ANNOTATED LIST OF THE SPECIES 141 HypophthalmusperporosusCope,1878,Proc.Amer.Phil.Soc.,XVII,673,Nauta; Steindachner, 1882, Denksch. KK. Akad. Wiss. Wien, XLVI, 4, Rio Huallaga. HypophthalmusedentulusCastelnau,1855,Anim.Amér.Sud,Poiss.,47. Rio Huallaga and Amazon northward 15653,5,230-3824mm.,Iquitos,Allen,September,1920.
15779,8,220-302mm.,Iquitos,Allen,September,1920.
15780,2,248and287mm.,Manaos,Brazil,Allen,December,1920.
15787,5,215-300mm.,Manaos,Brazil,Allen,December,1920.
15976,1,245mm.,Iquitos,Morris,1922.
Known among the inhabitants of Eastern Peru as the maparaté. Conspicuous for the extreme width and flatness of the barbels, as well as their jet-black color. The barbels may be seen extended forward during locomotion, near the surface of the water. Usually found in muddy streams. Steindachner was supplied with a Quichua name of mapa-racui. The larger specimens have less black on fins and barbels. Those from Brazil have broader, blacker, more taeniiform barbels than the Peruvian specimens, especially the mental pair. Our Peruvian specimens, 15653, have a larger head than the Manaos specimens, being contained more than four times in the total length in both cases.
Our specimens have the dorsal and ventral profiles moderately and uniformly convex, except for the elevated dorsal basis and slightly concave head; adipose short, elevated, thin; dorsal fin short, its origin well back, in the vertical from the tip of the pectoral, its distance from the tip of the snout 1.2 in its distance from the base of the middle caudal rays, the dorsal fin elevated, its spine long, smooth, slender, D. I, 6; pectoral fins well developed, a half longer than dorsal, with a similar spine only three fourths as long as the longest ray; ventral fins very small, the longest ray 2.6 in that of the pectoral, origin of fin at middle point of pectoral; anal very long, its basis nearly equal to distance of dorsal origin from tip of snout, anal rays about 64; caudal deeply forked, its upper lobe the narrower and somewhat longer, both lobes, especially the upper, sharply pointed. Eye strongly inferior, 9.3 in the head, the interorbital space variable, due to the paper-thinness of the skull and elasticity of the opercular region and visceral skull; head widest behind posterior nares and behind eyes, narrower midway; snout obtusely pointed; jaws equal; gill-membranes cleft to within two eye-lengths of the symphysis. The six barbels of moderate length, the longest being the maxillary, reaching to near the tip of the pectoral fin. The conspicuous branches of the lateral line canal pitted at frequent and irregular intervals; numerous dichotomously branching or netted systems of canals on the head.
Barbels nearly black, all rayed fins tipped in black, extending over distal half of anal rays, and on one specimen nearly the entire caudal fin. Family VII: Pygidiidae SiluroideitrichomycteriformesBleeker,1863,Nederl.Tijdsch.Dierk.,I,112. TrichomycteridaeGill,1872,FamiliesFishes,19.
PygidiidaeEigenmannandEigenmann,1888,Amer.Nat.,July,649; EigenmannandEigenmann,1890,Occ.PapersCal.Acad.Sci.,I,316; Eigenmann, 1918, Mem. Carnegie Mus., VII, 259-398, pls. xxxvi-lvi, text figs. 1-39.
142 FISHES OF WESTERN SOUTH AMERICA Throughout most of South America, abounding in moderately high to extremely high elevations, contending with Orestias for the distinction of attaining the greatest altitudes among the fishes of the world.
Air bladder poorly developed; without adipose fin; median fins short to moderate; unarmored; skull enclosed by muscular coat and naked skin; teeth villiform; nares remote; dorsal fin not united to skull by a bony plate. Mostly given to burrowing and holding, the fins well back in response to this habit. The names candirvi and carnero, implying the habit of parasitism, are met with in regions of lower elevation and the more tropical portions of the continent of South America. These terms are as loosely appled to a variety of fishes as the words perch or trout in North America. In fact they are so universally known that few distinctive, colloquial names exist for any member of the family in those regions. It is true, however, that the inhabitants are rarely able to distinguish one from the other, even between distinct subfamilies.
