairs; dactyl strongly curved andhalf as long as the propodus. Propodus of the female much smallerthan that of the male, only slightly narrowed distall}^; palm not sooblique as in the male, furnished with a few long hairs and a fewspines at the tip of the closed dactyl, but without a spine at the middleof the palm.Propodus of the second gnathopods in the male as long as in thefirst pair, broadest distally; lateral margins nearly parallel; the poste-rior margin furnished with numerous fascicles of hairs; palm oblique,but not so much so as in the first pair, concave in the middle, armedwith a few short hairs at the base of the broad lamellar edge, a stoutmedian spine, and three or four smaller ones at the tip of the closeddactyl; dactyl not so strongly armed as in the first pair, as long as
FRESH- WA TER AMPHIPODA? WECKEL. 41the palm. Second g-nathopods in the female much smaller than in themale; carpus elont^ated, as broad as the propodus; propodus narrow,twice as long as broad; lateral margins parallel; palm straight, trans-verse, without a lamellar edge and devoid of a median tooth.Coxal plates of the last three pensopods with the posterior marginserrate.The infero-posterior angles of the second and tiiird abdominal seg-ments are slightl\" produced, forming- less than a right angle. Thefourth, tifth, and sixth segments are slight!}^ angulated dorsally, each
Fig. 8.?Gammarus fasciatus. gn. 1 $, gn. 2 ?, first and second gnathopods of a female,OTHER FIGURES FRO.M A MALE; mil.. MANDIBLE, AND Vlxp., MAXILLIPED, DRAWN TO A LARGERSCALE THAN THE OTHER PARTS; ah. d., DORSAL SIDE OF THE TERMINAL PART OP THE ABDOMEN.bearing' three fascicles of well-developed spines, the median one ofwhich on the fourth and fifth segments is raised on a distinctprotuberance.First uropods exceeding the second pair in length and projecting-backward slightly beyond the peduncle of the terminal pair; outerramus of the terminal pair composed of two segments, the first seg-ment elongated, furnished on both margins with many long plumosehairs and on the outer margin with several small obtuse spines; thedistal segment very short, styliform, and furnished at the tip withseveral hairs, but having no hairs on the lateral margins; inner ramus
42 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxii.not quite so long as the first segment of the outer ramus, armed with afew spines on the inner margin and numerous long plumose hairs onboth margins.Telson cleft to the base, the lobes furnished distally with two orthree spines and a few hairs.Length, 10-15 mm.Distrihution.?Hudson River and Niagara Falls, New York; AnnArbor, Michigan; Lake Superior; Delavan and Lake Geneva, Wis-consin; Havana, Illinois; Burlington, Iowa; Redfoot Lake, Tennessee;Brookside, West Virginia; Washington, District of Columbia; St.Johns River, Florida.GAMMARUS LIMNiEUS Smith.Gammarus lacustris Smith, Aiiier. Jour. Sci., 3<l ser., II, 1871, p. 453; Prel.Eept. on Dredgings in Lake Superior, 1871, p. 1023.Gammarus limna'us Smith, 7th Kept. U. S. Geol. Survey, 1874, p. 609; Kept. U. S.Fish Com., 1872-73, p. 651, 1874.Gammarus robustus Smith, 7th Kept. U. S. Geol. Survey, 1875, p. 610.Eyes small, slightly elongated. First antennae about as long as thethorax; first and second segments of the peduncle nearly equal, twiceas long as the third; flagellum composed of about thirty elongatedsegments, which are furnished with few hairs; secondary flagellumcomposed of two to four segments, of which the terminal one is short.Second antennae slightly shorter than the first; peduncle exceedingthat of the first pair with the two distal segments nearly equal inlength; flagellum shorter than or nearly equal to the peduncle, con-sisting of about twelve segments.First gnathopods of the male with the carpus broadl}^ triangular, asbroad as the propodus; propodus a little less than twice as long asbroad^ much narrowed distall}"; palm very oblique, continuous withthe posterior margin, and having a lamellar edge furnished with a fewlong hairs, and two long obtuse spines near the middle, and three orfour smaller spines on each side near the posterior angle; dactylstrongly curved and about one-half as long as the propodus; lateralmargins convex, the posterior one furnished with a few long hairsand several short, obtuse spines. In the female the propodus is con-siderably smaller and proportionally shorter than in the male; palmoblique, with a narrow lamellar edge continuous with the posteriormargin, furnished with a few long hairs, and sev^eral spines on eachside of the tip of the closed dactyl; posterior margin furnished withseveral short spines, and numerous long hairs arranged in severalfascicles.Carpus of the second gnathopods in the male slightly longer than inthe first pair; propodus as long as in the first pair, slightl}^ broadestdistally; lateral margins nearl}^ parallel, posterior one furnished withfive or six fascicles of long hairs; palm slightly oblique, concave at the
