<> arerd sbasesaa es Hast —— Peeecece 3 ->4 ose Saas pone s—s- oe a bod nee Ne noe en be beer este S cett eee bese bee a Sa Se ae Wl a Oe 2 6 — 0 — oO 0 9 me OA OOO a Eee 2-9-9 — 8s eae eee eet ea retesecsrarstetes SIRS L i Bs Bias ire eee ee ececeeeeelereeseeee . bese Sa oe ea ee oees sae Sees =: | ee te tateletetetetetete: .s Sats tess > PE NO Oe —~ += oo oa ee ceceatetees roo atetetetetite: fe fetetetiiete pret gear e gene ene cenvet bee erheetesereee tie. eat etet beter etet aces ia lelelelereteleceler ie tye! : nrelatete! 2) sb ELLE A MM $9 2-0 Fmt 8 Ow ts scscebecesssseasecsceststetatetacststscatstststereestss O50 0-9-8 — 3 ~ 0 9 OF 8 oo 8 is SSS 9 OS 9 8 oO OOO 8-0 0 Ot tenes ae eet teins eee 2 a a ee ee oss eee e eee ee seer etetessceseeee pare ates o— + -o— 9 SS SSS RRS ne SSS SS + eee Ss tot = warts eas pe ig ieeeeeseese Peeretatts scsitee eee eletosest 1 See eee ee pe sae ess Se [PioSastte sess se ste. pose eee es SSS EE ESS Seer ete PRA SR ee e000 0 9 Oe eee RRR HS a cme ee esse te eae He ra en ene oo ere rect => Nita Se Setrtrististst: Sttese OO OO OOO OO OO OO 8 OO Sere es = FIIs telsis ale lec SO a a a Wee i teas a rr ee er ee eS 59S S~ 0= 0-9 O-O- OS oats Se See See ==. Te Se ee = ss BESS SSE DSSS mecelelelerrre rele elerecerecererertrrserereesereci senate setsraeaea ses eoe btn ese sede Petee sist Shit Sedna SS SSRRSe Rei Pi sePtea Et rte ee tt Te ee eer eeete ee eeese pescetes Sere hare Sabie Paes See ee See eee aati estate Stee Si See ei eee eae et ee et eaeae Sasa seea sr Se SSS SSS Ristecst steers Pe eteeetieete. Fivt partes aise ee et Satsist ss Poser ese 3 eee Sana srs apeentacese te setet epee e srt ree FSS SSS SEES EES ERS AAS t a $0 6 80-6 8 8 8 = crores poset rere trees: Fe eS acne wee te ewer tees tote ee +t Ht it ry ae peneeere; ese: z = 2525! + asses TS SS EEE SSS [SHANNA TSS eee $00 0 6 0 0 0 8 OP RO 8 mt PS Oe er or ee Oe et ees ere ta tate 9 F— 8 8 Fr h $= 82e-8- pee tss costes HRs > Ses Sees ees eS ae le See SSS wi aere Hs =o Pe SS ee eh ea re rs re re Oe a Oe Oe Ow Ow ee ee ee ee Ne Oe EEE epeelele. poses OO OO PO OO OO 8 OF OO On Ot oO 80 8 88 poorest i OO On eee eee tr One <—*. se; PEL SSeSSSSe Se debates es ebebesseslabesesesnssanseenssshasenseeees hh ee eee ere err eee casa Sa eee ea abd a eee be he bs ae etetin oe ree eeee tes tebe eee = Stasis Ss os ne as ae eae eens Fig hae eyes — _—~ SSS ESS EEE 5 00999088 98 09 8 8 Oe oe a a a ee rn a re ne ee en a a re ee han rE a ere ESN e OR EE eRe, Oe ne ee ee aeentes ee eee sect e ect e eee eee ee eee ee eee ee eee eee eee be ee b ee eee eee e eS aS 5S ae ee ee i> ideas SSS SESE PI ee et 999 9 9-989 - eae ae ee eet sree ear erty a2 3232 SIH SO OO oe OP i Or Oe OT OT OTe Otte eter t ae BAS Rh HASSE ee eee eee eS SS ESS MRIS Gt Ree = FO 8 Or Or Oe ee Oot Cr ener ee tere ete test! — > a ~>* -o-o—e- LEE EE EE ate ee ee oP OS Oe 9 = 9-0 0 0 > LMA A riatebetessgersceretetesspeeerepeens sary az retreat ity tate nate terscrees seaboec rise peters —2—o~ oa > O—o-— ete tstststsss > ss 3+ 9 88 Sretetetotetstes: Sececeseat ests eete he cesers pibtgsbssisscastessisseasaesyeeseseret apibesara ess Pariser sees ateeee ates teaese: ema tetetete te ete ere ene lene peace ng enone ee, Pips deessississsesaseaepapesseeeese tess tasepyeetgeasesateseeet Sa SRE iiss iteeiwisaes ets ees Seater sat ate teee iene neirneeearere 6-9 SS ete ro>es Seihprerizscorereeee THe .~ = Po 9 99 0 Ss 9 O09 90 0 Ot ice reteset a tata te seawater eee eee ee ee eS EES SSS Nee a 6 OS Or OO POR Pte 8 StF t FO 8 8 OF eT Oe 9 61-56 6 8 9 88 OO <3 505 8-9 9 OOo en P98 : Tet einiciste tet eUtae es ees $9 9 00 8 8 99-9 9-9-0 0-9 0 0 8 0 oes MARA AAA as => DSSS SSS SS SSS Sissesctscecstece te pe apes sseees TS SSS EEE EEE eietrtrertetreeteererses eitistsieisisiestessosseassesess tas — ToS SESS [oistatetesesesrsrsase ; By tet eta ese os Teatro 0 9 oe oS Seteataear ents irpeisesisseee iste er ee eres ee eee tee are Soieweseicsess atte FOr e918 9 8 8S 888 OFS 8 8 O88 8 OO 8 8 8 TE eS SUES See ata > ied we eereeiseeens bans rs oe TsFetet el os Ets ete ee saawees POO 0 tees Ste he Te Oe reassess, Po oo west sete. <= feastecse heat eset ro — 9 6 oo 8 — $8 9 8 9-9 OO Hits RRMA OF Fe OE Od — 8 — O— OO OO | eGo ome ae o> Se sot eee. oe a i re eee ee eee eee t eee =o = SSeS =" 32s ee ee we ee eS STEEL OEE E OM AMO tT ee ee <2 === FO OO On eee ee ee ee ete te bette 2 iH * Oe Oe eee ee eee eae en eeeeeeease eee ees Soe Oe eee ee tate tsleeee Sas Fe ee Se Se ee ents oS ie tt She RE eee e eee ee amet a tee wee eae e ee Owe weet eee eee Rae ARO ROR RE eee eet Oe RENE SHS TS EOE RE OER S AE TO Oe eres eso eoee Soe > S88 8 8 o— 2-59 i os > SSeS SS SEE ES stissbsbisses pee aess esse ares * — [aa raetete ast eeree treet ree Seer urnee wiririeeoe ewer se ess o> Sess Se Sst Ss Toe eat ete ee eee Ee Ree ee ee eee eee eae ee eee eee te eee Tet et eee eee 3 Tersceees + Parse ea erase e eet ra erea ea Seer er eae anita sats e iets hiss Seti reise sanaeteiealen te = 98 8 0 Oe On Oe Fe OO OTOP FO 1 9 8 8 8-9 OO OS OO Oe One Se Oe TOO Cee reat tes Se I OI I OF 8 8 8 FP OOO SSS ONES O STOOL OUT OT oo 8-8-9 9 es : Soeeysrese SESeeeOR SSM Nile eaee reece taaedeseetesetensie Sects ts Cece Sete OsES Pa ties Tors se st ae see OO Fe On nO OOo SO oe nt Oe ee OO Ont OOO On Oe OO oe eet Geee i Pe i ee Pan er te Pe Pe PO OO OO OOF PROS OPe On Te POCO e es eee ee oes etree =o — 0 00 6 86 — 0 8 5 OO OT OO er ere Ore ete ee eee eee ee te eee! Sie trpceresisasiete set teeeseneree Sr eraser eset esi = i} af ts ttisssNS . — eee oe o~ prers cos suuasreees Dane Sa aaed eeeeesee ett tes 84sec scecsarsre Peeeee ea thi ti sa ralesasardrorcrarere 3234 SIDS: see 3 ee ee aa ene eeeeeae beaten 5257 [esses ss ee ee eee eee Peri iors pose ese Srisisicsstetsore SsS235 on oo OO wisi ei eri seieetesssrttasipigetitte STesstesss ters sss sss rsarese oo Sei ieisastesecesstesteessietnteee Seier ti eisibitssis ces. Pgtasei grat ssatisitsesssccsssssa ties iist Tease SER Ea SSS SASSER Tete Pa eae teeta sees Pissrspisisteseessiitypstsssetpsgssse aes aeaeeeeeae teehee OO mo mm re 8 a ee oa Oo 0 $90 0 9 0 FOO 9 Essa sate bigdenss teeter teria tate i tie ste trae Lose ctatsetete ceneeetetetacesrcesesriressie terest eeraet b- 9 50 55 Sb oe Sic tetetetete odoararcstsarsesessretetrisssisisseneles <=. Seressrssiesrsserestateceee Seema eS epet esses tetas esele hata te eae: rs oo — 9 LACM A ABA A EAA << a abe b ot bebbeeebe bee eeee Ooo 1 OO Ot te eo eet ecleeen = = -o=s Lyte astete neta = = ertets es = >! 252525 LARA at ates == WII a ee eetersretetessessstesseersesesee st eeee Preteteeaeresectes Sg issqisieiessics SotrEsesee Siecesetereies Seaeseieae cece sreriterirsre stress etes ee Epi stetetetetesscsrscetssterssesststitecstsesesesesesere > Sie ees cea et see eee ae trae es teeter ee te neces ete tate ot eee tt ce seseree pi3sbibisirieseteetatieig ety ists: ere Seaeasct sibs tarieigtesear sear sries Srsripiesssssasesgieiesgess ee ease Siiataessseasaseterat sesso 53s ===s255 SEsSIetos paleo ~* 6-9-6 9 0 cos s+ = = = SSS = Soteteiesspee wiser trees te sears espe esses tate er iser ee ores aS a= SiS A eee > @- &— 0 6 FeO ~~ = Sesaseae ee = . eieeeteceiseces we eere rca rewens 5050 8-0 9 0-9-0 - 9 8-92 9-0 O08 Oo a et, ceeeesee 99 89S 9 OO 08 ORO 0 SOE Oe S99 SOO OSI ISO OO On On ee ge bette ent Seeaset sees Sel eiet esa ee reeew te eeee = -—o—o— e— o— © - © —6— 6 = & — § = O— & o-— = 4 THE JOURNAL EXPERIMENTAL ZOOLOGY EDITED BY WILLIAM E. CastTLe Jacques LorB Harvard University The Rockefeller Institute —. Epwin G. ConxkLin EpmMuND B. WILSON Princeton University Columbia University CHARLES B. DAVENPORT Tuomas H. Morcan Carnegie Institution Columbia University HERBERT S. JENNINGS GrorGE H. PARKER Johns Hopkins University Harvard University Frank R. LILuisz RAYMOND PEARL University of Chicago Johns Hopkins University CHARLES R. STocKARD Cornell University Medical College and Ross G. HarRISON, Yale University Managing Editor VOLUME 35 JANUARY—MAY, 1922 THE WISTAR INSTITUTE OF ANATOMY AND BIOLOGY PHILADELPHIA, PA. CONTENTS No. 1. JANUARY Mary JANE Hoaur. A comparison of an amoeba, Vahlkampfia patuxent, withepissue-culpure cells-maulthree: freunes® «29. bisy-in acest -elo es tie oie eee ciel 1 THEOPHILUS S. PAINTER. Studies in mammalian spermatogenesis. I. The spermatogenesis of the opossum (Didelphys virginiana). Eight text figures and three plates (iwenty figures) ..0... 4.2) -60. 2 veleesae ec awe seals 13 JosepH Hatt Bovine. The effect of light and decapitation on the rate of CO, output of certain Orthoptera. Three figures........................-5- 47 Dwicut E. Minnicu. The chemical sensitivity of the tarsi of the red admiral butterfly Pyrameis atalanta Linn. Three figuresy....)..)....5.0..5--.. 57 No. 2. FEBRUARY Ann Haven Morcan. The temperature senses in the frog’s skin. One IN ERUITRO 316 Sa cio. GeO 0 6 coe CGI CS OAT ec RPE ORE PS TEE: OSS Paka At DOE AUN tos 83 S. R. Detwiter. Experiments on the transplantation of limbs in Ambly- stoma. Further observations on peripheral nerve connections. Thirty- GWG RTO UTS te oc us AMR IN Scie s Pryeats re oa eho: Sanco aoe heal shal on dallche & chews Sem aeeTS.S 115 H. L. Wieman. The effect of transplanting a portion of the neural tube of Amblystoma to a position at right angles to the normal. Eighteen IPAUIRSSI, b mancns ie ORE OS Enc. ENG RL COT 2 CUO COREE ALD a een eR eRe ere 163 LoraNnpE Loss Wooprurr AND Hope Spencer. Studies on Spathidium spathula. I. The structure and behavior of Spathidium, with special reference to the capture and ingestion of its prey. Eight text figures andvonerplate, (UTES MUNE GO TWENEY) ..22nc 25 ss acces oo o's Seo os Glecie see 189 M. F. Guyer. Studies on cytolysins. III. Experiments with spermato- GORITUS See OVE GIT CUCM eae) ad tess, Ae vek peste od 15 5p OVS ae sishetiss to dressers Scere vse ab aecle Sy ey si 207 Bessie Noyes. Experimental studies on the life history of a rotifer repro- ducing parthenogenetically (Proales decipiens)..................000008- 225 No. 3. APRIL J.S.Nicnouas. The reactions of Amblystoma tigrinum to olfactory stimuli. OUENS TEEATTHDZ Geard An Sele Coorg EES CO REO Cd Ce ae eT aie ca Erte PSY A. FRANKLIN SHULL. Relative nuclear volume and the life-cycle of Hydatina SCLC OME a UCase WN aah xc eisai, heres MAAS Erase oe wens GPs oi shere ce gatats mene 283 JosrePpH Hatt Bopine. Anesthetics and CO, output. I. The effect. of anesthetics and other substances on the production of carbon dioxide byacertaim) Orthopherane civic i@Unes sc. e o-/actinecce oc om aelerneloiae seine 323 H.C. van per Heype. On the respiration of Dytiscus marginalis L. Three HIGEUITE Se Sra aen eae RO ere Payette eke oie ales w CPaT 51 er ene auese aiShel cha Tetetgey een saenOerere 335 iv CONTENTS No. 4. MAY - Wiut1am H. Cour. The transplantation of skin in frog tadpoles, with special reference to the adjustment of grafts over eyes, and to the local specificity of integument. Two text figures and four plates (twenty- LAY OS TEVITEB) nk cals = c.ccce Aamir coe Sete BC oot. & «. Seana an Ee ens. Soe 353 L.S.Sronr. Experiments on the development of the cranial ganglia and the lateral line sense organs in Amblystoma punctatum. Ninety figures... 421 PROMPT PUBLICATION The Author can greatly assist the Publishers of this Journal in attaining prompt publication of his paper by following these four suggestions: 1. Abstract. Send with the manuscript an Abstract containing not more than 250 words, in the precise form of The Bibliographic Service Card, so that the paper when accepted can be scheduled for a definite issue as soon as received by the Publisher from the Editor. 2. Manuscript. Send the Manuscript to the Editor prepared as described in the Notice to Contributors, to conform to the style of the Journal (see third page of cover). 3. Illustrations. Send the Illustrations in complete and fin- ished form for engraving, drawings and photographs being pro- tected from bending or breaking when shipped by mail or express. 4. Proofs. Send the Publisher early notice of any change in your address, to obviate delay. Carefully correct and mail proofs to the Editor as soon as possible after their arrival. By assuming and meeting these responsibilities, the author avoids loss of time, correspondence that may be required to get the Abstract, Manuscript and Illustrations in proper form, and does all in his power to obtain prompt publication. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 35, No. 1 JANUARY, 1922 Resumen por la autora, Mary J. Hogue Una comparacién entre una amiba, Vahlkampfia patuxent, y las células de un cultivo de tejidos. Los fibroblastos y células blancas de la sangre de embriones de pollo fueron cultivados en un medio de cultivo de tejidos y com- parados con la amiba Vahlkampfia patuxent. Las amibas fueron introducidas en los cultivos de tejidos, estudiaindose las reacciones de ambas clases de células bajo la accién de los colorantes vitales. Las amibas son cuatro veces mas sensitivas que las células del cultivo. Se compararon las mitocondrias de las células. Los granos de melanina de la retina del embrién fueron ingeridos por las amibas y las células del tejido cultivado. La autora describe el método de ingestion de dichas particulas por la amiba. Son expelidas sin digerir. Las células del tejido murieron sin haber expelido los grdnulos de pigmento. La motilidad se considera como un criterior de la vida de la amiba durante el estado de trofozoito, mientras que en el caso de la célula cultivada su reaccién hacia los colorantes vitales se ha usado para comprobar si esta todavia viva ossi ha muerto. Translation by José F. Nonidez Cornell Medical Collgee, New York AUTHOR’S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, DECEMBER 12 A COMPARISON OF AN AMOEBA, VAHLKAMPFIA PATUXENT, WITH TISSUE-CULTURE CELLS MARY JANE HOGUE Department of Medical Zoology, School of Hygiene and Public Health, Johns Hopkins University, and the Department of Embryology of Carnegie Institution of Washington THREE FIGURES INTRODUCTION While working with tissue cells grown in culture medium I was much impressed with the similarity of their appearance to that of amoebae grown on agar-agar plates, and determined to make a comparative study of these two kinds of cells. The particular cells used for this study were the fibroblasts and the white blood cells from the embryonic chick and an amoeba, Vahlkampfia patuxent (Hogue, ’21), which is a salt-water form, parasitic in the digestive tract of the oyster. The tissue cells were grown at 39° in hanging drops of Locke- Lewis solution (Lewis and Lewis, 715) inverted over hard vaselin rings on depression slides.!_ The amoebae were taken from agar- agar plates made up with 0.7 per cent sodium chloride solution, with 0.4 per cent peptone, and kept at room temperature. The comparison of these two kinds of cells was made by study- ing permanently stained preparations, by introducing the amoebae into the cultures where the tissue cells were growing and here studying their different reactions to various vital stains and pigment granules and by further testing out the effect of these stains and pigment granules on the different kinds of cells in their own culture media. 1 Mrs. Lewis kindly furnished me with tissue cultures of connective tissue and spleen. 