Among the genera Acanthopoma, Vandellia, Stegophilus, Branchioica, Urinophilus, Cetopsis, Hemicetopsis and Tridens, representing three subfamilies, there are supposed to exist one or more species of minute, elongate fishes which exhibit the following type of behavior (or misbehavior):
1. A positive tropism toward urine in the water.
2. A tendency to enter the urethra or other natural opening of the body of man or other animal.
Dr. Eugene W. Gudger, with his genius for documentation and the pursuit of fish-lore to its ultimate hiding-place, has assembled an impressive volume of evidence in support of this popular belief. Some of his information is direct, leading back to supposedly original observation; elsewhere he is obliged to rely upon hearsay. As a summary of present day information on this subject Dr. Gudger has done a well-nigh perfect piece of work. He rightly approaches the subject from the generic angle, not definitely ascribing these attributes to a particular species, as, of course, there is little precise or experimental evidence as to the habits of the respective species.
The fearful appellation candiru was adopted by Spix (1829) as the specific name of the suspected form. ‘‘This fish belongs to Cetopsis, a genus which we have described, but I do not know whether these pests are the younger individuals of the species described (C. candiru and C. coecutiens) or whether they are individuals of a third species of smaller fishes endowed also by nature with this cruel instinct.”’ (Martius, in Spix and Agassiz, 1829). Thus Martius in the preface of the same work leaves the matter of identity open; in spite of all the work done since it 1s still an open question as to whether Cetopsis candiru or other form is the candiru of tradition.
Eigenmann was greatly interested in establishing the truth as to the alleged tropism toward urine. (1918, ¢.) He was sufficiently convinced to attach the opprobrium of it to a new genus of Vandelliinae, Urinophilus. At his suggestion I made repeated efforts to collect these minute fishes by means of their attraction
ANNOTATED LIST OF THE SPECIES 143 to urine. This was done by saturating cotton and cheesecloth with urine and using it as bait in a glass-jar type of lead-in trap, with a small funnel opening. The experiment had only negative results.
Pearson (Gudger, 1930a, 180) collected Urinophilus erythrurus on a bit of floating fresh meat, demonstrating a chemotropism for meat or blood on the part of that species. Marcoy described a similar episode in which a turtle shell with the meat cut away was both bait and trap for swarms of carnero. My own efforts in the field warrant only the following conclusions which are briefly summarized without much discussion:
1. The occurrence of the several ‘“‘carnero” species along beaches, among sand and rocks, and their possession of retrorse, opercular spines like those of the Pygidiidae in general, show these fishes well fitted for holding their places on the beds of rapid streams. The evolution of this intrusive manner of living into a parasitic one, with the use of the spines as holdfasts, is not difficult to imagine. 2. A single specimen, described by Myers as Urinophilus diabolicus (q.v.), was recovered from a Pseudoplatystoma, half inserted into the abdominal cavity. This seems an indubitable case of parasitism, but shows no special probability of a urinophilous tropism, as the point of entry was at some distance from the vent. A similar case was reported by Ternetz to Myers (Gudger, 1930a, 183). 3. Trapping experiments were wholly negative and inconclusive. 4. My host at the Rio Pacaya, Sr. Medina, showed me a scar at the groin which he attributed to the ‘‘carnero,” but admitted that he did not have opportunity to make a satisfactory identification of his assailant.
5. There is a universal agreement on the part of the inhabitants as to the existence of urmophily, but no agreement as to the form or identity of the urinophile. In all the lowland regions visited the inhabitants were reluctant to enter the water at all, and did so only with some protection to the genitalia. Bathing was commonly engaged in as a daily, or twice daily, ritual. But it was nearly always done by dipping water over the person with a dipper or long-handled calabash, and on the stationary rafts at the water’s edge. Some households afforded a riverside enclosure, or bath-house, built on the raft landings, for greater privacy. Among the better-class whites and mestizos the bathing-hours were apportioned. (This practise, as reported by Algot Lange, somewhat mistrusted by Gudger.) 6. No specimens were collected which were sufficiently small to effect an entrance to the urethra.