NO. 1507. FEESH-WATEE AMPHIPODA? WECKEL. 43middle, with a broader lamellar edg-e than in the first pair; armed witha long, obtuse spine and a fascicle of long hairs at the center, two orthree small spines at the posterior angle, and a row of short hairs atthe base of the lamellar edge. In the female the carpus and propodusare much smaller than in the male and proportionality more elongated;carpus fully as broad as the propodus; propodus twice as long asbroad, palm straight, nearl}" transverse, without the median spine.Posterior margin of the coxal plates of the last three perwopodsserrate.Fourth and fifth abdominal segments rounded above, each bearingthree fascicles of small spines: the sixth segment with two lateral
Fig. 9.?Gam>iari"s limn.ecs. Ann Arbor, MirniGAX.fascicles and two or three short median hairs. Infero-posterior angleof the first abdominal segment rounded, in the second and third pro-duced posteriorly and forming an acute angle.First pair of uropods exceeding the second pair in length andextending slightly beyond the peduncle of the third pair. Outerramus of the terminal pair narrow, the basal segment furnished onboth margins with many long plumose hairs, and on the outer marginwith a few short spines; terminal segment short, styliform, lateralmargins as well as tip furnished with long plumose hairs; innerramus slightly shorter than the l)asal portion of the outer, with both
44 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxii.marg-ins furnished with long- plumose hairs and the inner with a fewspines.Telson cleft to the base, each division furnished with three or fourshort spines and several long- hairs.Length, 15 ram.Aroostook Count}', Maine; Caledonia, New York; Marquette,Ann Arbor, and Isle Ro3^aI, Michigan; dredged in Lake Superior;Fort Wing-ate, New Mexico; Wahsatch Mountains; Shoshone Falls,Idaho; Salt Lake City. Utah; Flat Head Lake, Montana; YellowstoneNational Park.This species closely resembles Gammarus fasciatus^ from which itcan be distinguished most easily by the presence of long- plumose-hairson the lateral margins of the terminal segment of the outer ramus ofthe third uropods. The other characters in which these two speciesdiffer can be appreciated only by a careful comparison of specimensof the two species. The antennae in Gammarus limnaeua are furnishedwith fewer hairs than in Gammarus fasciatus and the secondaryflageUum consists of two to four segments (in Gammarus fasciatusfive or six). The propodus of the first gnathopods of the male ofGammarus limnxas is broader at the proximal end than in Gammarusfasciatus; the lateral margins are more convex, and the posteriormargin is furnished with spines, which, if present in Gammarus fas-ciatus^ are very small. In the female the propodus of the firstgnathopods of Gammarus lininseus is much narrower distally and thepalm is more oblique than in the other species. In Gammarus lini-nx.us the spines on the abdominal seg-ments are smaller and themedian ones are not raised on distinct protuberances as they are inGammar us fasciatus.After a careful comparison of the description and figures of Gam-marus limnmus Smith and Gammarus rohustus Smith I have con-cluded that they are the same species. The two species correspond inall characters except the gnathopods, and the differences which arethere pointed out are not sufficiently marked to justify distinguishingthe two forms as separate species. In Gammarus limiuvus the pro-podus of the first gnathopods is longer in proportion to the width, thepalm is more oblique, and the posterior marg-in is armed with morespines than in Gam.marus rohustus. In the second g-nathopods thepropodus of Gammarus limndeus \s longer than in Gammarus rohustus.These are differences which appear in specimens of different sizes.From the lengths of the two species which Smith gives {Gammaruslimnxus l^-'iO n\n\. and Gammarus rohustus lO-lo mm.) he evidentlyexamined larger specimens of Gammarus lifnna^'us than of Gammarusrohustus, which fact would account for the differences in the gnatho-pods which he found.
FRESH- WA TEE AMPHIPODA? 1 1 ECKEL. 45GAMMARUS PURPURASCENS W. P. Hay.Gammarus purpurascens W. P. Hay, Proc. U. S. Nat. Mus., XXY, 1902, p. 433.Eyes large, reniform; larger than in Gammarus liw.nseim Smith,which species Gammarus jyia'puraxcens resembles. First antennalonger than in Gammarus llmnseus^ about half the length of the bod}';first and second segments of the peduncle nearly equal in length; thirdsegment about half as long as the second; flagelkim composed oftwenty-five to thirt}' segments; secondary fiagellum consisting of fou?"