1 2 MARY JANE HOGUE MORPHOLOGY AND MOVEMEN © At first the most noticeable difference between these cells is their diversity in size. The amoebae averaged 21 yu in length and were much smaller than the fibroblasts which measured 50 win length. ‘The exact size of the fibroblasts is hard to deter- mine, as their processes are so thin and fine that their farthest limits are difficult to determine. ‘The amoebae always have a definite outline, though they are continually changing shape. Still another difference is that the amoebae can be observed to move rapidly, while the movement of the fibroblasts is too slow to be noticeable, though they do move as can be seen by watching the growth of the culture from day to day. The amoeba moves by lunging in one direction and then in another. The cytoplasm is thicker and denser than in the fibroblasts where the processes are very thin and flat against the cover-slip. In order to study the two kinds of cells together, the amoebae and the bacteria growing with them were introduced by means of a platinum loop on the cover-slip where the tissue cells had been growing for from twenty-four to forty-eight hours. ‘The amoebae began at once to crawl actively around and over the fibroblasts without apparently disturbing them. The tissue cells had to be kept at 39°C. The amoebae ordi- narily live at room temperature, though, as J have already shown (Hogue, ’21), they can be kept at 35°C. for twenty-four hours, after which they showed a decided tendency to encyst some time during the following week, the encystment not following immedi- ately on their removal from the high temperature. When the amoebae were brought into the warm box, where the cultures were examined microscopically, their rate of motion was greatly increased. ‘They were now moving rapidly, and continued to do so as long as they were kept at this high temperature. REACTION TO BACTERIA The presence of the bacteria, which were chiefly non-patho- genic bacilli serving as food for the amoebae, was very harmful to the tissue cells. They evidently could not stand the toxic effect of the bacterial waste products. In a few instances some AMOEBAE VERSUS TISSUE CELLS 3 of the fibroblasts lived twenty-four hours after the introduction of the amoebae and the bacteria into the tissue culture, though most of the fibroblasts pulled in their processes and contracted into small oblong masses within a few hours after the introduction of the amoebae and bacteria. Later they went to pieces, while the amoebae moved about quite normally in this new medium (Locke-Lewis), and lived for over two weeks, when the experiment was discontinued. REACTIONS TO STAINS Neutral red In trying out the effect of vital stains on the cells it was found that a very weak solution of neutral red was sufficient to give an excellent stain. When neutral red 1 to 10,000, the proportion used for tissue-culture cells, was used with the amoebae, they would immediately round up, stain red, and die. This solution was sometimes allowed to dry on a cover-slip, which was then put down on the drop of Locke or normal salt solution containing the amoebae. The neutral red would gradually diffuse into the medium and stain the neutral-red granules and vacuoles of the amoebae without immediately killing them. Experiments with neutral red of different strengths were made; 1 to 40,000, 1 to 60,000, and 1 to 80,000 in Locke solution were all good and not sufficiently strong to kill the animals. Amoebae were kept in hanging drops of Locke solution with neutral red 1 to 40,000 for nine days; at the end of this time they were moving slowly and had a few neutral-red granules. In another set of experiments the amoebae were kept in Locke solution with neutral red 1 to 80,000 for twenty-four hours. At the end of this time most of them had lost the stain, so Locke with neutral red 1 to 60,000 was added, and they continued active with stained neutral-red gran- ules for seven more days, when the experiment was discontinued. This is especially interesting, as in tissue cultures the disappear- ance of the color from the neutral-red granules was always taken as an indication that the cells were dead. With the amoeba after the color has disappeared the granules can be restained with neutral red and the amoeba is still normally active. 4 MARY JANE HOGUE The disappearance of the neutral red from the amoeba seems to be due to oxidation. One small amoeba was watched for five hours. At 9:00 in the morning it contained fifteen neutral-red granules. The color gradually faded from these granules until by 11:30 only one granule had the neutral-red color and by 12:00 this color had disappeared. The granules themselves did not dissolve, but could be still seen in the amoebae. Unfortunately, janus black no. 2 had been also used, and this eventually kills the cells. However, at 4:30, this particular amoeba was still Fig. 1 Amoeba showing neutral-red granules of different sizes and neutral- red vacuoles containing one or two neutral-red granules. 12.5 ocular, 1.9 oil immersion. very active with numerous stained mitochondria, but the next morning it was dead. The neutral-red granules and vacuoles were of varying size andnumber. Some amoebae had many small granules and many large vacuoles, which contained from one to two granules (fig. 1). In the fibroblasts one frequently finds neutral-red channels, but they have never been observed in the amoebae. Brilliant cresyl blue 2 b The amoebae were much more sensitive to this stain than were the tissue-culture cells. One drop of a weak solution was added AMOEBAE VERSUS TISSUE CELLS 5 to the tissue culture. The fibroblasts responded at once in the usual way (W. H. Lewis, ’19) to this stain. Many of the vacuoles stained pink with one or more purple granules. The vacuoles were grouped around the nucleus and centriole, just like the neutral-red granules. This same solution applied to the amoebae stained the cytoplasm a pale blue, the granules a purple blue, and the karyosome of the nucleus a purple blue with a clear area in the center. The nucleus with its karyosome and clear area were pulled away from the endoplasm out into the ecto- plasm. The amoeba moved slowly and sent out many large clear pseudopodia, a kind of lunging movement similar to blebs in tissue cells. Within half an hour the granular endoplasm had contracted into a solid blue mass of irregular shape, the nucleus remaining in the ectoplasm. Motion had stopped and the amoeba was dead. This solution of brilliant cresyl blue was diluted one-half with Locke solution and applied to the amoebae. This time about half of the endoplasmic granules took the deep blue stain, the other half remained unstained and were quite refractile. The karyosome stained purple as before, but with a bluish purple area around it and the cytoplasm was a pale blue. The amoebae moved actively, but in a short time the granules lost their color and the nucleus and cytoplasm stained more deeply and the amoebae were dead. When the solution was again diluted one-half and added to a new hanging drop of amoebae, the effect was the same as that on the tissue cells with a solution four times its strength. The nucleus was not stained. The endoplasm contained purple granules as well as the pink vacuoles with their purple granules. The amoebae moved rapidly and progressed normally. Methylene blue The solution of methylene blue, when added to the tissue cells, did not stain the nucleus nor cytoplasm. The vacuoles stained blue and the one or more granules within them took a deep blue stain. Blebs were forming on many of the cells. The same solution added to the amoebae stained the karyosome a deep 6 MARY JANE HOGUE blue, the granules a deep blue, and the cytoplasm blue. The nuclear area around the karyosome remained clear and the amoebae moved freely. Later they rounded up into blue balls surrounded by a lighter blue ectoplasmic area. In this area the nucleus was usually found away from the endoplasm. The amoebae were dead. When the solution of methylene blue was diluted one-half, the effect on the amoebae was the same, though the animals did not die so quickly. When it was diluted one-fourth of the original strength, it stained the amoebae the same as the original solution had stained the tissue-culture cells. The cytoplasm was clear, the nucleus unstained and invisible except as a clear area in the endoplasm free from granules. Many of the granules were colored blue, though not all took the stain. To this stain also the amoeba cell is four times as sensitive as the tissue-culture cell. MITOCHONDRIA The amoebae were treated with janus black no. 2. A very weak solution was used, as otherwise the amoebae would contract and die without showing the mitochondria at all plainly, though the nucleus stained well. When the amoebae stained slowly the mitochondria appeared as small short rods, like bacilli, in among the neutral-red granules (fig. 2). Some of the mitochondria were round, and frequently after they had been stained for some time groups of three or four small mitochondria would be found in among the larger ones (fig. 3). No long branching mitochondria were found in the amoebae, though they are very common in the fibroblasts. The mitochondria retained the stain and were visible as long as the amoebae were alive. This stain seemed to have a stupefying effect, as the amoebae died with their pseudopodia extended instead of rolling up into balls. The mitochondria, when the amoeba is dying, have very marked brownian movement. In the later stages of death the mitochondria move out into the clear ectoplasm which before has been free from granules of all kinds. This does not occur in the tissue-culture cells, as I have already shown (Hogue, 719). “J AMOEBAE VERSUS TISSUE CELLS REACTION TO MELANIN PIGMENT GRANULES In studying the amoeba’s reaction to melanin pigment gran- ules, I first introduced the amoebae into tissue cultures of the spleen which had been growing with the melanin pigment from the retina of a chick embryo for from twenty-four to forty-eight hours. Later I simply put the granules into hanging drops of 0.7 per cent normal salt solution with the amoebae. As soon as the amoebae recovered from being transferred and were moving about with their large ectoplasmic pseudopodia, they began to take in the pigment granules. The amoebae reacted differently to these granules which were in great numbers Fig. 2. Amoeba stained with neutral red and janus black no. 2. The neutral- red granules are clear, the mitochondria are solid black. 12.5 ocular, 1.9 oil immersion. Fig. 3 Amoeba stained with janus black no. 2. The mitochondria are oblong and a few small ones are arranged in groups. 12.5 ocular, 1.9 oil immersion. and were showing brownian movement. Some of the amoebae at once took in large numbers of the granules, others only a few, and still others moved about freely among the granules without taking in any of them. There was evidently some difference in the physiological condition of the amoebae at that time. Many of the granules were oblong, so that it was easy to deter- mine their movement in the cells. ‘They always entered by way of the anterior pseudopodium, whether it was one large pseudo- podium or two smaller ones. The amoeba took in the granules in the same manner that it takes in its food. The granules had to be in the same plane as the advancing pseudopodium. If they were above it or below it the amoeba simply went past it 8 MARY JANE HOGUE without taking it in. The granules could lie in any position in relation to the advancing pseudopodium. They were horizontal and perpendicular to it and at times formed an angle with it. Often they were swung around from a perpendicular to a horizon- tal position and vice versa, and sometimes they swung from a perpendicular or horizontal position to a slanting one and vice versa as they entered the ectoplasm. Once in the ectoplasm their position was constant until they reached the endoplasm. As soon as they were there, they were whirled about by the endo- plasmic currents. As a rule, only one granule was taken in at a time, though I have seen a group of three go in at once, and frequently two gran- ules will be taken in at different parts of the pseudopodium at the same time. The amoeba is capable of taking in a large number of these granules. Four hours after the pigment granules were introduced into the hanging drop of amoebae, the amoebae were filled with the granules. They had at least a hundred granules and resembled the pigment cells of the retina, except for the fact that they were moving rapidly. The amoeba takes the granules in by simply pushing against the granules with such force that the granule pierces the ecto- plasm. For this reason the granule to be taken in must be in the path of the advancing amoeba. If it is to one side of the anterior pseudopodium it is passed by; if it is near the side of the pseudopodium, but still anterior to it, it may be taken in or it may be pushed off farther to the side, depending on how far to the side it is and also on how rapidly the amoeba is moving. If it is moving very fast the granules near the outer edge of the anterior pseudopodium are much more likely to be pushed away than to be taken in, or they may remain attached to the amoeba like the spines of a sea-urchin and be passed back along the edge until they reach the posterior end. Here they frequently remain attached for a long time, but are never taken in at this point. When the amoebae have been in a medium full of these melanin granules for some time many of them are found carrying at least fifty or more granules on their pointed posterior ends. They fairly bristle with the pigment granules at this point. AMOEBAE VERSUS TISSUE CELLS 9 After the granules have entered the endoplasm, they circulate freely through it. Sometimes two or more will unite and occa- sionally a vacuole will form about these fused granules. Some- times there will be as many as three or four clumps of granules in one vacuole. These vacuoles are heavy and lag behind in the posterior part of the amoeba. Eventually they come in contact with the outer edge of the amoeba. After that they soon break, discharging the pigment granules to the outside. Some of the eranules remain attached to the amoeba, others, more forcibly ejected, float away in the medium. In this way the amoeba gets rid of the melanin pigment. After twenty-four hours most of the amoebae had given off the granules and were moving about freely among the granules without taking them in again, at least not in such large quantities as when the granules were first intro- duced into the hanging drop. The amoebae never digested the melanin pigment. Small mononuclear blood cells were present in large numbers in the tissue cultures of spleen. These the amoebae took in, carried them about for a while, and then gave them off. Several of them would often remain sticking to the posterior end of the amoeba along with many pigment granules. All this weight did not seem to retard the movement of the amoeba. These results are somewhat similar to those obtained by Smith (21) with tissue-culture cells, except that the tissue cells take in the melanin pigment at all parts and the pigment granule must always be horizontal to the surface of the cell where it is entering. He notes the formation of vacuoles, but could not determine the ultimate fate of the granules, as the life of the tissue-culture cells is short, and they died without discharging or digesting the granules. DEATH Ordinarily when the nucleus and cytoplasm of a tissue-culture cell have taken methylene-blue or brilliant cresyl-blue or neutral- red stain, the cell is said to be dead. Again in tissue-culture cells, when the color of the neutral-red granules fades and they become clear, the cell is thought to be dead (Lewis, 719). With 10 MARY JANE HOGUE the amoeba a different standard must be used. In one experi- ment already cited the amoebae were treated with neutral red and with janus black no. 2 at 9:30 in the morning. The neutral red color had disappeared at 12:00, noon, but at 4:30 in the afternoon the amoebae were still moving actively. The next day they were dead. This is also true of both brilliant cresyl blue and methylene blue. The amoeba moves rapidly after its nucleus and cyto- plasm have both become stained. It is evident, therefore, that these old criteria of the death of a cell cannot be applied to an amoeba which still continues to move. Indeed, one can wonder whether the tissue-culture cells are really dead when their nuclei and cytoplasm have taken these stains. When these amoebae are alive they are actively sending out pseudopodia. Usually these come from only one part of the cell at a time, though any part may send them out. The amoebae not only send out pseudopodia, but they move rapidly. As the animals begin to die, progression is less rapid, though the pseudo- podia are formed very quickly. Then gradually the pseudopodia are formed more slowly, then only at one end of the amoeba, and eventually they are simply blebs which flow from one side of the amoeba to the other until finally all motion stops and the amoeba is dead. From the fact that locomotion and motion become gradually less and less as the amoeba is dying, it seems evident that the motion of an amoeba and not its reaction to vital stains should be the ultimate criterion of life in the trophozoite stage. AMOEBAE VERSUS TISSUE CELLS 11 BIBLIOGRAPHY Hoguz, M. J. 1919 The effect of hypotonic and hypertonic solutions on fibro- blasts of the embryonic chick heart in vitro. J. Exp. Med., vol. 30, p. 617. 