7. The mechanics of making the alleged entrance would seem impossibly difficult, and would meet with such prompt resistance, as to be but rarely successful, especially in the case of males.
8. It has been suggested that the odor, the salt, or the warmth of urine might be the attracting stimulus. At least in the case of the piranha blood is well shown to have such an attraction.
9. There is more reason to regard the behavior of the candirv, or candiris, as cases of parasite attacking a potential host, attempting to attach itself anywhere, at random. Of this there is no reason for entertaining doubt, in the light of
TOJEM UL O[IYM ommyzest Sur Ssolsou WO 9U0 PaAOUIAL PRY IY WAY Plo} OLYSB,) “ACT :SuoTyy -BOYLIBOS PYATOIal J[OSUILY $07/DADI ap NLipuva BIB 89h SOX -IBIM UO YSISUL SAATLBN sok | Josey ‘odAy snoprydoutim «oppeurs ‘soteds omy, | ‘S6QT powodat ‘yygT “yaeqor Supe ‘U0T}d9} soued poyeyndure niApueo “u0z ON sok -0ld 8 SB [[OYSs-jnU0I0;) doIY} Uses pey iloppe[q puB BiyjoM sloyuy | -BUTy ‘J68] (desuaTNog) Yorg uolRiedo pur YyRq ON JOY YAM poyBaly :
BLY JOIN .lo}Ud | GUO] “UL OMT, NBULY ‘FRg] ‘UsUTay}S Nhupuerod uozZeUy ON ssutuedo aqua sotloads oT, | ‘Oden ‘s.O28T ‘LO8T ‘UOWIO UMOUY IB ON ON SOMTEIBF PUB SloyZRG JO SadYWO Soyetjous_ | Bllepeyy ‘UOZeUTY ‘FLT ‘leT[oy SOSBO “MWY suajjauun puod uolyBiodo WOAf YYBop JO MoU NUpUBd BIT] oytoods AG JON ON -0YJUDIP SB poqkiuseq | :ssuruedo usppiy AT[etlvedso ‘Apoq syorpy | -[eNET ‘OLST ‘(4eT[MTV) SUTBAL YoDPINY JO uonrorydde yoy Syyeq SpIvd Joos SYoR]YR /SUOTSaL SUIMBIP ON ON SULINp WOTPRULIN Ploay poytqeyut ‘sotoys UO punoy ssoyout g OF Your * | Soars [RABI ‘GOST ‘AOoIepYy ON ON BIYjoin UL punoy FORT ‘toyjuneyy snsoipisur snjrydo WUQnoy JO YANOUL Sox ON Bag sty ATNyyqnog, | UL punofy SwuMyoot pue Bayjoin uolRijousg nupues ‘GGRT ‘ypareyUutey IdAL 0 ON OJUL posreyostp Jl ouLIM JO UUTN[OD pussy UOZBULY ‘GGRT ‘neuUye4se,) SNOWes BuUBINY) (6) ON -old SUOT Your = uMyoot puR BIyjoin syoryy | ‘narpuvo “OFST ‘Yysanquroyos oytoods & pnbipx Jo vom’ wm.