Fig. 10.?Gammakus purpurascexs, male. Nickajack Cave. ah. d., dorsal side of the termi-nal SEGMENTS OF THE ABDO.MEN.segments, the distal one ver}' short and tapering. Second antennoeslightly more than half as long as the first pair; peduncle exceedingthat of the first pair in length, the two distal segments elongated,nearly equal; fiagellum slightly shorter than the peduncle, composedof about ten segments, each of which bears a small olfactory club.First gnathopods of the male with the carpus ])roadh' triangular,nearly as broad as the propodus; propodus broad, narrowed distall}^;
46 PROCEEDINGS OF THE NA TIONAL MUSEUM. vol. xxxii.pahn very oblique, continuous with the posterior margin and furnishedwith a few short hairs on each side, a fascicle of long ones at the cen-ter, and a few spines at the tip of the closed dact}^; the posterior andanterior margins of the propodus convex, the posterior one providedwith a few fascicles of long hairs and one or .two small spines nearthe palm; dactjd strongl}^ curved, half as long as the propodus.Second gnathopods with the propodus rectangular, more than half asbroad as long; lateral margins only slightly convex, the posterior withabout eight fascicles of long hairs; palm oblique, furnished as in thefirst pair except that no spines extend upon the posterior margin; dac-t}^ not so strongly curved as in the first pair, extending slightly beyondthe palm.Basal plates of the third, fourth, and fifth perj\?opods narrow, broad-est proximally, with margins^'serrate and furnished with a few spines.Three posterior segments of the abdomen rounded dorsally, eachfurnished with three small fascicles of spines; the spines of the medianfascicles smaller than those of the lateral ones.First pair of ul-opods slightly exceeding the second pair in length,projecting somewhat beyond the peduncle of the third; third pair witha short peduncle onl}' slightly longer than broad; the outer ramus con-sisting of two segments, the basal one long and narrow with four orfive spines on the outer mai'gin and many long plumose hairs; termi-nal segment very short, styliform, and furnished distally with a few.long hairs; inner ramus about three-fourths as long as the basal por-tion of the outer, with many long plumose hairs and usually a fewspines on the inner margin. Telson divided almost to the base, witha spine and a few hairs on the outer margin, and three or four spinesand a few short hairs distally.Length, 10 mm.The t3^pe specimens of this species w^ere collected at the mouth ofNickajack Cave, Shellmound, Tennessee. Specimens of a smaller sizehave also been found in a spring at Rossville, Georgia.GAMMARUS PROPINQUUS W. P. Hay.Gainmarus prophujaus W. P. Hay, Proc. U. S. Nat. INIus., XXV, 1902, p. 224.This species is ver}^ similar to Ganvmarus fasciattis Say, but difl'ersfrom it in the following characters: The secondary flagellum consistsof from two to five segments; in Gaviinm'nsfasciatusiYom. five to six.The palm of the first gnathopods of the male in G. fasdatus is moreirregular and more concave. In the second gnathopods it is moreconcave and slightly more oblique than in Gaminarus j^i'opmquus.The best distinguishing characteristic, however, is in the third uropods,the inner ramus. of which in Gammarus propinquus is much moreslender than the outer and half or onlv slightly more than half as long.
FRESH- WA TER AMPHIPODA? WECKEL. 47In GammarKS fasciatus the inner ramus is only slightly narrower thanthe outer and at least three-fourths as long.Length, 10-15 mm.Specimens of this species have been collected in several localities inregions about Mammoth Cave, Kentuck}', the type coming from a
Fig. 1L?(iAMMARus pROPiNyrrs, male. Mammoth Cave, Kentucky.
spring about 2 miles north of this cave; collected also in springs in theOzark Mountains, Missouri, and in West Virginia.GAMMARUS Ci^iCUS, new species.Eyes absent. First antenna more than half as long as the body;second segment of the peduncle slightly longer than the tirst andabout three times as long as the third; liagellum composed of twentyto thirty elongated segments, each bearing a few short hairs at thedistal end; secondary iiagellum reaching slightly beyond the thirdsegment of the primary flagellum, composed of four segments, thedistal one short and furnished with long hairs.Second antenna^ are about two-thirds as long as the tirst pair withthe peduncle extending fai- beyond that of the first pair; ultimatesegment of the peduncle only^ slightly longer than penultimate whichis greatly elongated and about equal in length to the antepenultimate;flagellum composed of about twelve segments, which are shorter thanthose of the first antennae and furnished with more hairs.The carpus of the tirst gnathopods of the, male is triangular andelongated with the anterior margin furnished with a few long hairsand numerous short ones; propodus narrower than the carpus, twiceas long as broad, with the anterior margin concave, armed sometimeswith a fascicle of hairs, the posterior margin convex, and the palm
48 PROCEEDINGS OF THE NATIONAL MUSEUM. VOL. XXXII.almost straight, slightl}^ convex, and armed with four or five spinesand a few short hairs; dact^yl as long as the palm and fitting it closely.Second gnathopods with a carpus broader than in the first pair butsimilarly armed; propodus not so l)road as the carpus, about twice aslong as broad and larger than in the first gnathopods; posterior mar-gin almost straight; anterior margin slightly convex and usually fur-nished with one or two fascicles of hairs; palm very oblique, slightlyconcave at the center, armed with five or six spines at the tip of theclosed dactjd, and one or two spines and a few short hairs on themargin; dactyl strongly curved, as long as the palm.