1921 Studies on the life-history of Vahlkampfia patuxent n. sp. parasitic in the oyster, with experiments regarding its pathogenicity. Amer. J. Hygiene, vol. 1. Lewis, W. H. 1919 Degeneration granules and vacuoles in the fibroblasts of chick embryos cultivated in vitro. Johns Hopkins Hosp. Bul., vol. 30, no. 1. Lewis AND Lewis 1915 Mitochondria and other cytoplasmic structures. Am. Jour. Anat., vol. 17, p. 339. Smitu, D. T. 1921 The ingestion of melanin pigment granules by tissue cells. Johns Hopkins Hosp. Bull., vol. 32. Resumen por el autor, Theophilus 8. Painter Estudios sobre la espermatogénesis de los Mamiferos. I. La espermatogénesis del opossum (Didelphys virginiana) El presente trabajo ha sido emprendido con el fin de aclarar los resultados contradictorios sobre el numero de cromosomas obtenidos por Jordan y Hartman. El autor ha hallado 22 cro- mosomas en las espermatogonias. Los dos mds pequefios con- stituyen un complejo cromosémico X-Y tipico, que puede obser- varse durante la maduracién. En la primera divisién madurativa existen en el huso 11 cromosomas, diez de los cuales son tetradas y el otro el cromosoma X-Y. Los elementos X e Y se separan durante esta division de tal modo que los espermatocitos secundarios llevan 10 autosomas més X o 10 autosomas mas Y. Ambos tipos de cromosomas sexuales se dividen ecuacional- mente en la segunda divisién de maduraci6n. Las células somaticas de embriones, en' vias de divisi6n, presentan en todo caso 22 cromosomas, pero en el caso de-los machos el complejo X-Y existe y en las hembras encontramos 2X. El autor llega a la conclusién que Jordan no pudo hallar mas de 17 cromosomas espermatogoniales a consecuencia de la fusion de los cromosomas en su material. El ‘‘cromosoma accesorio”’ descrito por dicho autor es probablemente una tetrada desplazada. El error de Hartman sobre el ntmero de cromosomas se debe a una divisién longitudinal de una de las tetradas en el huso del segundo glébulo polar. Translation by José F. Nonidez Cornell Medical College, New York AUTHOR'S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, DECEMBER 12 STUDIES IN MAMMALIAN SPERMATOGENESIS I. THE SPERMATOGENESIS OF THE OPOSSUM (DIDELPHYS VIRGINIANA)! THEOPHILUS S. PAINTER EIGHT TEXT FIGURES AND THREE PLATES (TWENTY FIGURES) INTRODUCTION The opossum, along with many other mammals, is a form for which the diploid or somatic chromosome number is a matter in dispute. In 1911 Jordan gave an account of the spermatogenesis of this animal. His results as they touch the chromosomes are briefly as follows: 17 chromosomes were found in dividing sper- matogonia; in the primary spermatocytes 9 chromosomes were present—one of these was bipartite in form and passed apparently undivided to one pole of the spindle; in part of the second sper- matocytes this bipartite element was found, in others it was lacking. Because of its behavior and distribution, Jordon inter- preted this bipartite chromosome as the true sex-chromosome, and from his description it is clear that one half of the sperm carry this element and one half lack it; in other words, that the male opossum possessed the X—O type of sex-chromosome. No further work upon the male opossum has appeared except a short note by the author (Science, May 27, 1921) in which the main results of this paper were outlined. Recently, Hartman (’19) working upon the early stages of opossum embryos, reports that in the ova (polar-body spindles) he finds from 10 to 12 chromosomes. Since 12 chromosomes are found in several spindles where the chromosomes are most favor- ably placed for counting, Hartman concludes that this is the true reduced or haploid number. From this it would follow of 1 Contribution no. 150, from the Department of Zodlogy, University of Texas, Austin, Texas. 13 14 THEOPHILUS 8. PAINTER course that the diploid or somatic chromosome number for the female opossum is 24.? The discrepancy between Jordan’s and Hartman’s counts amounted to 6 chromosomes—a number far too large to be explained on the basis of differences in the sex-chromosome complex. The author undertook the present work with a view of deter- mining, if possible, the cause for the difference in the chromosome counts of the two investigators whose work has just been cited. The problem, however, is not simply a matter of chromosome numbers, but one which has a broad and very important bearing upon the whole subject of spermatogenesis and sex-determination in the mammals. For the discrepancy in the works upon the opossum is typical of the confusion in the literature over the number of chromosomes for many other mammals; in fact, there is less difference in the counts for the opossum (6 chromosomes) than for such forms as the pig (where 18 and 40 have been given as the diploid numbers), or for the male of the human species, which has been variously reported to have 22, 24, and 47 chro- mosomes.’ It is clear that, as long as the total number of chro- mosomes possessed by an animal is a matter of doubt, we cannot safely accept any conclusions regarding the sex-chromosomes of that form. In a paper on the spermatogenesis of lizards, the author (21a) has shown that the presence of a bipartite body at one end of a spindle in maturation divisions does not neces- sarily mean that it is a sex-chromosome as has been so often assumed in vertebrate spermatogenesis. It is frequently either a whole tetrad or half of a tetrad which has divided, the other half remaining in the equatorial plane of the spindle. In the present paper (p. 30) a number of such false sex-chromosomes 2 Hill (17), working on the South American opossum, Didelphys aurita, estimated that there were 12 (haploid) chromosomes in the egg of this species. 3’ Wodsedalek (13), gives 18 as the diploid chromosome number for the pig, while Hance (’17) shows that there are 40 chromosomes in both the germinal and somatic cells. Among the recent counts for the male of Homo sapiens we have the following: Guyer (710), 22; Montgomery (’12), 23 or 24; von Winiwarter (712), 47; Wieman (717), 24. MAMMALIAN SPERMATOGENESIS—OPOSSUM 15 are shown (text fig. 7). It is not to be doubted but that some of the ‘accessory chromosomes’ reported by different investigators of mammalian spermatogenesis are of this false type. In addition to the question of chromosome number and of sex- determination in the opossum, a number of points of rather special interest to cytologists are touched upon; chief among these is the phenomena of so-called ‘double reduction’ first described by Guyer (’09) and reported for the opossum by Jordan. MATERIAL AND METHODS The present study is based upon the testes of four opossums and upon the dividing somatic cells of eight embryos. In Texas the breeding season for the opossum begins in Janu- ary. The males, upon which this study was made, were oper- ated on as follows: one male on October 8, 1920; one male on November 15, 1920, and two males on January 14, 1921. Mature sperm, dividing spermatogonia, and cells in all stages of matura- tion were found in all the males but the maturation stages were more abundant in the individuals whose testes were preserved in January. In the first male, an attempt was made to secure fixation by injecting the fixing fluid into the blood system, as Allen (’19) recommends. Only partial success attended the use of this method. Spermatogonial chromosomes (fig. 1) show up with sharpness, but the first and second spermatocyte divisions are badly masked by chromosome fusion. This may be due to the ether used in anaesthetizing the animal. In the other cases, no ether was used. The males were tied down, the scrotal sacs cut open with a razor and the testis removed. The testis was cut into several pieces with scissors, and then one or more of these pieces dropped into the fixing fluid, and the tubules quickly teased apart, so that complete penetration would be secured. A period of less than thirty seconds elapsed between the time the tubules were receiving blood from the animal and were being bathed in the fixing fluid. Two preserving fluids were used. One was modified Bouin’s solution, suggested by Allen (’16, 719), warmed to about 38°C. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 35, NO. 1 16 THEOPHILUS S. PAINTER The other was the cold Flemming method, with urea, as recom- mended by Hance (717). The modified Bouin’s fluid gave superb preparations for all stages of maturation, and was the more satisfactory of the two preservatives. However, it is my experi- ence that for spermatogonial and somatic divisions, no preserva- tive can surpass the results obtained by cold Flemming to which urea is added. A great deal of care was used in handling the material after preservation. When modified Bouin was used, the suggestions of Allen (19) for washing and dehydration were closely followed. As an agent for securing rapid mixing of fluid, I used air which 4For the convenience of the reader, the exact technique employed is given below. The modified Bouin fluid is made up as follows: A. Picric acid, saturated aqueous solution................. 75 ce. BOT Ol: ) Gee a Pe 85 SBS on boo cree ae 25 ce. Glagial aeenre ACO. esas ee oh as ss ye se ee 10 ce. BMEChromicracie "Cry Stalssemne eee ae fe eo eee: See ee 1.5 grams Urea; aryatalls: (ii; fee ete Be, Gaited: Co. CR eas 2.0 grams The ingredients under A may be mixed and kept as a stock solution. Just before using the chromic acid and urea should be added, in the order named, and the solution heated to about 38°C. In operating for the removal of tissue, the fixing fluid is kept in a water-bath on the same table. Immediately on removal, the testis is slit open and a small mass or two of the tubules are cut out with scissors and dropped into the fixing fluid. The tubules are quickly teased apart so that each of them will be exposed to the action of the fixative. Fixation at 38°C. is carried on for about an hour. The tissue is then placed and washed in 5-per cent alcohol. Using the drop method throughout (an ordinary burette is useful), the tissue is brought through the following solutions: 10 per cent alcohol, to which a few drops of a saturated solution (aqueous) of lithium carbonate has been added. It is kept in this solution until most of the picric acid has been removed. Then 50 per cent alcohol is added until the strength of the solution is about 35 per cent. Then a mixture of equal parts of aniline oil and 50 per cent alcohol is dropped in. This is fol- lowed by a similar mixture of 70 per cent alcohol and aniline oil. Next, pure aniline oil is added and the tissue cleared in this. The aniline oil is replaced by wintergreen oil. From this it is embedded in paraffin, but the tissue is passed through some six or eight changes in which there is an increasing proportion of paraffin. The time taken for embedding is around three hours. In the second method, urea is added to ordinary strong Flemming solution to the extent of 0.5 gram per 100 cc. The solution is then chilled with ice to about 4°C. The tubules are teased out in this fluid and kept in it on ice for twenty- four hours. Subsequently it is washed and treated in the ordinary way. From 95 per cent alcohol one may pass to oil of cedar (using the drop method), then zylol, and finally paraffin. MAMMALIAN SPERMATOGENESIS—OPOSSUM 17 was forced to bubble through the solutions containing the material. After using Flemming, Hance’s suggestion of clearing with cedar oil directly from 95 per cent alcohol was followed. The embryos employed were both unborn and pouch young. Only the cold Flemming method was used for preserving this material. Embedding was carried on in the usual way and sections cut at from 6 to 8u. Iron-haematoxylin was the method used for staining. The author is indebted to Dr. Carl Hartman for much of the living material used for the present work, and also for the priv- ilege of examining some of his preparations of polar spindles, which will be found figured on page 36. SPERMATOGONIAL DIVISIONS Dividing spermatogonial cells, in equatorial plate view, show 22 chromosomes (figs. 1 to 6, see also somatic divisions, fig. 5, on page 26). Of these 21 are elongated rods, of various sizes, bent or shaped in characteristic ways, and one small rounded chro- mosome. There are no typical V-shaped chromosomes in the opossum complex. Usually the twenty largest chromosomes form a ring about the two smallest (figs. 1 to 4, also somatic divisions), but occasionally this condition does not obtain (figs. 5 to 6). On inspection of the figures it will be noted that there are a number of pairs of chromosomes, similar in size and shape; and, further, that the two chromosomes lying in the center of the spindle are decidedly smaller than the remaining twenty elements. This fact is important and is shown in all figures (figs. 1, 2, 3, 4, and 6) where there is no great foreshortening of the chromosomes. It is brought out, best, perhaps, by arranging the chromosomes of any spindle in their approximate size relations. In text figure 1 the chromosomes of the cells shown in figures 2, 4, and 6 are copied by the aid of a copy camera lucida, and arranged in approximately the order of their size. It will be seen, first, that the twenty larger chromosomes can all be paired up, with more or less accuracy, and, secondly, that the two smallest chromo- somes not only are decidedly smaller, but that they have no 18 THEOPHILUS S. PAINTER synaptic mates of the same size. These two smallest elements are, as will be shown, the sex-chromosomes, the rounded element being the Y-component and the bent rod the X-component. In o> és, a ae as = ad & x ee me Res. & i = aE — — we @s a @u oe: tee ea Cae. pel Bre ln SEP ae Se Pe EE, ee Pro SER ae &*. eu P= es - = ve ae. eee Clad es ere ¢ Aes Text fig. 1 Showing the chromosomes of three spermatogonia, arranged in approximately the order of their size. The X- and Y-chromosomes are labeled. figures 1 to 6 the Y-element appears as a rounded body, very conspicuously smaller than the other chromosomes. The X- component frequently shows one end of the rod larger than the MAMMALIAN SPERMATOGENESIS—OPOSSUM 19 other (fig. 4) or it may even appear as two unequal egg-shaped chromosomes joined as in figure 1. This inequality in the two ends of the X-element persists throughout maturation. The autosomes are all more or less rod-like, but in some cases there is a tendency for the inner end of a chromosome to be smaller than the outer; the inner end may even be bent at an angle to the rest of the chromosome. The cells figured (1 to 6) are taken from three different males; there is no appearance of any fusion of the chromosomes and little overlapping (fig. 1, 2, or 4). In some cells there is a tend- ency for certain chromosomes to occur together, either lying side by side (fig. 6) or one partly on top of the other (fig. 5). In poorly preserved material such chromosomes would doubtless appear as one element. In addition to the counts figured in plate I a great number of other drawings of dividing spermatogonial cells were made by the aid of a camera lucida, and counts then verified. In every case 22 was found to be the number of chromosomes present. The division of spermatogonia offers no point of especial interest, as it proceeds in the usual way. No lagging behind of any elements has been seen. FIRST MATURATION DIVISION The author has made no attempt to follow in minute detail the changes which the chromosomes undergo from the early growth period to the time when they enter the first maturation spindle. Jordan has clearly figured and described the essential facts; my own preparations show stages similar, in the main, to those found by him. A few points illustrated in text figure 2 will be touched upon. The so-called ‘diffuse stage’ (text fig. 2, A) is characterized by three nucleoli, which lie scattered among the faintly staining chromatin knots. The two smaller nucleoli are unequal in size. With the appearance of the leptotene threads, some deeply staining areas are found in the nucleus, which probably represent one or more of these nucleoli. The pairing of the leptotene threads is seen with almost schematic clearness in the opossum. In text figure 2, B are three pairs 20 THEOPHILUS S. PAINTER of threads, drawn as they lie in one nucleus. We have three steps of the fusion illustrated here—the threads lying more or less together, a pair of threads in which partial fusion has taken place and a pair of threads in which all but the ends have joined to- Text fig. 2. Various stages taken from the growth period of the first sperma- tocytes. gether. The single leptotene threads are distinctly knotted, and in synapsis these knots join together. There is some contraction (synezesis) of the nuclear content at this time, but it is not very marked. The ‘bouquet stage’ (text fig. 2, C) is a striking fea- ture of many of my slides. In addition to the thread loops, one MAMMALIAN SPERMATOGENESIS—OPOSSUM IAS often sees clearly the deeply staining chromatin masses labeled X and Y. Following this period, the chromatin threads expand until they fill the entire nucleus and the chromosomes become woolly in appearance (text fig. 2, D and E). The whole cell remains in this phase for a considerable time. In early stages (text fig. 2, E) three nucleoli are found, but the homologies of these with those described in the early growth period (text fig. 2, A) arenot clear. As this stage advances, however, two of the nucleoli seem to disintegrate, they swell, lose their affinity for the stain, and eventually disappear from view. Some indication of this may be seen in text figure 2, E. The third nucleolus persists and frequently has a form (text fig. 2, D) which suggests that it is made up of the X and Y chromosomes lying side by © side. The contraction of the diplotene threads follows the dis- appearance of the two nucleoli, and the tetrads of the first matura- tion spindle take on their characteristic forms. In the first maturation division spindles, there are 11 chromo- somes (figs. 7 to 11). It has proved difficult and unsatisfactory to determine the number of chromosomes present from ordinary equatorial plate views. The reason for this is not because of a fusion of the elements, but because of the shapes of the chromo- somes themselves, as a glance at the figures will show. In side views of the spindles, particularly just as the tetrads are entering the spindles, and in slides in which destaining has been carried out to an advanced degree, one can usually make out ten typical tetrads and an eleventh chromosome which is tripartite in form, that is, made up of two egg-shaped elements, and a blunt rod joined end to end. This is the X-Y chromosomes complex. (In fig. 10 two of the chromosomes, which lie underneath the rest are drawn out at one side.) The ten tetrads have such characteristic forms, that after experience one can usually identify all of them and determine which are lacking, in case the spindle has been cut in two. In text figure 3 the typical shapes of these tetrads, as taken from two cells (figs. 7 and 10), are shown. A comparison of the chromosomes figured in text figure 3, with the spindles in figures 7 to 12, will enable the reader to identify most of the elements a2 THEOPHILUS S. PAINTER for himself. Once one learns to know the shape of the individual tetrads, then polar views of equatorial plates are intelligible and in all complete spindles 11 chromosomes are found.