UL paplodg poamno pue ‘vursva 07 pord -de pnbpxr YAM poyRod} UBWIOM MBS J[OSUUTET Raven, “uozeury N.up -uk) ‘o1oued ‘gegy, ‘stdde0g [HyAqnop Ayyuepy SLOYVeq syouyzyeR ‘ wstdo.ouLtyy) harpue ‘fizBtg ‘TERT ‘sniyaepy BITR}UIS puey puovag ON WA0d Soondoaid ory, powuyuoo pouiguos UIST} IStAt / westdo14 SJUBJIGVYUL JO SUOTIOBIY cae -OUuLL() suoumoadg poayrodoada ByeqT auBu ‘AYITBIOT ‘ayep ‘1oyyNy OMIGNV,) AHL 40 AUOLSIF{ ASV) AO SISKIVNY “AT ATAVL 144
| stued Wolf [BMBIPYYIA | gSoX | Jo osBo JO MOUY Sosvo [BULSBA ‘JIL UO pozBilod¢g, |(SovURPY) BIVETY ep opadpy “Aq (sted) Joppeylq ‘stued ‘sour ¢ 10 Z iON sox | poyeiedo pry oy uRULoWULy “qT WOLy y10d0yy Id}JO] SABI Ss = = = — ————————— — —— _ = ———— — = = = = se es ~ —— =— (BUISBA) | (snue ‘osou “1ve) /RUISBA iSO OJUL dUBIZUD JO asBo pURY PUOVIS sjz0doyYy Jaye] ‘sodwey “1q Autds upueBd (9L0UL ONS || Aways ‘onfq ‘soyout ¢ 07 dn ysy tawoy pRtouory |-epy) Ladepy onp ‘ozeL ‘vey, WALL UL SULCOpUNe] }BOUL | PIA poyoezyV “URWIOM JOYJOUR WoL, PpoAout nlp {Sok {sok uO Uoye} snpeydouls;) | -of DAIpUBS UOos PBY JIM osSOYM UBUT YoY | -WRo ‘TUE ONY ‘TZE] ‘UOsUReg SULUUIMS oryiseaed ATperueyod 10 BUIPBA JO BUBPIOAB |SLaY}O $4SoY UO popoo][Oo OIG OLOUIBI WIG VARY 0} podvklpaq OLOULLE SIX ON ‘S]yBt puw sosnoy ye | eyseaed opuyop UC) | (BOS B UMOYS SBAL !SOLI0}S ay} JO yo paBoFT | ‘UOZeUTY UBIANIEg ‘OZ6T ‘UV poe 3 2S S eee eee = aes ———— = ee Se es es = UBUIOM UBIPUT JO VULSBA WHOL] | 9UO PIAOUd :SosBo E MoUY :Soinqiode ABUL iON {sox pue [vue soyetoued !(-uld g-G) soyout Z'E-z NUpuwd ‘ZZ ‘o}UlLoD oT snyiydours) auieUu ANYRIO}L] SoPBoL | NLpupI se poyyuopt 9.INYRAOYT] ON ON -10) puR Ssaqtiosepay | A[eyUyep you ‘Aueyy AMIYMATIAI WOLF SOLLOJS PUBY-PUODeS AURTY | SMoLADY ‘STG, ‘uuBUTUESIGT ON oN | qoRty AIBULIN LoyuyT | Narpuko ‘uozeury ‘SuLIyy UO A ALIYM ALIA NapuBsd suoz ON ON | Jo paBoy ‘poqreq ‘suol soyour omy uRY) sso'T | -vury todd: ‘FI6T ‘oyoupoo Ay (SUIZURDOT) ON ON sutusdo Apoq Auw out ysnd 0} wo8ey | vUBNBeI ‘TIE, ‘esnery IOJVM | loyUa 0} posTqo way BIBS GULIOAOD SUBIP poyedodo uvur Jo YYvop JO MouY S4sBy ON “uy ‘Sjyet uo surged | | | ploy se wif jpswop ss1ayyeq Jo ssutuedo yBainye Yy oompuRy ‘AIBAB LE ‘ZT G] ‘osueT 145
146 FISHES OF WESTERN SOUTH AMERICA parasitism in general. It would seem to account for the lack of specific points of attack by the laws of chance.
10. The apparently genuine cases supporting the tradition would inevitably receive more attention than the random attacks of the fish upon other parts of the body. This may be partly due to the fact that the indigenous tribes of the region have a superstitious attitude, perhaps reverence, toward the organs of reproduction. Nor is the psyche of the white population without its rich erotic vein and consequent emphasis upon phallic matters.
If a study of the porpoises or manatees of the Amazon should yield abundant piscine parasites on or about the genitaha, a better interpretation of what Gudger ealls ‘‘the only vertebrate parasite of man’ could be made. In the foregoing table I have attempted to bring into a bird’s eye view the evidence pertaining to the candiru. Most of the cases given are taken direct from Gudger, or from sources cited by him.