Pig. 12.
?
Gamm.\rus cecus, male. Modesta Cave, near Canas, Cuba.Both margins of the coxal plates of the third, fourth, and fifthpeneopods are serrate and furnished with spines, those on the anteriormargin being smaller than those on the posterior.Postero-lateral angles of the third and fourth abdominal segmentsare produced backward and end in a blunt tooth. The last two orthree abdominal segments are furnished dorsally with a few shortspines.The first uropods project slightly beyond the second pair. In bothpairs the rami are about equal in length and slightly longer than thepeduncle. The third uropods were broken off in the few specimenswhich I had for examination. Telson cleft to the base, armed distallywith a few short spines.
NO. ir.07. FRESH- WA TER AMRllIPODA? WECKEL. 4 V>Length, 10 mm.Collected 1)}' Dr. C. H. Eig-enmaiui among the roots of a tree,Modesta Cave, near Cana.s, Cuba.Ti/pe.?Cat. No. 3268i?, U.S.N.M.Genus CRANGONYX Bate.KEY To Sl'ECIES.A. Telson nearlyas broad as loiij^, with a slight eiuargiuat ion (Ustally; sec-oiid uropodswith rami equal < '. /it reus, ]>. 49.A A. Telson entire, two-tliirds as broad as long; rami of second uropods unequal, theouter ramus about half as long as the inner ,.<\ tennis, p. 50.CRANGONYX VITREUS (Cope).
Styyubroiiiitx vitreus Cope, Amer. Nat., VI, 1872, }>. 422; ord and 4th AnnualRepts. of Geol. Survey of Indiana, 1872, p. 181.Crangonyx vitreus Packard, 5th Annual Rept. Peabody Academy of Science,Salem, 1873, p. 95.?Smith, Rept. U. S. Fish Com., 1872-73 (1874), p. 656;Amer. Jour. Sci., 3rd series, IX, 1875, p. 476.Blind. First antenna' more than halt" as long- as the t)od\'; thethree segments of the peduncle nearly equal in length; Hagellumcomposed of about fifteen segments; secondary flagellum reachingbeyond the first segment of the primary one, consisting of two seg-ments. Second antennte slightly less than half as long as the tirstpair; peduncle extending I)eyond that of the tirst pair with the twodistal segments elongated and nearly equal; flagellum very short,scarcely exceeding the distal segment of the peduncle in length andcomposed of six segments.Propodus of the first gnathopods in the male slightly longer thanbroad; lateral margins convex, the posterior margin furnished with afew long hairs arranged in four or five fascicles; palm oblique, con-vex, armed with four or tive groups of long hairs and nine or tennotched spines, one of which at the posterior angle is nmch longerthan the other; dactyl stout. In the second gnathopods the propodusis more elongated and more narrowed distally than in the first pair;palm ver}^ oblique, convex, armed as in the first pair l)ut bearing afew more spines.Both margins of the coxal plates of the last three pairs of perao-pods are serrate and armed with small spines.First and second uropods extending beyond the telson; the thirdpair projecting onl}' about half way to the tip of the telson; peduncleabout as broad as long; the single ramus very small, about one-third aslong as the peduncle, slightl}- longer than broad, and furnished distallywith two short hairs. Telson nearly as broad as long, a very slighttriangular emargination at the broad distal end, on either side ofwhich are four or five spines.Proc. N. M. vol. xxxii?07 4
50 PROCEEDINGS OF THE NATIONAL MUSEUM. VOL. XXXII.Length, 5 mm.Specimens from Mammoth Cave, Kentucky, and Orleans, Indiana,In this species the telson has a slight emargination, a feature whichdoes not accord with the definition of this genus. The other charac-ters of the species, however, are so much like those of the recognizedmembers of the genus Crcmgonyx that it seems best not to assign it toa new genus. The specimens upon which this description is basedwere sent to me from the United States National Museum under thename Orangonyx vitreus Packard. The specimens agree in the mainwith Packard's description of this .species; however, he has described
Fi(_i. 13.?Crangonyx vitreus, male. Mammoth Cave, Kentucky.his specimens so imperfectly that I can not be certain that the speci-mens which I have described as C. vitreus belong to the same speciesas Packard's. I have followed Packard in quoting Stygohromus vitreusCope, as a synonym of Crangonyx vitreus but from Cope's ver}^ shortdescription of his specimens it is very doubtful if they belong to thesame species or even genus as Packard's Crangonyx vitreus.CRANGONYX TENUIS Smith.Crangonyx tenuis Smith, Kept. U. S. Fish Com., 1872-73 (1874), p. 656.I have had no specimens of this species for examination, but quotethe original description:A slender, elongated species, with very low ejiimera, reseinblinu- more in 'orm thespecies of Niphargm than the typical species of ( 'rangonyx.