® The history of the sex-chromosome complex is given in the adjoining text figure 4. A close inspection of this complex (labeled XY, figs. 7, 10, 12 and text figs. 3 and 4) shows that it is made up apparently of three elements arranged end to end. (Only two spermatogonial chromosomes are involved.) There are two egg-shaped masses, the X-element and a blunt rod which is the Y-chromosome. The egg-shaped mass lying distal to the equatorial plate is somewhat larger than the other similar body. RRAS Del Ld $44 dq por Text fig. 3 The chromosomes of the first maturation spindle. All are tetrads except the X-Y elements. The Y-chromosome may appear as bulged at one end (figs. 7 and 10), but more frequently it is simply an el chromatin (text fig, 4, A to E). The behavior of the X-Y chromosome compl in figures 7, 8, 10, 12 and in text fig.4. The time o be early (text fig. 4, A, B, and C) or late when these elements are found lagging behind in the spindle (text fig. 4, D and E). In either case the plane of division is such that the egg-shaped ele- 5 The fact that mammals have elongated tetrads in the first maturation divi- sion (the author has observed them also in the bat and the striped skunk) un- doubtedly explains the great difficulty which has been experienced in making counts at this period. MAMMALIAN SPERMATOGENESIS—OPOSSUM 23 ments, or X-chromosome, go to one pole, while the Y-chromosome goes to the other. No exceptions to this have been noted. As a result of the first maturation division, the secondary sper- matocytes all receive 11 chromosomes, but in half of these cells Text fig. 4 Showing various stages in the separation of the X- and Y-chromo- some elements in the first maturation division. we have the X-element and in the other half the Y-chromosome is present. A distinct resting stage follows the first maturation division, as Jordan has described and figured. I fail to find, however, any very striking differences in the nucleoli of such resting cells. 24 THEOPHILUS S. PAINTER Frequently, in first spermatocyte cells, a chromatoid body is found (labeled chr. in fig. 3, text fig. 4, B and D). It usually lies well out of the spindle, and no attempt has been made to follow its history during maturation. SECOND MATURATION DIVISION In equatorial plate views all secondary spermatocytes show the presence of 11 chromosomes (figs. 13 and 17); of these the autosomes are all typical bivalent or dumb-bell-shaped elements. The sex-chromosome varies in appearance, depending on whether we are dealing with the X-component (fig. 13) or the Y-component (fig. 17). The X-element usually appears as a quadripartite (fig. 13) and the Y-element as a short rod (fig. 17) sometimes bipartite. The history of the second spermatocytes is given in figures 13 to 20. In figure 18 we see the X-chromosome as it appears in side views of the spindles, and in figures 16 and 19 we can follow the distribution of this element to the spermatids. It will be noted that each spermatid received a pair of egg-shaped chromo- somes which are joined together. Figure 14 is a polar view of one end of a second spermatocyte spindle (telophase) showing the X-chromosome plus 10 autosomes. When the Y-chromosome is present in the spindle (fig. Lie it divides equationally (figs. 15 and 20), each spermatid receiving a Y-element. Figures 19 and 20 are two cells, which in the testis lie end to end. They are probably the daughter cells of a single first spermatocyte, as no other cells in a similar stage are close to them. In the one daughter (fig. 19) the X-element is seen dividing and in the other daughter (fig. 20) the Y-chromosome is dividing. It seems to have been the common experience of a number of workers on mammalian spermatogenesis that in the second spermatocyte there is a tendency of the chromosomes to fuse in pairs, thus bringing about what has been termed by Guyer ‘double reduction.’ In my slides the fact cannot be denied that some of the chromosomes do fuse in a way which is suggestive MAMMALIAN SPERMATOGENESIS—OPOSSUM 25 of double reduction. ‘This is especially apt to be the case where the second spermatocyte cells lie deep within the tubule. On the other hand, when the second spermatocyte cells lie near the periphery of the tubule, there is no trace of such a fusion. The chromosomes are as distinct and separate as one could wish. THE CHROMOSOME NUMBER OF OPOSSUM EMBRYOS A study of the spermatogenesis has shown that in the male opossum there is a pair of chromosomes which segregate out in the first maturation division and are distributed to the sper- matids in a way comparable to the typical X-Y chromosomes of the insects. From a study of the male alone it could not be defi- nitely stated, for the opossum, which was the X-element, or female- producing component, and which the Y-element, or male-pro- ducing component. In order to clear up this point and to further verify the spermatogonial chromosome counts, a study of the chromosome complex of opossum embryos was undertaken. Five very young embryos of unknown sex were used together with three female embryos, whose sex was previously determined by Doctor Hartman. In all cases cold Flemming solution, with urea, was used as the fixative. The dividing cells principally of the nerve cord were used for making counts. At first the five embryos of unknown sex were studied. Embryo no. 4 showed the condition illustrated in text figure 5, A and B. There are 22 chromosomes in the spindles, two of these are smaller than the rest and are similar to the small ele- ments found in the spermatogonia. Embryo no. 4 clearly pos- sesses the chromosome complex of a male opossum. Embryo no. 5 (text fig. 5, C and D) showed, like no. 4, 22 chromosomes, including the small X-Y elements. Again, we are dealing with an embryo clearly possessing the male chromosome complex. Embryo no. 6 (text fig. 6, K and L) showed consistently 22 chromosomes. The small round chromosome was absent, how- ever, and in its place there is a small elongated element similar to the other sex-chromosome. This condition of the chromo- MAMMALIAN SPERMATOGENESIS—OPOSSUM 27 Text fig. 6 Equatorial plate views of dividing somatic cells of four female embryos. 28 THEOPHILUS S. PAINTER somes was found in all cells of this embryo. The absence of the small rounded chromosome and the presence of a second elongated rod was interpreted as meaning that embryo no. 6 was a female with the 2-X condition. This showed, then, that the small elongated rod was the X-component, and on this basis the rounded chromosome of the male was the Y-component. Embryos nos. 7 and 8 (text fig. 5, E, F and G, H) both turned out to be males. Thus it will be seen that among the first five embryos studied four were found to possess the chromosome complex of the male while only one showed the female condition. Although the study of the chromosome complex was consistent in that it showed all males to have the X-Y condition and the one female always to have the 2-X condition, a further check was deemed desirable, especially as the material was at hand. Doctor Hartman has found that he can identify the sex of the pouch young of the opossum soon after birth, by the presence of the scrotal swellings in the male and the absence of the pouch rudiments and, conversely, in the female the rudiments of the pouch are plainly visible and the rudiments of the scrotal sacs are absent. Doctor Hartman selected for me three embryos which he identified as being of female sex. ‘They were preserved in cold Flemming and sectioned. Text figure 6, M to R, shows the chromosome complex of these three individuals. The figures show that in every case the 2-X condition of the sex-chromosomes is found just as in embryo no. 6 (text fig. 6, K and L) and also that there are 22 chromosomes in all the cells. (Text fig. 6, M and N from embryo no. 10; O and P from embryo no. 11, and R from embryo no. 9.) In none of the eight embryos did I find any evidence for a fragmentation of the chromosome elements, as Hance did in a similar study of pig embryos. Among the hundreds of cell plates examined, I invariably found 22 chromosomes. In em- bryos nos, 4, 5, 7, and 8 the X-Y chromosome condition was observed, and in embryos nos 6, 9, 10, and 11 the 2-X condition prevailed. MAMMALIAN SPERMATOGENESIS—OPOSSUM 29 DISCUSSION 1. Sex-determination in the opossum In the foregoing pages evidence has been presented which seems to demonstrate conclusively that sex-determination in the opossum is of the X-Y type for the male, and not the X-O type, as reported by Jordan. The facts adduced to prove this con- clusion are briefly as follows: In the spermatogonia there are 22 chromosomes (figs. 1 to 6). These can all be paired up except the two smallest which have no synaptic mates; these are the X-Y chromosomes. The X-component is a small, somewhat elongated rod, the Y-element is a single rounded chromosome. In the growth period of the first spermatocytes we find a nucleo- lus which is probably the X-Y components lying side by side (text fig. 2, D). In the first spermatocyte spindles there are 11 chromosomes, showing that all 22 spermatogonial chromosomes have mated (figs. 7 to 11). Of these eleven chromosomes, ten are conspicuous tetrads, while the eleventh is made up of three parts arranged linearly. The figures, however, show that the tripartite chromosome is really the X and Y elements joined end to end. When division occurs the three elements are separated so that the two egg-shaped parts (the X-components) go to one pole, while the single rod-like Y goes to the other. In the second- ary spermatocytes we find in part of the cells the X-element alone, and in other cells the Y-chromosome occurs by itself. In both cases the sex-chromosome divides equationally, so that all spermatids receive 11 chromosomes. Half of the spermatids receive 10 autosomes plus the X-chromosome, and half receive 10 autosomes plus the Y-chromosome. As an additional bit of evidence substantiating the facts outlined above, are the dividing cells of opossum embryos, in which we either have the X-Y con- ditions (males) or the 2-X condition (females). Sex determina- tion for the opossum follows the scheme given below: Sperm contain: Eggs contain: 10 autosomes + X, plus 10 autosomes + X, 10 autosomes + Y, plus 10 autosomes + X, 20 autosomes + 2X (female) 20 autosomes + XY (male) 30 THEOPHILUS S. PAINTER The question now arises, what is the explanation for the bipartite body which Jordan figures, as going undivided to one pole of the first maturation spindle, and which he interpreted as being the X-chromosome? ‘The probable answer is, that what Text fig. 7 Side views of the first maturation spindle, showing displaced tetrads which would appear as true sex-chromosomes in poorly fixed material. The X-Y chromosomes are visible in two of the spindles. Jordan saw was either a displaced tetrad or half of a tetrad, the other half remaining in the equatorial plane of the spindle. Such conditions are not rare in my material. In text figure 7, several cells are figured, in which a tetrad has been displaced and appears MAMMALIAN SPERMATOGENESIS—OPOSSUM at to be passing undivided to one pole. That these bodies are false accessory chromosomes is proved by the fact, first, that the X-Y elements (the true sex-chromosomes) can be clearly seen in most cells, and, secondly, that these false accessory chromosomes usually have the typical shapes of tetrads which reveal their true nature. If the preservation of the material were not so good, these tetrads would appear as more or less bipartite, and would doubtless be interpreted as a true accessory chromosome if the X-Y elements were not in prominence. The cause of this occasional displacement of tetrads, which one observes here and there in the spindles, is probably due to the technique employed (either the sectioning razor or to diffu- sion currents set up during fixation and subsequent treatment of material) and doubtless does not occur in the living cell. 2. Sex-determination in mammals The opossum has long been cited as a mammal in which we have the typical X-O type of sex-chromosome for the male. And since the appearance of Jordan’s work in 1911 many investigators, reporting on the spermatogenesis of other mammals, have de- scribed a body similar in appearance and behavior to the ‘acces- sory chromosome’ found by Jordan, and have interpreted it, as Jordan did, as the true sex-chromosome. It may be added that in most cases no attempt was made to work out the complete history of the chromosomes in the way Jordan had done. And yet as the present work seems to prove clearly the true sex- chromosomes of the opossum are of the X-Y type, and that what Jordan saw was probably a displaced tetrad, or possibly half a tetrad which had passed early to one pole, the other half remaining in the equatorial plane of the spindle. Both of these conditions have been found by the author in his opossum material. In view of these facts, it seems that we may, with all fairness, raise the question, whether or not in some of the other mammals the same error of interpretation has not been made; in brief, whether or not the true sex-chromosome has been found for many forms. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 35, NO. 1 32 THEOPHILUS S&. PAINTER Table 1 gives in a condensed form the results of all workers on mammalian spermatogenesis in which the sex-chromosome has (This list is complete to date as far been figured and described. as known to the author.) A glance at the table will show that both the origin and the form of the sex-chromosome are variable. It is interesting to note that in the guinea-pig, the rabbit, and, one author claims, for man, the X-Y condition has been described, while in all the TABLE 1 DIPLOID TYPE OF FORM oe ee a Ua oun AUTHOR Homo sapiens..... 22 X-O | Ist div. reduc. | Guyer, 710 Homo sapiens...........| 23-24 | X-O | 1st div. reduc. | Montgomery, 712 Homo sapiens...........| 47 X-O | Ist div. reduc. | Winiwarter, ’12 Homo sapiens,......... 24 X-Y | 2nd div. reduc. | Wieman, 717 PGI x8 Ae Paice Wee 21 X-O | Ist div. reduc. | Malone, 718 Cat, domestic... o.-. ....|e 35 X-O | Ist div. reduc. | Winiwarter and Sainmont Armadillotseoee betaine cole X-O | Ist div. reduc. | Newman and Pat- terson, 710 WpossuMi. Noo. eke. ss Kf X-O | Ist div. reduc. | Jordan, ’11 Guines=pigs..38 Shee.) “56? X-Y | Ist div. reduc. | Stevens, ’11 Blab Pie ccaskis cise Sess <3 22 X-Y | Ist div. reduc. | Bachhuber,.’16 Wihitemat.... ) erence 37 X-O | Ist div. reduc. | Allen, 718 House mouse:=.2........| 39 X-O | 2nd div. reduc. | Yocum, 717 Horses... fi) eee ae X-O | Ist div. reduc. | Wodsedalek, ’14 Li) Na ey BEE ee | © a 18 X-O | Ist div. reduc. | Wodsedalek, ’13 LO AL ee ee eae a 37 X-O | ist div. reduc. | Wodsedalek, ’20 rest of the cases we have the X-O type of sex-chromosome. It should be added, however, that a great many investigators have failed to find evidence for sex-chromosomes in many of these mammals for which it is reported.® Unquestionably, the evidence for the existence of the X-O type of sex-chromosome is very complete in some cases, such as in the rat (Allen) or cattle (Wodsedalek), but in many other cases 8 Mention is made only of the papers in which an accessory chromosome has been described and figured. A complete review of the literature will be found in Ethel Browne Harvey’s paper in the Journal of Morphology, vol. 34, 1920. MAMMALIAN SPERMATOGENESIS—OPOSSUM 33 the evidence is not convincing, and in view of the results obtained for the opossum it is to be desired that the other mammals where the old technique was used should be reworked with the newer methods of Allen and of Hance. Only when this is done shall we be able to accept as settled the question of the sex-chromosome complex of any given mammal. 3. Double reduction Is ‘double reduction’ of the second spermatocyte chromosomes a process which normally occurs in spermatogenesis, or may we take the term as being synonymous with poor preservation and technique? As far as the opossum goes, and the author suspects this will hold for all vertebrates, it is perfectly certain that a fusion of the chromosome elements is only obtained when poor preservation is in evidence. In case the material is badly fixed one may get not only double, but complete reduction into one chromatin mass! The following facts are illuminating. In the case of the opossum, it has been found that it was useless to attempt to study any second spermatocyte spindles which did not lie near the periphery of the testicular tubule. If one will observe this precaution, absolutely no evidence for any fusion of the chro- mosomes will be found. On the other hand, if the second sper- matocyte spindles lie deep within the tubule, the chromosomes are all more or less fused, or tangled up so that accurate study of the elements is impossible. It seems clear that the explana- tion for this is a matter of the quickness with which the cells are killed. If they lie on the periphery of the tubule, then they are quickly killed, and their true form retained. But if they lie deeper, so that penetration goes on slowly, then a great deal of distortion and fusion will take place before fixation is complete. It is believed, therefore, that, as a normal process, there is no such phenomenon as ‘double reduction,’ and that when any investigator obtains such a fusion of chromosome elements, he had better take steps to insure more rapid and complete pene- tration of his fixing fluid. 