A summarization of the table will warrant the following:
1. That the exact identity of the culprit is not clear, nor the limits of its depredations, whether it is the original Cetopsis candiru of Spix, or whether a number of species are involved, even whether some unknown species exists which is the guilty one. No specimen known to have attacked man has been preserved in a museum, so far as is known. Orton reported, and others have repeated, that two species exist. The Peruvians told me that there are two, a larger attacking females only. This may be the same form that Jobert calls candiru de cavallo, or “horse eandirt.”’ 2. The credibility of some of the witnesses may be impugned, when accepting such accounts as that of the fish ascending a jet of urine from the surface of the stream, or the dorsal fin being used as a holdfast, and the fish being a myxino (hagfish or lamprey).
3. A schooling habit is sometimes described, not easily reconciled with other habits described.
4. It cannot be questioned that members of this family sometimes attack other fishes or the raw flesh of fishes and other animals. They have been attracted and trapped by their taste for flesh. It would be an easy step, considering their intrusive habits, for them to attach themselves to the mouths or gills of larger catfishes, as witnesses report, or even to invade other parts of the host’s body, or to be attracted to any opening which they find.
5. Persons have been attacked while bathing, and have received minor wounds not very clearly chargeable to these fishes.
6. Some of the cases of operations are doubtfully due to the candiru, others are rather directly reported and convincing, none quite first hand. 7. It will be noted that among the stories coming through popular sources, and also most of the older accounts, that the candiri is given to attacking males, while the more recent and more direct reports show that females risking themselves in the water are usually the objects of attack.
8. There is little clear-cut evidence by which we can definitely sort out cases
ANNOTATED LIST OF THE SPECIES 147 of urinotropism from parasitism. Some of them could be as readily interpreted one way as the other. The case for urinotropism is very weak indeed, and few cases if any cannot be ascribed to parasitism. However, it is an incipient or facultative degree of parasitism at that, and not at all an obligate one. It lacks the ear-marks of true, obligate parasitism, for (a) there is no evidence that they cannot turn to predatism, or do not rely upon it nearly altogether; (b) there is no especial readaptation of the body for the inactivity of parasitism, unless, as in the case of Hemicetopsis, a very great capacity for soft, liquid food; and (c) there are no known modifications of life history for such a mode of life. Haseman, 1911b, page 315, not quoted in the table above, agrees that some species of candirvi are said to live in the branchial cavities of the larger siluroids. “In fact they attach themselves to any kind of fish or animal, including man. By means of suction for which their mouths are adapted, they fasten themselves to their victim, and then painlessly cut the skin, and gorge themselves upon its blood. The fishes brought into the market at Manaos often show many wounds inflicted by the Candirtis. Below the first fall in the Madeira river it is difficult to take a catfish which has not been bitten several times by the Candiris.”’ This may be considered as much in the nature of predatism as parasitism. Numerous popular authors give some brief space to the candiru. Prodgers, Bolivia, page 115, calls them the kandiro, ‘‘a slimy leech, three to four inches long, which get up into the rectum,”’ where they erect the dorsal fin and have to be cut out. In his Peru the same writer, page 210, calling them candero, 14 to 3 inches long, ‘‘and a shark in mineature, with similar fins and teeth... when the fins fold down it looks like a worm... its favorite point of attack is the fundament.”