NO. 1507. FRESH-WATER AMPHIPODA?WECKEL. 51Eyes not observable in alcoholic specimens. Secondary flagelluni of the anten-nulae very small, composed of two segments, of which the terminal is very short.First and second pairs of legs diffei'ing but httle in the two sexes. First pairstouter than the second, and with the palmary margin of the propodus much moreoblique; the palmary margin of the propodus of both pairs, and in both sexes, armedon each side with a series of stout, obtuse si:)ines, with a notch and a cilium near thetip.First three segments of the abdomen longer than the last three of the thorax;fourth, fifth, and sixth together scarcely longer than the third. Caudal stylets allextending to about the same point. First pair with the rami subequal, scarcely halfas long as the peduncle. Peduncle in the second pair reai'hing a little beyond thepeduncle of the first pair; the rami very unequal, the outer only half as long as theinner. Posterior pair scarcely as long as the telson; the single terminal segmentvery small, and tipped with four or five setiform spinules. Telson two-thirds asbroad as long, tapering very slightly toward the entire and slightly arcuate posteriormargin, which is armed with about ten slender spinules.In the largest male seen, 13.5 mm. in length (excluding the antennae), the anten-nulge are about 5 mm. long, the flagellum being twice as long as the peduncle andcomposed of about twenty-two segments, wliile the antenn;e are stout, fully 6 mm.long, and the flagellum as long as the peduncle, and composed of fifteen segments.All the females and most of the males which I have seen are much smaller, being6 mm. to 8 mm. in length, and in these the antennula? are longer than the antennae;and the flagellum of the antennulie is composed of sixteen to nineteen segments,while that of the antennae has only eight to ten.The only specimens which I have seen were found in wells at Middletown, Conn.,and were sent to me by Mr. G. Brown Goode.Genus STYGONECTES M/ . P. Hay.STYGONECTES FLAGELLATUS (Benedict).Crangonyx JiageUatus Benedict, Proc. U. S. Nat. Mus., XVIII, 1896, p. 616.Crangonyx boiversll Ulrich, Trans. Amer. Microscopical Soc, XXIII, 1902, p. 85,pi. XIV.Slygonectesflagellatiig W. P. Hay, Proc. IT. S. Nat. Mus., XXV, 1902, p. 429.This blind species is closely related to Eucrangonyx macronatus(Forbes). It is, however, larger and much stouter. The head, whichis slightly longer than the tirst thoracic segment, is proportionallynarrower and more concave at the insertion of the upper antennte thanin Eucrcmgonyx inucro7iatu)<.First antenna? as long as the body; first two segments of the peduncleabout equal and three times as long as the third; flageHum composedof forty to sixty segments; secondary flageHum extending slighth'bej'ond the basal segment of the primary one, composed of two seg-ments, the first about four times the length of the second. Secondantenna slightly less than half as long as the first pair; peduncleexceeding that of the first pair in length; two distal segments muchelongated and nearly equal; flagellum not quite so long as the peduncle,composed of twelve to eighteen segments.Propodus of the first gnathopods in the male about two-thirds asbroad as long; lateral margins only slightly convex, the posterior
52 PROCEEDINGS OF THE NATIONAL MUSEUM. VOL. xxxir.margin furnished with three or four fascicles of spines; palm veryoblique, slightly convex, with a nan-ow lamellar edge furnished oneach side with a row of about twentj^-four bifurcate spines, two orthree of which extend upon the posterior margin; dactyl stout andstrongly curved.In the second gnathopods of the male the propodus is about equalin size to that of the first pair, but with the palm less obli([ue and theposterior margin correspondingly longer.First uropods extend to the tips of the second pair, the peduncleslightl\' longer than the two rami; second uropods al)out two-thirds
Fig. 14.
?
Stygonectes flagellatus, male. San Makcos, Texas. The rudimentary inner ramusof the third uropod was broken in the specimen drawn.as long as the first ones; the first and second pairs reaching far beyondthe third pair (in Eucrangonyx m.ucronatun they extend only slightlybeyond the tip of the third pair); terminal uropods very small, abouttwo-thirds as long as the peduncle of the second pair; inner ramusrudimentary; the outer one about as long as the peduncle and fur-nished distally with a few spines. Telson entire, about three times aslong as broad at the base, tapering slightly to the distal end, which issubtruncate; furnished with several spines at the tip and four or fivespines on the lateral margins.