34 THEOPHILUS S. PAINTER 4. The chromosome number for the male and the female opossum It has been shown in the foregoing pages that the diploid or somatic number of chromosomes for both the male and the female opossum is 22, and not 17, as stated by Jordan, or 24, as concluded by Hartman. How are the results of this paper to be reconciled with the chromosome counts of Jordan and of Hartman? The following suggestions are offered. In the case of the males, a careful comparison of Jordan’s figures with my own preparations indicates that Jordan had con- siderable fusion of the chromosomes in his slides. The Y-chro- mosome seems entirely lacking from all his figures, and the X-component is present in only one or two cells. The main error in counting, however, was in the chromosome ring. The fact has already been pointed out in the early part of this paper that in spermatogonia one frequently finds cells in which two, sometimes four, chromosomes lie close together, or even partially overlap. Unquestionably, with fixing fluids less adapted to vertebrate chromosomes than the newer methods, these chro- mosomes would fuse or lie so close together that the most com- petent investigator could not distinguish their separate outlines. Here, it is believed, is the explanation for Jordan’s failure to find more than 17 chromosomes in the spermatogonia. In this connection it is illuminating to consider Hance’s (’17) work upon the pig. Wodsedalek (’13) had reported 18 as the spermatogonial number of chromosomes for the pig, while Hance, using cold Flemming solution with urea, found 40 chromosomes in both somatic and germinal cells. Hance made careful com- putations of the mass of chromatin in Wodsedalek’s figures, and then, after making similar computations for his own preparation, he compared the results. He found that, in Wodsedalek’s figures with 18 chromosomes, there was approximately the same mass of chromatin as in his cells showing 40 chromosomes. While the author has made no attempt to compare his results in this way with those of Jordan, there is little doubt but that the same mass of chromatin would be found in both cases. MAMMALIAN SPERMATOGENESIS—OPOSSUM 35 The fusion which masked the true spermatogonial chromosome number in Jordan’s preparations is doubtless responsible for his failure to find 11 chromosomes in the first spermatocytes. I found it very difficult to make counts in equatorial plate views of the first spermatocyte spindles until after I had practically learned to recognize each chromosome by its size and shape. To make convincing drawings of such views is very difficult. In side views of the spindles, however, under favorable conditions (early phases of the spindle in cells fully differentiated), it was a simple matter to make accurate counts. In his second spermatocyte cells, Jordan reported ‘double reduction’ which does not occur at all in the best of my material. Altogether, it seems clear that Jordan failed to find 22 chromo- somes in the male opossum, because there had been a fusion of some of the chromosome elements in his material. This fusion has been universally met with by all investigators on vertebrate germ cells up to, perhaps, the last five years, and it seems very probable that the chromosome numbers for nearly all vertebrates, as they have been reported before this period, are to be accepted with some reservation. The truth is, the old methods of pres- ervation were simply inadequate to handle the vertebrate chromosomes, and it would be grossly unfair to the pioneer workers in this field to regard their work as carelessly or inac- urately done. The reason Hartman found 12 to be the haploid number of chromosomes is due, as it turns out, to a precocious splitting of one of the tetrads in the second polar spindle of the egg. Figures 13 and 18 of the present paper will show that the X-chromosome splits so early in second spermatocytic division that, if one did not know the origin and fate of the halves, he would interpret them as being two separate chromosomes. And yet, as figures 14, 16, and 19 show, each pole of the cell receives one of these dumb-bell-shaped halves. Before I had an opportunity of examining Doctor Hartman’s preparations, I had concluded that probably this is what happened in the eggs where he found 12 chromosomes. In other words, that in such cases Hartman’s preparations would show 10 tetrads plus two diads. 36 THEOPHILUS S. PAINTER Later Doctor Hartman was able to find for me one of his clearest cases of twelve chromosomes in the second polar spindle, and placed the slide at my disposal for study. In text figure 8, A I have drawn the chromosomes of the egg illustrated by Hartman in plate 14, figures 3 and 4, of his paper. There can be absolutely no question of the accuracy of Hartman’s count, as there are 12 chromosomes present in this egg. However, on closer examination, it will be seen that 10 of these chromosomes are tetrads, while two of them are bivalent or are diads. (This condition was indicated in Hartman’s figures but the magnifica- tion of the cell was not sufficient to make the fact clear.) Judging = \ | gues eg | rin Le \ of oo ‘ag | s ie \ i @ B A Text fig. 8 Showing the chromosomes in the second polar body spindle of two eggs. A. Twelve chromosomes are present in this spindle, of these ten are tetrads and two are diads. This spindle was figured by Hartman (’19) in figure 3 and 4, of plate 14, of his paper. B. Eleven tetrads are present in this spindle. (One of the tetrads is drawn to one side to show its morphology.) from my experience with the X-chromosome in the second sper- matocytic division, these two diads are halves of what should be the X-chromosome tetrad. As a further check on this conclusion, I selected from Hartman’s slides an egg showing 11 chromosomes. ‘This egg is illustrated in text figure 8, B. A glance at the figure will show that all 11 chromosomes are tetrads, and that no diads are present. It is clear, therefore, that the female opossum has 11 chromo- somes (tetrads) in the reduced number, and it would follow from this that the diploid number is 22, as I have found to be the case in the somatic cells of female embryos. MAMMALIAN SPERMATOGENESIS—OPOSSUM 37 SUMMARY There are 22 chromosomes in the spermatogonia of the opos- sum. The two smallest of these constitute a typical X-Y sex- chromosome complex, which can be followed through the growth period of the first spermatocyte. In the first maturation spindle there are 11 chromosomes—10 tetrads and the X-Y chromosome. The X and Y components segregate from each other in this division so that the secondary spermatocytes contain either 10 autosomes plus X, or 10 autosomes plus Y. There are 11 chro- somes in all secondary spermatocytes, of these one is either the X-component or the Y-component. In either case the sex chro- mosome divides equationally. Half the sperm carry the X- and half carry the Y-chromosome. A study of dividing somatic cells of embryos showed that in both male and female embryos 22 chromosomes were present. In the males the X-Y condition was found, while in the females the 2-X condition existed. It is concluded that Jordan’s failure to find more than 17 chro- mosomes in the spermatogonia was due to faulty fixation. The ‘accessory chromosome’ which he described was not the true sex-chromosome, but probably a displaced tetrad. Hartman’s error in count was due to the precocious splitting of the X-chro- mosome tetrad in the polar spindles. BIBLIOGRAPHY ALLEN, Ezra 1916 Experiments in technique. Anat. Rec., vol. 10. 1918 Spermatogenesis in the albino rat. Jour. Morph., vol. 31. 1919 A technique which preserves the normal cytological conditions in both germinal and interstitial tissue in the testis of the albino rat. Anat. Rec., vol. 16. BAcHHUBER, L. J. 1916 Thespermatogenesis of the rabbit. Biol. Bull., vol. 30. Guyer, M. F. 1909 The spermatogenesis of the domestic chicken. Anat. Anz., Bd. 34. 1910 Accessory chromosomes in man. Biol. Bull., vol. 19. 1916 Studies on the chromosomes of the common fowl as seen in testes and in embryos. Biol. Bull., vol. 31. Hance, Ropert T. 1917 The fixation of mammalian chromosomes. Anat. Rec., vol. 12. 1917 The diploid chromosome complexes of the pig. Jour. Morph., vol. 30. 38 THEOPHILUS S. PAINTER HarTMAN, Cart 1919 Studies in the development of the opossum (Didelphys virginiana). Jour. Morph., vol. 32. Harvey, Erxen Browne 1920 A review of the chromosome numbers in the Metazoa. Part IJ. Jour. Morph., vol. 34. Hitt, J. P. 1918 Some observations on the early development of Didelphys aurita. Q. J. M.5., vol. 63. JorDAN, H. E. 1911 The spermatogenesis of the opossum, with special refer- ence to the accessory chromosome and the chondriosomes. Arch. f. Zellforschung, Bd. 7. MatoneE 1918 On the spermatogenesis of the dog. Trans. Amer. Mic. Soc., vol. 37. Montcomery, T. H. 1912 Human spermatogenesis, spermatocytes, and sper- miogenesis. Journ. Acad. Nat. Sei., vol. 15. NEWMAN AND Parrerson 1910 The development of the nine-banded armadillo. Journ. Morph., vol. 21. ~ Parnter, T. S. 1921 The Y-chromosome in mammals. Science, N. S., vol. 53, pp. 503-504. 1921a Studies in reptilian spermatogenesis. I. The spermatogenesis of lizards. Jour. Exp. Zodél., vol. 34. Stevens, N. M. 1911 Heterochromosomes in the guinea-pig. Biol. Bull., vol. 21. Wireman, H. L. 1917 The chromosomes of the human spermatocyte. Am. Jour. Anat., vol. 21. WINIWARTER AND Satnmont 1909 Nouvelles recherches sur l’ovogénése et l’organogénése de l’ovaire des Mammiféres (chat). Arch. Biol., T. 24. Wintwarter, H. von 1912 Etudes sur le spermatogénése humaine. Arch. Biol., vol. 27. WopDsEDALEK, J. E. 1913 Spermatogenesis of the pig, with special psiose to the accessory chromosomes. Biol. Bull., vol. 25. 1914 Spermatogenesis of the horse, et special reference to the accessory chromosomes and the chromatoid body. Biol. Bull., vol. 27. 1920 Studies on the cells of cattle with special reference to spermato- genesis, oogonia, and sex-determination. Biol. Bull., vol. 38. Yocum, Harry B. 1917 Some phases of spermatogenesis in the mouse. Univ. Cal.-Publ., vol. 14. PLATES EXPLANATION OF PLATES All of the figures represent a magnification of about 3,000 diameters as they are reproduced, except text figures 2, 4, and 7, which were drawn at about 3,000 diameters and reduced by 2. The camera lucida was used for drawing together with a ;'s oil immersion (B. & L.) and a no. 15 ocular. 39 PLATE 1 EXPLANATION OF FIGURES 1 to 6 are equatorial plate views of dividing spermatogonia. Fig. 1, from male no. 1, figs. 3 and 5 from male no. 2, and figs. 2, 4, and 6 from male no. 3. 40 PLATE 1 MAMMALIAN SPERMATOGENESIS—OPOSSUM THEOPHILUS 8S. P AINTER 41 PLATE 2 EXPLANATION OF FIGURES 7 Side view of first maturation spindle, 11 chromosomes present. 8 Same as above, with 11 chromosomes present. The X- and Y-components have separated. 9 Same as above. Two of the tetrads have been displaced. 10 Same as above. Two of the chromosomes which in the spindle under the others have been drawn out to one side in order to show their shapes. 11 Same as above, showing 11 chromosomes. Only the X-component is visible. 12 Same as above, showing with especial clearness the morphology of the X-Y chromosome complex. Not all chromosomes are present in this spindle. 42 MAMMALIAN SPERMATOGENESIS—OPOSSUM PLATE 2 ; THEOPHILUS 8. PAINTER PLATE 3 EXPLANATION OF FIGURES 13 Equatorial plate view of dividing second spermatocyte. Note the quadri- partite form of the X-chromosome. Eleven chromosomes present in this spindle. 14 End view of late telophase stage, showing chromosomes at one end of the spindle. Eleven chromosomes are seen, including the bilobed X-chromosome. 15 Late anaphase of second spermatocytic division showing the Y-chromo- some at each pole of the cell. There are 11 chromosomes clearly visible at the lower pole of the cell. 16 Same as 15, except that the X-chromosome is seen. 17 Equatorial plate view of dividing second spermatocyte, showing the presence of the Y-chromosome. 18 Side view of second spermatocyte spindle, showing the precocious separa- tion of the halves of the X-chromosome. 19 Late anaphase of second maturation spindle, showing the presence of the X-chromosome. 20 Same as above, except that the Y-chromosome is seen dividing. The cells represented in figs. 19 and 20 are probably daughter cells of a single spermato- cyte, as explained in text. 44 PLATE 3 MAMMALIAN SPERMATOGENESIS—OPOSSUM THEOPHILUS S. PAINTER 19 Resumen por el autor, Joseph H. Bodine Los efectos de la luz y la decapitacién sobre la cantidad de CO, producida por ciertos ortopteros. Los datos consignados en el presente trabajo demuestran que cuando se barnizan de negro los ojos de los saltamontes supri- miendo de este modo la iluminaci6n se observan cambios definidos en la tensién o tono de los mtisculos y que estos estan asociados con cambios visibles en la cantidad de anhidrido carbonico pro- ducido por el organismo. Los individuos decapitados también producen menos anhidrido carbdénico. Translation by José F. Nonidez Cornell Medical College, New York AUTHOR'S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, DECEMBER 27 THE EFFECT OF LIGHT AND DECAPITATION ON THE RATE OF CO, OUTPUT OF CERTAIN ORTHOPTERA JOSEPH HALL BODINE Zoological Laboratory, University of Pennsylvania THREE FIGURES As a result of the work of various investigators, particularly of Moleschott,! J. Loeb,? and C. Ewald,? it has been pointed out that the effects of light on the respiratory exchange of organisms are rather variable. In some cases definite increases in the respira- tory exchange have been noted, while in others no direct effects have been detected. More recent investigations, especially those of Holmes,‘ Garrey,> and J. Loeb,® on the tropic responses of organisms to light, have definitely shown that the primary effect of light consists in changes in the tension or tonus of muscles. It is the intention of this paper to show that such changes in muscle tension or tonus in an insect are associated with corre- sponding detectable changes in the rate of carbon-dioxide output of the organism. The choice of suitable material and a method delicate enough to detect slight differences in CO, output are of prime importance in such an investigation. Grasshoppers have been chosen because of the ease in handling, their relatively small size, the fact that they show rather definite tropic responses, and, lastly, that by proper handling their body movements can be practically elimi- nated during experiments. Nymphs or adults of the following species of grasshoppers were used: Chortophaga australior, Chortophaga viridifasciata, Melanoplus differentialis, and Dichro- morpha viridis. All animals were kept in the laboratory and fed grass, lettuce, ete.—the same as those used for other experi- mental work. Carbon-dioxide determinations were made by the indicator method described by Dr. N. H. Jacobs,’ of this 47 48 JOSEPH HALL BODINE laboratory. With this method the CO, output was ascertained by noting the time required by the animal to produce definite amounts of CO.. In blackening the eyes of the animals asphalt varnish was used. This varnish was found to give off no sub- stances which affected the indicator solution. It has been repeatedly pointed out, as the result of experiments to explain the functions of the nervous system of insects, that the brain is the seat of peripheral nerves, the center for inhibiting reflex movements, and for controlling the tonus of the muscles (Bethe*). More recent investigations, and especially those on the functions of the brain of the grasshopper (Ewing!), have definitely shown that the brain controls the tonus of the muscles. It was further shown by this author that neither the supra- oesophageal ganglia or brain of the grasshopper nor the sub- oesophageal ganglion was the center for respiratory movements. Each ganglion of the thoracic and abdominal ventral cord was found to be the center for respiratory movements and reflex actions of the segment and the appendages to which it belonged. It was also pointed out that not only the whole abdomen, but different segments of it continued their respiratory activity when severed from the body. Loeb® has shown that light acting on the eyes of an animal produces definite effects upon the tension or tonus of the muscles. Lyon!