Orton found the name applied also to three species of Serrasalmo, by reason of the same dread inspired in bathers. ‘‘The natives accuse Vandellia of entering the nether openings of bathers’? but he did not meet with any confirmatory cases. Guenther, 1931, alludes to it as a little, thread-like Silure...an unpleasant acquaintance ... forces its way into the urethra of bathers and owing to the spines on the gill-covers it cannot be extracted.”’ Marcoy, 1875, II, 184-188, provides us with a drawing which has more resemblance to Hemicetopsis than to other suspects, as you might also suppose from the stated length of 5-6 inches... “it wages relentless war on the calves of the natives who come within its reach; darts impetuously at the fleshy mass, and rends a portion away before the owner of the calf has time to realize his loss.”’ Then he goes on to say that “the smaller of the species are 2-3 lines long and dangerous in quite another respect ... introduce themselves into the secret parts of bathers, where their extended fins retain them captive... whispered warnings to avoid passing urine into the water ...one remedy alone the Ucayali doctors (sic) ...a tisane made with genipa or hwitoch apple... hot... dissolves (they pretend) the animal which obstructs.” I (W.R.A.) found numerous natives using a fruit with the same name, rubbing it into the skin of hands, face, and feet, where it gave an intense, purplish black stain, said to be repellent to mosquitoes and sand-flies.
148 FISHES OF WESTERN SOUTH AMERICA Marcoy describes the procedure of fishing for the kandiroo with “turtle shells having a little flesh adhering... the fish getting as fat as a goldfish in a globe... the excess fish being fed to the poultry.”
Subfamily: CETOPSINAE Dorsal fin entirely forward of the ventrals; vomer with a single series of teeth; head compressed; anterior nares almost labial in position; eye almost entirely concealed under the skin; a single maxillary barbel; two pairs of mentals; opercle unarmed, and gill-membrane broadly united with the isthmus. Genus 52: HEMICETOPSIS Bleeker CetopsisAgassiz,1829,Sel.Gen.etSpec.Pisce.Bras.,11,sp. Hemicetopsis Bleeker, 1862, Versl. Akad. Amst., XIV, 403; Bleeker, 1863, Nederl. Tijdsch. Dierk., I, 111; Eigenmann and Bean, 1907, Proc. U.S. Nat. Mus., XXXI, 664; Eigenmann, 1910, Rept. Princeton Univ. Exped. Patagonia, III, 398; Eigenmann,1912,Mem.CarnegieMus.,V,210.
Cetopsis(Hemicetopsis)EigenmannandEigenmann,1890,Occ.PapersCal.Acad.Sci.,I,318. Type: Silurus candiru Spix Amazons and Guianas Dorsal fin forward of the vertical from the ventrals; teeth conical, vomerines in a single series.
106. HEMICETOPSIS CANDIRU (Spix) SiluruscandiruSpix,1829,Sel.Gen.etSpec.Pisce.Bras.,pl.x,fig.1. CetopsiscandiruAgassiz,1829,Sel.Gen.etSpec.Pisce.Bras.,13; Eigenmann and Eigenmann, 1890, Occ. Papers Cal. Acad. Sci., I, 319; Pearson, 1924, Ind. Univ. Studies, XI, no. 64, 16, Rurrenabaque, Bolivia, a specimen of 242 mm.
Cetopsis candira Cuvier and Valenciennes, 1839, Hist. Nat. Poiss., XIV, 386; Giinther, 1864, Cat. Fish. Brit. Mus., V, 199; Cope, 1878, Proc. Amer. Phil. Soc., X VII, 678, Peruvian Amazon; Steindachner, 1882 (1883), Denksch. KKK. Akad. Wiss. Wien, XLVI, 5, Rio Huallaga, five specimens, 190-280 mm.; Eigenmann and Eigenmann, 1888, Proc. Cal. Acad. Sci., (2), I, 157. HemicetopsiscandirwBleeker,1862,Versl.Akad.Amst.,XIV,403; Eigenmann and Bean, 1907, Proce. U.S. Nat. Mus., XX XI, 664; Eigenmann,1910,Rept.PrincetonUniv.Exped.Patagonia,ITI,398; Fowler, 1939 (1940), Proc. Acad. Nat. Sci. Phila., XCI, 232, two, Contamana. Rios Ucayali and Huallaga down the Amazon, to Bolivia 15801, 15, 118-174 mm., Rio Ucayali, Nueva Ytalia, near Cashiboya, Allen, August, 1920. 15814,1,1387mm.,Bretana,RioPuinagua,Allen,September,1920. When taken, subcylindrical, and turgid with blood and offal. Specimens therefore much shrunken and wrinkled in alcohol. A favorite haunt of this species