NO. 1507. FRESH- WA TER AMPHIPODA? WECKEL. 5 3Length, 18 mm.Specimens examined from an artesian well San Marcos, Texas, thetype locality.OnuKjony,!' Ixurerml Ulrich is undonl)teclly the same species as Stij-gcmectes fngt'llatas. The specific ditierences which Ulrich points out,tog-ether with the rather small size of the single specimen which heexamined, indicate that he had an immature specimen. The only dif-ferences which he gives are in the nnml)er of segments in the antennieand the number of spines on the propodus, which in Crangonyx hcnv-ersl! are less than in Sfygoticctes ffageUatdis. These are characterswhich vaiy with age, there being fewer segments in the anteima^ andfewer spines on the propodus in the immature forms.This is the type species of the genus Stygonecte.s to which it wasassigned by Prof. W. P. Hay. This genus closely resembles the genusCrangonyx, but differs from it in having the last two segments of thourosome coalescent.Genus APOCRANGONYX Stebbing.APOCRANGONYX LUCIFUGUS (O. P. Hay).Crangonyx Iuc ifuguti O. P. Hay, Am. Nat., XV], 1882, p. 144.Apocrangoni/.v lucifugtis Stebbing, Trans. Linn. Soc. London, 2d ser., VII,1899, p. 422.This species is the type species of the genus Apncrangonyx to whichit has been assigned ])y Stel^bing. This genus diti'ers from Crangonyx.,in which the species was placed by Hay, in the rudimentary thirduropods which are without rami.I have had no specimens of this species for examination, l)ut quotethe original description:This is a small, rather elongated species that was obtained from a well in Abing-don, Knox County, Illinois. As befits its subterranean mode of life, it is blind andof a pale color. In length the largest specimens measure about (> mm.Male.?Antennuhe scarcely one-half as long as the body. The third segment ofthe peduncle two-thirds as long as the second; this two-thirds the length of the first.Flagellum consisting of about fourteen segments. The secondary flagellum veryshort and with but two segments. Antennae short, only half as long as the antennula*.Last two segments of its peduncle elongated. Flagellum consisting of but about fivesegments, and shorter than the last two segments of the peduncle taken together.Second pair of legs stouter than the first. Propodite of first pair quadrate, withnearly a right angle l)etween the palmar and posterior margins. Palmar margin oneach side of the cutting edge, with a row of about six notched and ciliated spines,one or two of which at the posterior angle are larger than the others. The cuttingedge is entire. Dactylopodite as long as the palmar margin, and furnished along theconcave edge with a few hairs.Propodite of the second pair of legs ovate in outline, twice as long as broad. Thepalmar margin curving gradually into the posterior margin. The cutting edge of thepalmar surface uneven, and having near the insertion of the dactyl a square projec-tion. The palmar surface also armed with two rows of notched and <"iliated spines,five in the inner row, seven in the outer. Dactyl stout and short.
54 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.xxxii.Two posterior pairs of thoracic legs longest of all and about equal to each other.All the legs are stout and their liasal segments squamiforni.Postero-lateral angle of first abdominal segment rounded, of second and thirdfrom obtuse angled to right angled.First pair of caudal stylets extending a little farther back than the second; theseexceeding slightly the third. The peduncle of the first pair somewhat curved, withthe concavity above, the rami equal and two-thirds as long as the peduncle. Thepeduncle of the second pair little longer than the outer ramus. Inner ramusnearly twice as long as the outer. Third pair of caudal stylets rudimentary, i-onsist-ing of but a single segment. This somewhat longer than the telson, broadly ovate,two-thirds as broad as long and furnished at the tip with two short spines.Telson a little longer than wide, narrowing a little to the truncated tip, which isprovided at each postero-lateral angle with a couple of stout spines.Female.?In the female the propodite of the anterior pair of feet resembles closelythat of the corresponding foot of the male. The palmar margin of the second pro-podite is less oblique than in the second foot of the male and does not pass so gradu-ally into the posterior margin. It is also destitute of the jagged edge and the squareprocess of the male foot. There are fewer spines along the margin. One of thespines at the posterior angle is very long and stout.This species appears to resemble C. Itennis Smith, but is evidently different. Inthat species, as descri))ed by Prof. S. I. Smith, the first pair of feet are stouter thanthe second and have the palmar margin of the propodite much more ol)lique. Thereverse is true of the species I describe. Nor do I understand from the descriptionof C tenuis that the posterior caudal stylets each consist of a single segment. Thereare some minor differences. From C. vitreus, judging from Professor Cope's descriptionin Ainerican Naturalist, Vol. 6, p. 422, \t must differ in the caudal stylets. "Penul-timate segment, with a stout limb with tw^o equal styles," is a statement which willnot apply to my species, whichever the '' penuliiinate" segment may be.Family ORCHESTIID^].Genus HYALELLA Smith.HYALELLA KNICKERBOCKER! (Bate).Allorchestes knickerbockerl Bate, Cat. Amphip. Crust. Brit. Mus., 1862, p. 36,pi. VI.Hyalella deniata Smith, Rep. U. S. Fish Com., 1872-73 (1874), p. 645, pi. ii.Hyalella inermis Smith, 7th Rept. U. S. Geol. Survey of Colorado, 1873 (1875),p. 610.Allorchestes dentatus Faxon, Bull. Mus. Comp. Zool., Ill, 1876, p. 373, figs. 35, 36.