® and Garrey,® by blackening both eyes of insects, were also able to show a decrease in the neuromuscular tonus which was normally maintained reflexly by the action of light on the eyes. Both the experiments on decapitation and blackening of the eyes have been repeated on grasshoppers by the author, and results which agree with those of the above-mentioned investi- gators have been obtained. It was thought, however, that-some quantitative measure of the effect of light and of the brain on muscle tension or tonus could be gotten by estimations of the CO, output of the organism. MUSCLE TONUS AND CO. OUTPUT 49 BLACKENING BOTH EYES When one eye or parts of one or both eyes of an animal are blackened, the characteristic changes in posture, etc., are pro- duced; but slight disturbances to the animal due to the partial blackening cause it to try to remove the varnish with its front legs. Such movements are not easily eliminated, and hence no satisfactory results on the CO, determinations are possible. However, when both eyes are completely blackened, the animal makes no attempt to remove the varnish, but remains motion- less. Carbon-dioxide determinations on the animal before blacken- ing its eyes are, with proper handling and manipulation, rather easily and accurately made. In the experiments herein reported at least two or three separate determinations, in which there were no detectable body movements, were made and the average rate of CO, output for the normal animal then obtained. No dif- ference in movements of the animal in the normal and eye-black- ened condition could be detected, so results are not, to any ap- preciable extent at least, modified by CO, produced as the result of body movements. Table 1, in which are listed results of some ten typical experi- ments, shows the time taken to produce the same amount of CO, by an animal when normal and with eyes blackened. In almost every case a marked decrease in the rate of CO, output is noted. Those animals, in which the decrease in rate of CO, out- put was not so marked with eyes blackened, when decapitated also showed a comparatively small decrease in rate. These slight decreases in rate of CO, output in certain animals can doubtless be attributed to the physiological condition of the particular organism. Figure 1 shows graphically the average decrease in rate of CO, output for fifty individuals. To show that this decrease in rate of CO, output was due to the effects of blackening the eyes, and that after removal of the varnish the animal assumed its normal conditions, the following experiment is cited. Individuals were taken, the normal rates of CO, output determined, and then the eyes blackened. The JOSEPH HALL BODINE 50 ¢'9OT , 0ST 0°8 & 91 “AV 40 °9T § SI “AV 40 FI PS “AV 40°6 0S2z 0 FIEPV IOT[BIJSNG “C) 2 61 ¢°9OT (ete Ge, OBES SN; 0°6 GG “AY e-G &8E “AV 4G°E 0cSE 0 YUPV TOT[VI}SNB “C) é GG ey : O02 gg 0 Y Ont ni GG & TT “AV 40'TT 8 “AV OL Oo AY: 5GeG OFGS 0 FINPV TOTBAISNG “C) 2 &@ 0'GI 6°G 0°9 0°06 0°04 GGL & 61 “AV 40°06 G°8T ‘AV 50'8T & @l AV 4G IT OSZT 0 FRPV TOT[BIFSNB "P) eS &% 0ST CLT eal 0 °F Gr OL 9°66 “AV 50 °ZG 8°6I “AV sGGL 84 “AV 499°9 OFIZ 0 FOP V LOTLVIJSNG “C) 2 &G 0°GE bean 0°OL swunub Do pezeqidvoag poueyovyq soAny [euI0N ane | Rear satoaas 2 aes 200 JO ENOOWY ANVS TOAGOUd OF SALANIW NI GWE payppdnoap uayn pup ‘pauayon7q safia ym “Qousou uaym suaddoyssp.ib fo ndjno %*Qy fo ayp4 ay} Buamoys T HIaVL MUSCLE TONUS AND CO. OUTPUT 0°66 “AY 0°96 “AV 6 8] “AV 9 TT AV T°@] “AV So i=) Onn © ydurd yy ydutd yduraé Ny yduté Ny ydurd J SIPRIFUOIOYIP “TW SI]BI}UOLOYIP “WW SIPRIFUOIOYIP “TW BYBIOSVJIPLILA “OC TOTLVAJSNB “C) &% GZ LG ae JOSEPH HALL BODINE rate of CO, output of the animals with blackened eyes dropped characteristically and remained so. After several hours the var- nish on the eyes became dried and brittle and could easily be scraped off with fine forceps, leaving the eyes again exposed. Immediately after removal of the varnish the rate of CO, output increased considerably, being slightly higher than the original normal rate. ‘This increased rate, however, lasted only a rela- tively short time, when the normal rate was resumed. Figure 2 shows graphically the results of such an experiment. N A B Fig. 1 Curve showing the effect of blackening the eyes and of decapitation on the rate of CO, output. Based on average of fifty individuals. Ordinates represent the rate of CO2 output. The normal rate (which is taken as 100 per cent in each case) corresponds to the time to produce a definite amount of CO: (the same amount in any experiment). Points on abscissas indicate different experimental condition of animals, N, normal animals; A, animals with eyes blackened; B, decapitated animals. That the decrease in rate of CO, output was a permanent con- dition as long as the eyes were blackened is shown by the follow- ing. Several individuals of the same species and kept under as near the same conditions as regards food, etc., as possible were taken and their rates of CO, output determined. Half the indi- viduals were then subjected to the eye-blackening process and the characteristic decrease in rate of CO, output obtained. The remaining individuals were kept as controls. All the animals were starved for twenty-four hours under the same laboratory MUSCLE TONUS AND CO, OUTPUT 53 conditions and CO, determinations were then made. It was found that the animals with blackened eyes respired at a much lower rate than the normal individuals. The varnish of the blackened-eyed animals was next removed and a marked increase in rate of CO, output resulted. Figure 3 shows graphically the results of such an experiment. We may summarize the results of the above experiments by stating that a marked decrease in rate of CO, output results from a cutting off of the illumination of the eyes and that this effect is a permanent one so long as the eyes are blackened. 100 a s Yo 75 50 26 N A B Fig. 2 Curve showing the effect of blackening the eyes and of the removal of the varnish from the eyes on the rate of COz output. Based on average of six individuals. Points on abscissas, N, normal animals; A, animals with eyes blackened; B, animals from whose eyes varnish has been removed. Ordinates same as in figure 1. For further description see text. DECAPITATION Since it has been pointed out that the brain of insects exerts a marked effect on the tonus of muscles, but does not control the respiratory movements, it was thought that by decapitation a check on the previous results on blackening the eyes could be made. Decapitated animals were used three-quarters to one hour after the operations, so that any effects of the operation would not interfere with the experiments. The posture, etc., of the animal after decapitation were much the same as when the eyes were 54 JOSEPH HALL BODINE blackened. Both animals whose'eyes had and had not been pre- viously blackened were used, but no differences in response were noted. Table 1 and figure 1 show that a decrease in the rate of CO, output was always associated with decapitation. The degree of this decrease varies slightly for different animals, and it is usually lower than that for individuals with blackened eyes. When decapitated, the amount of actual respiring tissue taken off with the head is rather large and can account for much of the decrease in rate below that for blackened-eyed individuals. The N A B Cc D Fig. 3. Curves showing the effect of blackening the eyes, shortly after the operation, twenty-four hours after, and the removal of the varnish after twenty- four hours on the rate of CO, output. Based on average of six individuals. Solid lines for experimentai animals. Broken lines for normal animals. Points on abscissas, NV’, normal animals at start; A, animals with eyes blackened—at once; B, animals with eyes blackened—after twenty-four hours; C, animals from whose eyes varnish was removed—after twenty-four hours; D, normal animals after twenty-four hours. Ordinates same as in figure 1. For further description see text. operation itself also doubtless exerts some effect. The combined effects of these two factors undoubtedly account for much of this decreased rate, and as a matter of fact it has been found that when the head is put into the respiration tube with the decapitated animal the decrease in the rate of CO, output is strikingly similar to that observed for blackened-eyed individuals. The above results then show that decapitation results in a marked decrease in rate of CO, output not greatly different in magnitude from that observed for blackened-eyed individuals. MUSCLE TONUS AND CO. OUTPUT 55 SUMMARY The action of light on the eyes of an animal (like the grass- hopper), affecting the tonus of muscles, is associated with a decrease in the rate of CO, output of the organism. A similar decrease in rate of CO, output is also found in decapitated indi- viduals. LITERATURE CITED —_ Mo.escHotr 1855 Wien. med. Woch. no. 43 (from Krogh ’16, Respiratory exchange of animals and man. Longmans, Green & Co.). LorB 1888 Pfliiger’s Arch., Bd. 42, S. 393. Ewatp 1892 Journ. of Physiol., vol. 13, p. 847. Hotmes 1916 Studies in animal behavior. Boston. GaRREY 1918 Journ. Gen. Physiol., vol. 1, p. 101. Lors 1918 Forced movements, tropisms and animals conduct. Phila- delphia: J. B. Lippincott Co. Jacoss 1920 American Naturalist, vol. 54, p. 91. BetHe 1897 Pfliiger’s Arch., Bd. 68, S. 494. Ewine 1904 Kansas Univ. Se. Bull., vol. 2, no. 11, p. 305. Lyon 1900 American Jour. Physiol., vol. 3, p. 86. Doe W bo _ oo on THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 35, NO. Ll Resumen por el autor, Dwight E. Minnich La sensibilidad quimica de los tarsos de la mariposa Pyrameis atalanta Linn. En la mariposa Pyrameis atalanta Linn. cada uno de los cuatro tarsos ambulatorios posee quimoreceptores de contacto cuya estimulacién apropiada produce una respuesta en la forma de una extensién de la proboscis. Esta respuesta sin embargo varia algo con la naturaleza quimica del estfimulo y con la con- dicién de la nutricién del animal. Un estudio intenso de estas variaciones ha hecho posible establecer con certeza algunas de las substancias que la mariposa puede distinguir mediante sus tarsos. De los cuatro estimulos probados, tres de ellos, a saber, el agua destilada, I M de sacarosa y 2 M de NaCl son claramente distinguidos por la mariposa. El cuarto, M/10 de cloruro de quinina es claramente distinguido del agua destilada y de la solucién 2M NaCl y probablemente también de la IM de sacarosa. Puesto que los 6rganos tarsales son quimoreceptores de contacto que sirven para distinguir el agua y el alimento normal tal como la sacarosa, pueden propiamente considerarse como eels ceo iscstilse cs 80 INTRODUCTION In a previous paper (Minnich, ’21) I showed that the tarsi of the butterfly, Pyrameis atalanta Linn., are sensitive to contact with certain substances in solution, and hence must possess con- tact chemoreceptors. In the same paper I also presented some preliminary data indicating what kinds of substances the animals were thus enabled to discriminate. It is the purpose of the pres- ent paper to give a further account of the qualitative range of sensitivity of these tarsal organs. In the present work I have employed the same general methods as before, with two exceptions. First, instead of using both captured specimens and those hatched in the laboratory, only butterflies hatched in the laboratory have been employed. The complete adult history of every specimen was thus a known and controlled one. Second, a few animals have been studied, under varying nutritional conditions, over a long period of time, rather than a number of animals, in the same nutritional condition, for a short period of time. Indeed, the data to be presented have been obtained from but eight specimens, and chiefly from four of these. These four butterflies, however, were kept under very close observation for 19, 27, 29, and 30 days, respectively. 57 58 DWIGHT E. MINNICH GENERAL METHODS The experiments were carried out chiefly during the month of July. The laboratory was a basement room of northwest expo- sure, but direct sunlight was kept out by white-cloth shades at the west windows. Ventilation was accomplished entirely from the interior of the buildng, so that it was possible to avoid sudden and pronounced fluctuations of temperature. During the entire period of experimentation the extremes of temperature registered in the laboratory were 21°C. and 25.9°C. For the great majority of days, however, the temperature ranged from 22° to 24°, the variation for the day being less than 1°. Ventilation from the inside of the building also insured a more uniform condition of humidity than outside ventilation would have allowed. Every precaution was thus taken to maintain as great constancy of the general laboratory environment as possible. The specimens employed were all exceptionally large and perfect. After hatching, they were housed in large cages (0.9m. x 0.4m. x 0.4m.) consisting of a light wooden framework covered with mosquito bar. Excepting periods of starvation, which were practiced from time to time during experiments, the animals remained in excellent condition. In one or two instances, near the end of an experiment, one leg of a specimen became more or less stiff and functionless as far as locomotion was concerned. Also the wings, in addition to being clipped slightly for purposes of indentification, became somewhat frayed and considerably rubbed in their outer portions, due to the effects of the holder in repeated trials. But as far as I was able to observe, these slight mutilations in nowise affected the reactions which were being studied. In general, the hairy proximal portions of the wings and the rest of the body remained in almost perfect condition— much more so than is the case with animals in a state of nature. Thus butterflies caught in the field frequently have lost por- tions of the labial palpi, but of the eight animals in my experi- ments not one suffered such a mutilation. Even in their most vigorous states, however, the butterflies flew but little. It was not that they were unable to fly, but that they merely did not. Neither did they creep much, usually CHEMICAL SENSITIVITY OF TARSI OF PYRAMEIS 59 remaining at or near the place where they were released upon completion of a trial. This inactivity was doubtless due in a large measure to the effect of the spring clothes-pin holder on the wings in repeated trials, to continuous confinement, and to the constancy of the environmental conditions, viz., temperature, humidity, and light. As in my previous work, the results of the present experiments have been obtained entirely through a study of the conditions which effect an extension of the proboscis. I have described this response in considerable detail (Minnich, ’21, p. 178), so that it will be sufficient merely to mention the essential features here. In the unstimulated animal, the proboscis remains compactly coiled against the head, but upon appropriate stimulation it is extended and begins to probe the substrate. Not infrequently, however, a given stimulus fails to elicit a complete extension, producing only a partial extension followed by a subsequent recoil. Indeed, the partial extension may be so slight that the compact coil of the proboscis merely exhibits a jerk or two with no further sign of activity. Between such slight reactions and complete extension, all gradations may be observed. It is clear that all extensions of the proboscis, whatever their degree, represent responses. But it is equally clear that these responses differ in intensity. To measure such differences is difficult. Nevertheless, I have endeavored to approximate a measurement by weighting all responses in which the proboscis was uncoiled less than one-half at 0.5, and all in which it was uncoiled one-half or over at 1. In figure 1, no. 1, the proboscis is shown as it normally appears in the unstimulated animal. In the same figure, nos. 2 and 3 illustrate responses which would be weighted at 0.5, while nos. 4 and 5 show responses which would be weighted at 1. As a matter of fact, the case illustrated in no. 4 is virtually never encountered, for the proboscis is rarely extended one-half or more of its length without being completely extended. This scheme for measuring the intensity of the pro- boscis response, therefore, virtually amounts to weighting small partial extensions at 0.5 and complete extensions at 1. I shall employ this scheme in all comparative statements, in order that 60 DWIGHT E. MINNICH evaluations of groups of responses may be compared as well as the mere numbers of responses. Q Ss ee wens ve MZ Fig. 1 Diagrams showing the proboscis in the unextended condition and in various stages of extension: 1, unextended; 2 and 3, partial extensions which would be weighted at 0. 5; 4, partial extension which would be weighted at 1; 5, complete extension, which would be weighted at 1. In the present experiments every trial was immediately pre- ceded by a preliminary trial held under the same conditions. The purpose of the preliminary trial was fourfold: 1) to overcome CHEMICAL SENSITIVITY OF TARSI OF PYRAMEIS 61 the death feint, although this reaction is virtually negligible in Pyrameis; 2) to eliminate the immediate effects of mechanical manipulation incident to the trial; 3) to accustom the animal to the grip of the holder on the wings, and, 4) to make certain that there was no sign of proboscis extension in the absence of the chemical stimulus to be applied. The last-mentioned point is of particular importance, for extensions of the proboscis were occasionally observed when no external chemical stimulus was apparent. Responses of this sort were most frequently observed in individuals which, through prolonged starvation, had become extremely sensitive. Most of these responses were of slight magnitude, although some consisted of complete extensions. The following examples will illustrate the cases in point. In all trials butterflies were manipulated by means of a spring clothes-pin in which the wings were firmly held. The most simple method of placing an animal in the holder was often by direct use of the hands. This involved grasping the wings, and later the thorax and wing bases with the fingers. On grasping the wings with the fingers, as the specimen was pulled off the wall of the cage where it had been resting, there was occasionally a slight jerk of the proboscis coil. Again, while the butterfly was being held by the fingers in one position or another, further incomplete or, in rare instances, complete extensions of the pro- boscis were noted. Because of these occasional responses, direct contact between the hands and the body of the animal was avoided whenever possible. But even when the holder was applied directly to the wings, there were one or two instances of response. ‘The stimuli effecting these responses cannot be postu- lated with certainty. I am inclined to believe that distance and contact chemical stimuli from the hands were chiefly responsible although it is possible that in conditions of extreme sensitivity mechanical stimuli may also exert some influence. In several instances a butterfly which had remained undis- turbed in its cage fifteen minutes or more was found exhibiting either a partial or complete extension of the proboscis. The pre- vious trial having been completed fifteen to twenty minutes before, it was impossible to interpret such a.case as a persisting 62 DWIGHT E. MINNICH response. Internal stimuli due to prolonged inanition, or exter- nal chemical stimuli of an adventitious sort, too dilute for my detection, may perhaps explain these responses, though a certain statement concerning them is not possible. Without multiplying examples, the above are sufficient to illustrate the behavior in question. One of the chief purposes of the preliminary trial was to detect responses of this sort and thus avoid an occasional misinterpretation. If there was but one barely visible jerk of the proboscis coil as the animal was pulled off the wall of the cage, and no further sign of extension during the preliminary trial, experimentation was continued. If, how- ever, there was any more significant movement of the proboscis during the preliminary procedure, experimentation was discon- tinued and not resumed for at least fifteenminutes. Usually there was no evidence of response in the second preliminary, but in ease there was, experimentation was again discontinued for a minimum period of fifteen minutes. The maximum duration of all trials, preliminary and final, unless otherwise stated, was one minute. Failure to observe any visible movement of the proboscis during this period constituted a ‘no response.’ If the proboscis was partially extended early in a trial, the trial was continued, to ascertain whether complete extension would result. If the proboscis was completely ex- tended, the trial was immediately terminated, and the animal returned to its cage. ‘Trials thus lasted one minute or less, depending upon the response. The chemical stimuli employed consisted of distilled water, and three aqueous solutions, viz., 1M saccharose, 2M NaCl, and M/10 quinine hydrochloride. The saccharose and quinine hydro- chloride were USP quality; the sodium chloride, CP quality. The solutes being non-volatile, the distance stimulus afforded by each of the solutions was identical with that afforded by distilled water, viz., water vapor. In other words, the four stimuli used could be distinguished, if distinguished at all, only through direct contact. With the exception of quinine hydrochloride, the stimuli selected were substances frequently encountered by Pyrameis in CHEMICAL SENSITIVITY OF TARSI OF PYRAMEIS 63 its natural environment. The importance of water in this con- nection is too obvious to require comment. As for saccharose, it is one of the chief organic compounds present in the substances on which this species feeds, viz., fruit Juices, exuding sap, and nectar. Thus apple juice is very attractive to these animals, and, if available, may constitute one of their chief foods in the autumn. In orchards where fallen apples litter the ground, I have observed great numbers of the butterflies feeding, just prior to hibernation. According to Browne (’99, pp. 9 and 10) the flesh of the average ripe apple contains 4 per cent of sac- charose, this being, with the exception of water (84 per cent) and invert sugar (8 per cent), the only substance constituting more than 1 per cent of the total composition. Sodium chloride was selected as a common inorganic salt encountered by butterflies in surface waters. It is well known that many species of lepi- doptera congregate about drying pools of water, and pools which have been contaminated with urine or manure seem particularly attractive. In such situations, NaCl is present in considerable quantity. Saccharose and sodium chloride, in addition to the fact that they are frequently encountered by Pyrameis in its natural habi- tat, happen also to be substances which afford adequate stimuli for two of the four primary taste sensations in man, viz., sweet and salt. Quinine hydrochloride was chosen for experimentation, because it affords the adequate stimulus for a third of the human taste sensations, viz., bitter. With Pitter (11, p. 608) I agree that there is not the slightest reason to suppose that a substance which affects the human taste organs in a certain way will affect the taste organs of a lower animal in the same way. It was not to ascertain whether Pyrameis could distinguish bitter that quinine hydrochloride was chosen, but rather to discover whether this salt, which produces such a bitter and disagreeable sensation when applied to the human tongue, would produce or fail to pro- duce a reaction in the butterfly. 64 DWIGHT E. MINNICH EXPERIMENTS WITH LOCAL STIMULATION OF INDIVIDUAL TARSI Three butterflies were kept for five days after hatching with- out access to food or water. At the close of this period, they were subjected to a series of trials in which various chemical stimuli were applied locally to individual tarsi. The stimuli were applied on small cotton swabs ca. 1 em. long and 1 to 2 mm. in diameter, the swab consisting of a bit of absorbent cotton wound on the end of a dissecting needle. Care was exercised to keep the cotton chemically clean while being handled. As a further precaution, swabs used to test the effect of dry cotton alone were heated prior to each experiment, in order to drive off any excess moisture. The butterfly to be tested was placed in a holder (fig. 2) with the four ambulatory feet resting on a small platform of wire screen. In this position, the specimen was closely observed for one minute, this constituting the preliminary trial. In no case was any sign of response observed. Immediately following the preliminary trial, a dry cotton swab was applied to the ven- tral surface of the distal end of the ambulatory tarsus which it was desired to test. Except for a very few partial extensions this also failed to elicit any response. The dry swab was then replaced by a swab saturated with distilled water, and to this the animal almost invariably responded. Upon completion of the trial with distilled water, the specimen was returned to its cage for a minimum interval of fifteen minutes, after which another tarsus was tested in the same manner. This procedure was con- tinued until each of the four ambulatory tarsi had been tested twice. On completing the trials described above, the butterfly was placed on absorbent cotton saturated with distilled water, where- upon the proboscis was immediately extended. In this situation— presumably drinking in water continuously—it was allowed to remain as long as it would. After two or three minutes, how- ever, the proboscis was recoiled, and the animal either crept away or was removed. Some minutes later, it was replaced on the wet cotton, and thus given a second opportunity to drink, but in no case was there any response. ( CHEMICAL SENSITIVITY OF TARSI OF PYRAMEIS 6: After access to water, the butterfly was allowed to rest for a minimum interval of fifteen minutes, whereupon trials were resumed. ‘The specimen was again placed in the holder with its four feet upon the screen platform and carefully observed for one minute. As before, however, there was never any indication of response during this preliminary trial. Next, the dry cotton Fig. 2. Photograph showing holder employed in experiments on stimulation of individual tarsi. Note the appearance of the proboscis, P, in the unstimu- lated animal. swab was applied to the tarsus, but this now failed to elicit even a single partial extension of the proboscis. The dry swab was followed by a swab saturated with distilled water. Prior to the administration of water, this had produced a very high per cent of response. It now failed to produce a response in nearly every trial. An additional stimulus was, therefore, employed, viz., a swab saturated with a 1M saccharose solution. ‘To this stimulus 66 DWIGHT E. MINNICH there was but one failure to respond. On completion of the trial with the sugar solution, the tarsus which had been tested was carefully rinsed by immersing in distilled water, and the butterfly was returned to its cage. Fifteen minutes later another tarsus was tested in the same manner, and so on, until each of the ambulatory tarsi had again been tested twice. The data obtained from the above experiment are presented in table 1. It will be noted that specimens, both before and after access to water, exhibited essentially the same behavior toward the tactile stimuli afforded by the metal of the wire screen and the cotton of the dry swab, viz., failed to respond. It is true that in the case of the starved animals there were three trials in which partial extensions of the proboscis were produced by contact with dry cotton. Possible explanations of these responses will be discussed later. In comparison with the other stimuli employed, however, the response to dry cotton was virtually zero. With the cotton soaked in distilled water, the situation was very dif- ferent. Before the animal was allowed access to water, this form of stimulus was 87.5 per cent efficient in producing a response, whereas after access to water, it was but 4.2 per cent efficient. Yet after this virtual disappearance of the response to water, a 1M saccharose solution was still 91.7 per cent efficient in evoking a response. As noted above, a dry cotton swab applied to the tarsus occa- sionally elicited a slight response. Precisely what was the effec- tive stimulus afforded by the cotton? Two possibilities may be suggested: first, the mere contact (pressure) of the cotton; second, the hygroscopic water present on the cotton fibers. The fact that the few slight responses observed were in starved animals and that these responses disappeared after access to water lends support to the latter suggestion. A final statement, however, as to the effective stimulus in these responses is not possible, at least not from the present data. Returning to a consideration of table 1, the following facts are brought out clearly by the data there presented. First, contact (pressure) stimuli alone, such as those afforded by wire screen or dry cotton, when applied to the tarsi have little or no effect in 67 CHEMICAL SENSITIVITY OF TARSI OF PYRAMEIS ‘yuao rod €°9 = FZ + GL “PZ Weg OABY P[NOA ‘[eII} ATOAD UT UOTSUd}xa oJoTduI0D B Used d10q} pBYy “FY SIM [8}0} oY, “GT 10 ¢° X & ST YSIOM [v}0} AToY} ‘UOTsUa}xo [eTJIed [[eUS B ATWO SyUasoIded YOR ddUTG “poysI] o1¥ Sasuodsa.t da14} “E uUINjod UL ‘sNYyy, ‘[el4} ALaAe UT stosoqoid oy} Jo WOTSUa}x9 9Ja[dUIOD B UdEq 919} PBY POUTE}GO U90q SAVY P[NOA TOIYA FYSIOM [v}0} oy} Aq ‘GOT 03vd uo pouyop sev ‘pouteyqo sasuodsad Jo 4YSIOM [e404 oY} SUIPIATp Aq po}E[NoTVO SI ssousATOayo Jo yuad rod OT, 1 1°16 GP 0 0 ¢°18 Sa9 0 "** *SSOU9ATPIIZO JO U9 19g 6G i 0 0 IG cia 0 ‘“**-gsuodsar poyysiom [290], 1G |G Le NEG Cor a0. a, Ome re Os OS epee Obie slab a Oe ae EGO} | gO! ae 1G eee see aie tage ge ee STeIOL See O ee Ome Or ser He AO. Se On acer FOF si Om ese srk, oe Oe a0 I ANA [LO te VE iy I Deeb tte ZnO Ome er iOlma Ore se. ioe iG EO PD I ss && L Se OR Oe Oar Se OT Oma ae 30m Oke (as Fez Bohs Om sO) eile 10" 18: cE Te5 lore ee | Sole sl ee altel ae. ice sal: Slanal eh vel mtcile on] Sener one efle2| F laf) s2| 2 leSlet| Flaglet| o laglst| o lasle2l fle] s2| ¢ or Coal oS. 2B g chs A a g eee Stet 3 35 e = e ef B = 3 33 5 g 3 ao. Bg. Z oY | eye fe sy 54 Sal) rele! ullieo tame toe ty |) tereee e sh |] pen || ey ta} he) ell CES | ksh Eee eye ll ke) SE\Bo| 8 le®lBo| 8 |SelBo| £ |eelBe| S |S7|Ba| 8 |S8|Bol 8 |S Eo! 8 e Maee| oiescam | WG) alee mle alee A tS Mlsgo| 5 a 8 snsie} 9u0 snsie} suo snsiv} ou0 1818} snsie} 9u0 snsie UgadWON IVWINY I 4 au0 1s18} ales dena | re eee UT | qqr yowquoo ur | yyt joequ00 ut | T* ccs aT | yar youquoo ur | qjt youQUOo ur (1) asoreyooes J T | 107M poTystq SL estat SEASON: 10}BM por[iysiq Seana Ata eas ON (8) (1) (9) (¢) (F) (8) (2) ACLVM ALTIILSIa UGLVM YON dood AAHLIAN JO NOILVUISININGY UddaV ‘y NI SV SaITANaLLAg aAWVS QGAITOAA GVH HOIHM GIO SAVa G SAITINALLIOG a Vv 18.0} JONparpUr UO WNUWI}s Qualaffip fo sjoaffa ay? Bunoys T ATAaVL 68 DWIGHT E. MINNICH producing an extension of the proboscis. Second, the same stim- uli plus certain chemical stimuli will effect an extension of the proboscis. Pyrameis must, therefore, discriminate through its tarsi the presence of the chemical stimuli. Third, under certain conditions, distilled water is one chemical stimulus which is very efficient in evoking a response. Fourth, one condition which determines the responsiveness to water is the degree of inanition with respect to water. Satiety with respect to water inhibits the responsiveness thereto. This response may, therefore, be con- trolled, approaching 100 per cent or 0 per cent, according as the animal is or is not allowed access to water. Finally, although Pyrameis fails to respond to water in a condition of satiety with respect to the same, it, nevertheless, continues to respond vigor- ously to a 1M saccharose solution. The butterfly must, there- fore, be able through its tarsi to distinguish sharply between water and an aqueous solution of a non-volatile substance such as saccharose. EXPERIMENTS WITH SIMULTANEOUS STIMULATION OF ALL AMBULATORY TARSI Because of the mechanical difficulties involved in stimulating individual tarsi, a more satisfactory method of experimentation is to allow all the ambulatory tarsi to come in contact with the stimulus at the same time. Experiments using this method were, therefore, more generally employed. In these experiments the wings of the butterfly were held in a spring clothes-pin, which was manipulated by the hand. Upon removal from the cage, the butterfly was first subjected to a preliminary trial of thirty seconds, in which the feet rested on clean filter-paper. This trial was carried out within 3 to 5 cm. of the place where the sub- sequent trial with a given chemical stimulus was to be made, so that the environment in the two trials was practically identical. During the trial, the butterfly was gently lifted and let down again at intervals of ten seconds, in order that it might become thoroughly accustomed to this sort of manipulation. If there was no significant response during the preliminary trial, as was usually the case, the butterfly was lifted from the CHEMICAL SENSITIVITY OF TARSI OF PYRAMEIS 69 filter-paper, across the few intervening centimeters, and set down with the feet in contact with a thin layer of cotton, contained in a Syracuse watch-glass and saturated with the solution to be tested. If the animal failed to respond promptly, various slight manipulations were employed to make certain that the ventral surface of every tarsus was afforded ample contact with the stimulus. Thus, if there was no response during the first few seconds of the trial, a slight, even pressure was given the holder, thereby causing the distal portions of one or more tarsi to press more firmly against the cotton and thus become immersed in the solution. If still there was no evidence of response after twenty to thirty seconds, the holder was turned slightly, forcing the butterfly to shift the position of some of its legs. And if both the above measures proved ineffective, toward the end of the trial the animal was occasionally lifted gently and let down again. At the close of each trial the butterfly was placed over a watch- glass containing distilled water, and the tarsi were thoroughly rinsed. This was necessary in order to prevent contamination of the stimulating substance in one trial by adhering material from previous trials. If the proboscis remained extended after the trial, care was exercised to prevent the animals from drinking at this time. After the legs had been well rinsed, the butterfly was placed on clean filter-paper for a moment to absorb the excess moisture, and was then returned to its cage. In the above manner, butterflies were tested with distilled water and solutions of 1M saccharose, 2M sodium