'
Allorchestes dentatus yar. inermis Faxon, Bull. Mus. Comp. Zool., Ill, 1876, p. 373.Hyalella inermis Stkbbii<io, Proc. U. S. Nat. Mus., XXVI, 1903, p. 929.Hyalella faxoni Stebbing, Proc. U. S. Nat. Mus., XXVI, 1903, p. 928, pi. lxi.Eyes round or nearl}^ so; their distance apart slig'htly greater thantheir diameter. First antenna shorter than tlie second pair; first andsecond segments of the peduncle about equal in length, slightly longerthan the third; flagellum about twice as long as the peduncle and com-posed of sev^en to nine segments; second antennas about half as longas the body; peduncle exceeding that of the first pair with the twodistal segments elongated and nearl}^ equal; fiagelliim usuall}^ butlittle longer than that of the first pair and composed of eight to fifteenseo-meuts.
NO. 1507. FRESH-WA TER AMPHIPODA? WECKEL. 55First onathopods in the male more than half as broad as long; lateralmargins strongh^ curved and furnished with minute spinules; palmtransverse, concave at the middle, provided with a few small teeth andshort hairs at the posterior angle and a tuft of long stiff hairs at theanterior one; dactyl strongly curved, shorter than the palm. Firstgnathopods in the female closely resemble those in the male.Second gnathopods of the male much longer than the tirst ones;carpus not longer than broad, with a long narrow lobe extending alongthe posterior margin of the propodus; propodus stout, more than halfas broad as long, much broadened distally, especially in the mature
ur. IFig. 15.?Hyalella knickerbockeei. Ann Arbor, Michigan, (jn. 1$, gn. 2 9, first and secondGNATHOPOD.S OF THE FEMALE; OTHER FIGURES OF THE MALE. gv. 2', SP;COND GNATHOPOD OF AYOUNG male; gtl. 2, SECOND GNATHOPOD OF A MATURE MALE.specimens; anterior margin convex; posterior one straight, oblique,and unarmed; palm oblique, the middle portion slightly arcuate, witha notch near the middle and one or two slight emarginations at theposterior angle; provided with a submarginal row of spines; palm inimmature specimens transverse, emarginations very slight, dactjd notso strongly curved as in the adult. Second gnathopods of the femaleslender and weak; carpus and propodus elongated and much narrowerthan those of the male; carpus more than twice as long as broad, pos-terior projection proportionally not so great as in the male; propodusmore than twice as long as broad; palm forming less than a right
56 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxii.angle with the posterior margin, furnished with fewer siibniarginaljspines than in the male; dactyl not strongh^ curved and litting thepalm closel}'
.
Posterior margins of the coxal plates of the third, fourth, and fifthperiBopods serrate.Postero-lateral angles of the first two abdominal segments slightlyproduced, forming less than a right angle; the first two or three seg-ments usually produced dorsall\% forming a well-marked median pos-terior tooth.First uropods almost twice as long as the second pair; third pairvery short, not exceeding the basal segment of the second pair inlength; peduncle short, nearly as broad as long, furnished distallywith a few hairs; the single ramus slender, tapering, about equal inlength to the peduncle and provided with a few slender seta^ at the tip.Telson short, entire, as broad as long, and furnished distally with aslender seta at each side. Length not exceeding T mm.Abundant throughout the United States. Collected at Caribou,Maine; Cambridge and Quisset Fond, Massachusetts; Ann Arbor,Michigan; Isle Koyal, Lake Superior; Wisconsin River, Lake Geneva,Lake Delavan, and Winnebago Lake, Wisconsin; Omaha, Nebraska;Urbana, Fekin, Clifton, Havana, McHenry County, Meredosia Lake,and Lake Michigan, at South Chicago, Illinois; New Fhiladelphia andTuscarawas River, Ohio; Finey Branch, District of Columbia; FointFinellas, Florida; San Marcos, Texas; Lake Merced, Fresno, LosAngeles, West Berkeley, and San Francisco, California; YellowstoneNational Park; Voh^an Reventado, Costa Rica.Through the kindness of Dr. S. J. Holmes, who has compared thetypes of Bate's AUorchestes knickerbockeri in the British Museum withspecimens of Ilyalella dentata Smith, I am enabled to state definitelythat the latter are of the same species as Bate's types. From Bate'sdescription of Allorchestes knlckeThockerl in the Catalogue of Amphi-podous Crustacea in the British Museum, one can not establish thisidentity, for some of Bate's statements in this description do not agreewith his own specimens, and he has also confused the two sexes in hisfigures of the gnathopods as well as in his description. Bate's speci-mens were sent to him by Say under the name Gammarus minus Say;but Say evidently misidentified his own species, for the specimenswhich he sent to Bate, under the name Gaininarus tniniis Sa}^, agree inno way with his description of this species. The specimens were thendes(;ribed by Bate as AUorchestes ktiickerhocherL The specific nameknickerhockeri has precedence of dentata^ consequently the name ofthe species should be Ilyalella knickerhockeTi (Bate).In some specimens of this species the dorsal spines are absent onthe abdominal segments. These npndentate forms were first describedby Prof. S. I. Smith, and were given, the specific name inerinis.
NO. 1507. FRESH- WA TER AMPHIPODA?WECKEL. 5 7After comparing- a large number of the dentate and nondentate forms1 find they differ only in this one feature, and at present it seems bestnot to make this difference one of specific value; consequently, I havemade Hyalella Inermix synon3"mous with Jlyalella htilckerhockei'l.I have also compared Doctor Stebbing\s type specimens of Jlyalellafaxoni^ from Volcan Reventado, Costa Rica, with a number of speci-mens of Hyalella hnickerhodverl and feel convinced that they belongto the same species. The differences which are pointed out by DoctorStebbing between Hyalella fa,ron> and the nondentate forms ofHyalella JinicJierhocJcer! are very slight and are differences which areoften found upon comparison of a number of specimens of the samespecies. The principal difference between the two, as pointed out byDoctor Stel)bing, is the presence of three setw on the inner plate ofthe first maxillge of Hyalella faxonl instead of two. 1 have foundspecimens, however, from Yellowstone National Park and also fromthe Isle Royal, Lake Superior, which agree with Hyalella knicker-hocket'i in every detail, but on which 1 found three sette on the innerplate of the first maxilla'; consequently this distinction between thetwo forms is not valid. BIBLIOGRAPHY.Bate, C. Spence. Catalocjue of the Spei-iineiis of Amphipodous Crustacea in the Col-lection of the British Museum, 18()2.Benedict, J. E. Preliminary Descriptions of a New Genus and Three New Specieaof Crustaceans from an Artesian Well at San Marcos, Texas. Proceedings of theU. S. National Museum, XVIII, 1896, p. 616.Cope, E. D. On the Wyandotte Cave and its Fauna. American Naturalist, VI,187L', p. 406.
. Descriptions of Species from Mammoth Cave. Annual Report of the Geo-logical Survey of Indiana, 1872, p. 181.De Kay, J. E. Zoology of New York, Part VI, Crustacea, 1844, p. 37.Faxon, Walter. Exploration of Lake Titicaca by Alexander Agassiz and S. W.Garman. Crustacea. By Walter Faxon. Bulletin of the Museum of ComparativeZoology, III, 1876, pp. 361-375.Forbes, S. A. List of Illinois Crustacea, with Description of New Species. Bulletinof the Illinois State Laboratory of Natural History, I, 1876, p. 6.Hay, O. p. Notes on Some Fresh-water Crustacea, Together with Descriptions ofTwo New Species. American Naturalist, XVI, 1882, p. 241.Hay, W. p. Observations on the Crustacean Fauna of Nickajack Ca^-e, Tennessee,and Vicinity. Proceedings of the U. S. National Museum, XXV, 1902, p. 429.Holmes, S. J. Phototaxis in the Amphipoda. American Journal of Physiology,\\ 1901, p. 211.
. Synopses of North American Invertebrates. XVIII. The Amphipoda.American Naturalist, XXXVII, 1903, p. 267.Sex Recognition in Amphipoda. Biological Bulletin, V, 1903, p. 288.Packard, A. S. The Fauna of Nickajack Cave. American Naturalist, XV, 1881,p. 880.Say, Thomas. An Account of the Crustacea of the United States. Journal of theAcademy Natural Science, Philadelphia, I, 1817, p. 374.Smith, S. I. Preliminary Report on Dredgings in Lake Superior, 1871, pp. 1020-1026.
58 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxn.Smith, S. I. Crustacea of the Fresh Waters of the United States. Report of ther. S. Commissioner of Fish and Fisheries, Part 2 for 1872-'7;> (1874) , pp. 637-665.
. The Crustaceans of the Caves of Kentucky and Indiana. American JournalScience (3), IX, 1875, pp. 476-477.
. Report on the Amphipod Crustaceans. Annual Report of the U. S. Geologi-cal Survey of Territories for 1873 (1875), pp. 608-610, pis. i-ii.Crangonyx vitreus and Packardii. Memoirs of the National Academy
i
Science, IV, 1888, pp. 34-36.Smith, S. I., and Vekrill, A. E. Notice of the Inverte.brata Dredged in LakeSuperior in 1871, by the U. S. Lake Survey. American Journal of Science (3),II, 1871, pp. 448-454.Stebbixg, T. R. R. Amphipoda from the Copenhagen Museum and other Sources.Transactions of the Linnean Society, London (2), Zoology, VII, 1899, p. 423.
. Amphipoda from Costa Rica. Proceedings of the U. S. National Museum,XXVI, 1903, p. 929.Ulrich, C. J. A Contribution to the Subterranean Fauna of Texas. Transactionsof the American Microscopical Society, XXIII, 1902, p. 85.