chloride, and M/10 quinine hydrochloride, four trials being made daily with each of these substances. The order in which the four stimuli were employed was varied from time to time and a minimum rest period of fifteen minutes was allowed between consecutive trials. The responses of each butterfly were studied under three nutri- tional conditions: first, a condition in which the animal was receiving neither food nor water; second, a condition in which it was receiving water only, and, third, a condition in which it was receiving both food and water. The general plan was the following. Immediately upon hatching, the specimen was placed 70 DWIGHT E. MINNICH in a cage where it was kept without food or water. After three days, trials were begun with all four substances and were con- tinued until the butterfly gave 100 per cent response to distilled water; that is, complete extension of the proboscis in each of the four trials of the day. In a few cases the experiment was con- tinued for a day or so beyond this point before making any change in the nutritional conditions. In general, however, this was not possible because of the growing weakness of the specimen and the danger of its death. ’ When the experiment had reached the stage described, at the conclusion of the trials for the day the butterfly was placed on absorbent cotton saturated with distilled water and allowed to drink all it would. On the following morning it usually appeared quite revived. Before resuming trials, however, the animal was again given an opportunity to drink. If it failed to respond to the water, I several times forced an extension of the proboscis by touching one of the tarsi with sugar solution. Thus while the animal could not be compelled to drink, it could be compelled to bring the proboscis in contact with water. In this manner it was offered water in the morning, one hour before trials were begun, and in the evening, immediately after trials were concluded. ; The butterfly was continued on this water diet until it again became so weak that further trials were impossible, whereupon experimentation was discontinued for the remainder of the day, and 1M saccharose was administered. Of this solution the animal always imbibed freely, and it was not until the abdomen was greatly distended that it ceased to feed. On the following morning it would again appear quite restored, and trials were re- sumed. The saccharose diet was continued for 3 to 4 days, administrations being made twice daily as in the case of water. Following the period of saccharose diet, the butterfly was again kept without water or food until the response to water rose to 100 per cent. Then followed a period of water diet, and when this became insufficient, the sugar solution was again adminis- tered. In other words, the three nutritional states described above were repeated. CHEMICAL SENSITIVITY OF TARSI OF PYRAMEIS 71 In continuing periods of starvation as long as possible, the risk of death became very great, and of eight specimens employed, six died in the course of the experiment. The longevity of these individuals was 8,8, 9, 10,19, and27 days, respectively. Inall these cases death was most probably due either directly or indirectly to the effects of starvation. In several of the shorter-lived speci- mens, I might have saved them had I appreciated the gravity of their condition soon enough, for animals in a state of complete collapse may sometimes be resuscitated in a surprisingly short time by the administration of 1M saccharose. ‘Two specimens, however, survived the entire experiment, and several days after their last trials appeared vigorous in every way. They were killed for subsequent morphological study, having been under observation for twenty-nine and thirty days, respectively. Of the eight animals, [ shall present the data from the four longest- lived only. The data from the four shorter-lived individuals are less complete and show nothing not shown by the others. The observations on the four butterflies which survived longest are presented graphically in figure 3. The data for each animal numbered, respectively, 11, 12, 13, and 22, are presented in the form of four curves, each of which represents the responsiveness to a single substance. In these curves, the total weight of daily response, as defined on page 59, is plotted against age in days, the nutritional state for each day beingindicated. An examina- tion of the four curves for any one animal shows, with one excep- tion, viz., no. 22, that no two coincide. In other words, there were differences of response to the different stimuli. It should be borne in mind at the outset of this discussion that the fact that a butterfly responds identically to two substances does not necessarily mean that it fails to distinguish them. This may merely indicate a positive response to both substances. On the other hand, differences of response do not necessarily show dis- crimination, unless they are differences of pronounced nature and regular occurrence. Are the differences noted above such as to indicate discrimination of the various stimuli or not? In order to answer this question, let us examine and compare the responses with respect to: first, relation to nutritional condition; second, THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 35, NO. 1 "2 DWIGHT E. MINNICH intensity, as indicated by the total amount of response pro- duced, and, third, time aspects, viz., latent period and period of execution. Let us take first the response to distilled water. This response was characterized by its close relationship to the nutritional condition of the animal. As shown above, in the experiments on stimulation of individual tarsi, table 1, the responsiveness to water could be increased or decreased by preventing or allowing access to water. The same results were obtained when all four tarsi were stimulated simultaneously. The curves of response to water in figure 3 show in every instance that with continued starvation the response finally rose to 100 per cent, while directly after water was administered it dropped to 0 per cent, where it generally remained as long as water or an aqueous solution was accessible. In other words, the response to water depended directly on the nutritional state of the animal. In this respect it was absolutely unique, for in no other response was any intel- ligible relationship to nutritional condition evident. The re- sponse to distilled water is thus sharply differentiated at the out- set from the other responses studied. Consequently, we may omit it from further consideration, confining our attention solely to a comparison of the responses to the three solutions. A survey of the curves of response to 1M saccharose (fig. 3) shows that each of the four butterflies responded to this solution in every trial, irrespective of nutritional condition. My notes show that during periods of inanition the response often became more persistent, lasting for some minutes after the trial had been completed. But even during periods when 1M saccharose was being fed twice a day, there was never a single failure to respond. Thus, a butterfly which had ceased to feed and had crept away, if replaced on the cotton soaked with the solution, immediately responded anew. And, if after ceasing to feed, the specimen remained on the cotton without creeping away, mere seizure of the wings or other slight agitation was usually sufficient to induce a fresh response. The response to 1M saccharose was thus entirely independent of the varying nutritional conditions of the animal. CHEMICAL SENSITIVITY OF TARSI OF PYRAMEIS 73 Not only did the specimens always respond to 1M saccharose, but the extension of the proboscis was in every case a complete one. The total weighted response of all four animals under all conditions was, therefore, 100 per cent. Furthermore, the responses were rapidly executed. As shown in table 2, the average time required for this solution to produce a complete extension of the proboscis was between 1.7 and 3.4 seconds. The response to 1M saccharose was thus characterized by absolute constancy under all nutritional conditions, by maximum inten- sity, and by rapidity of execution. These facts all indicate that this stimulus must be a very powerful one. Indeed, it was by far the most effective of the stimuli tested. TABLE 2 Showing the average time in seconds required by different stimuli to effect a complete extension of the proboscis M/10 QUININE £ é s D 1 M SACCHAROSE 2 M SODIUM CHLORIDE ISTILLED WATER HYDROCHLORIDE Number | Average | Number | Average | Number | Average Number | Average of trials time of trials time of trials time of trials time il 80 3.4 35 $224 19 12.5 76 25.4 12 60 3.2 37 5.4 13 16.7 29 43.7 13 90 La 30 4.2 16 4.5 86 10.8 22 68 1.8 52 ee 10 4.3 68 6.1 Unlike 1M saccharose, the responsiveness to 2M NaCl was characterized by extreme variability. One has only to examine the curves in figure 3 to become convinced of this fact. Take, for example, specimen 11. During the first three days of its life, no trials were made. On the 4th, 5th, and 6th days, the responses to this solution were 0. On the 7th, 8th, and 9th days, there was a variable number of responses for each day. On the 10th day the number of responses again fell to 0. On the 11th day it rose to 100 per cent, where it remained until the 16th day. The 16th, 17th, and 19th days (no observations were made on the 18th) the responsiveness was a little less than 100 per cent. On the 20th day it dropped to 0, only to rise to 100 per cent the 21st day, and again fall to 0 the following day, where it remained for the last two days of experimentation. The responses of this 74. DWIGHT E. MINNICH specimen were thus extremely variable, and specimens 12, 13, and 22 showed essentially similar conditions. A careful study of each of these cases fails to show any apparent relationship between the nutritional condition of the animal and its response to the stimulus under consideration. Doubtless this response is very definitely determined, but the determining conditions are not evident from the present data. For the present, therefore, the outstanding characteristic of the response to 2M NaCl is its variability. It may be added that .this variability is very great not only from one specimen to another, but also from time to time in the same specimen. The variability of response to 2M NaCl means, of course, that the total amount of response to this substance was much less than to 1M saccharose. Considering the weighted responses of all four specimens collectively, the sodium chloride produced but 51.6 per cent response, as compared with 100 per cent for the sugar solution. On comparing the curves of the two responses, we find some days when specimens responded indistinguishably to both stimuli. But we also find days when there was a clear- cut difference of response. With every specimen there were periods, ranging from two to seven days, during which the animal gave 100 per cent response to 1M saccharose solution, and yet failed to evince even the slightest indication of response to 2M NaCl. The only plausible explanation of these facts is that the animal discriminated clearly between the stimuli. The fact that it responded at times to both is in nowise incompatible with this interpretation, while the fact that at other times it responded 100 per cent to one and 0 per cent to the other can hardly be explained in any other way. Clearly, therefore, the tarsi enable Pyrameis to distinguish a 1M saccharose solution from a 2M sodium-chloride solution. There remains for consideration the response to quinine hydro- chloride. In certain specimens, for example, no. 12 and to a slight extent no. 11, the curve of response to this substance (fig. 3) also shows some variation from day to day, though much less than with sodium chloride. No. 22, however, responded with a complete extension of the proboscis in every trial, and no. 13 Discarded i ga i i al 5 2 § Ta * still vigorous “ee of re: areani er ae Se ee ee ae a FEO ~Ofed nor 1M sacch. Ne aii Soe SCP COON A fete 7 VS 1 ae 4 ; Citas Ral Ain. ..BB) 2a) ety BEN Ebates 2a Ube = Saree anoiaemer! Given faa FeO Offered PG Offered Given ‘= Pr fered 5; 20 nor 1M sacch. JM sacch. nor 1M sacch 25 26 27 28 ra] 30 3 4 ay ar) Offered HeO Weight of responses 15 Weight of responses o i 2 3 4 5 6 Z 8 9 70 fins 8 is 5 6 7718 19 20 _y»_2l (ama ae |, Sa” ay | TEE | a. A ge ip RS Given neither HzO Offered HzO “OWered 1M sacch. Given neither HeO Offered HeO Offered nor 1M sacch. nor IM sacch. ig JM sacch, * Co © of OI a, Se eS 4 Specimen PSH vigorous Discarded aA Allowed fo recuperale Weight of responses fe Age in days ° pone, 3 oe 8 6 td 8 9 7o* (2 [4 5 16 (a ee a a a ace ae Ts ae a eT a wee oe fe Offered HzO “Gifered 7 pa Given neither HeO nor IM sacch. Offered HeO “Offered 1M sacch. ec Fig. 3 Curves of response to various solutions and distilled water. ---------- 1M saccharose;........ M/10 quinine hydrochloride; --------- 2M sodium chloride; ———— distilled water. Except where indicated by asterisk, the ordinate length represents the total weighted response, as defined on page 59, obtained from four trials made in the course of a single day. The asterisks indicate cases where for some reason, usually the weakness of the animal, it was impossible to complete the four trials of the day. In these instances the ordinate length represents the per cent of response based on the number of trials actually made. 75 = 76 i's vs CHEMICAL SENSITIVITY OF TARSI OF PYRAMEIS 79 did the same excepting the last day of the experiment, when its responses suddenly dropped to 0. The total weighted response of all four specimens to the quinine solution was 84.7 per cent, as opposed to 100 per cent for 1M saccharose and 51.6 per cent for 2M NaCl. The amount of response produced by quinine was, therefore, intermediate between the other two substances. A comparison of the curves for the quinine solution with those for sodium chloride (fig. 3) shows in every animal that there was a great diversity in the two responses. At times, the amount of response to each stimulus was the same; at other times, it was totally different. Here again, therefore, we must conclude that the two stimuli were differentiated. A comparison of the saccharose and quinine curves yields no such conclusive evidence as the case above. For, while the two curves exhibit a rather wide divergence in animal no. 12, they very closely approximate one another in the other three animals. In this instance, however, there were distinctive differences of another sort. The rapidity with which the proboscis was ex- tended to 1M saccharose has already been pointed out. As a comparison of the figures in table 2 will show, the response to quinine was very much slower. Thus, in animal no. 11 the average time required by quinine was seven to eight times that required by saccharose; in no. 12, thirteen to fourteen times; in no. 13, six to seven times, and in no. 22, three to four times. With 1M saccharose the extension of the proboscis began very shortly after the application of the stimulus and was rapidly completed. This was not the case with quinine. As a rule, the application of this stimulus was followed by a latent period lasting from a few seconds up to as many as sixty seconds, during which there was no sign of response. Then followed a period of reaction, beginning with slight relaxations of the proboscis which more or less gradually increased until ex- tension was complete. The period of reaction also lasted from a few seconds up to fifteen or twenty seconds or even longer.! 1 The long latent period together with the long period of extension necessitated the prolongation of a number of trials with the quinine solution from the usual duration of one minute to two minutes. 80 DWIGHT E. MINNICH The average response to M/10 quinine hydrochloride thus dif- fered strikingly from that to 1M saccharose. That this differ- ence, together with the slight differences of distribution noted, indicates a discrimination of the two stimuli, seems to me not only possible, but very probable. A final statement, however, cannot be made with certainty. From the evidence presented, it is quite clear that the tarsal organs of Pyrameis are chemoreceptors of a rather wide range of sensitivity. Through them the butterfly is able to differentiate such solutions as 1M saccharose and 2M NaCl from distilled water and from one another. It is also able to differentiate M/10 quinine hydrochloride from distilled water, from 2M NaCl, and probably from 1M saccharose. The appropriate stimulation of these organs leads to an extension of the proboscis, the initial act in food taking. ‘The tarsal organs are thus organs of chemical sense, concerned in the discrimination of food substances, and may be properly considered as organs of taste. GENERAL SUMMARY AND CONCLUSIONS 1. In Pyrameis atalanta Linn. each of the four ambulatory tarsi possesses contact chemoreceptors. 2. The appropriate stimulation of these receptors produces a response in the form of an extension of the proboscis. 3. The manifestation of this response varies somewhat, depend- ing upon the chemical nature of the stimulus and the nutritional condition of the animal. 4. An intensive study of these differences of response shows that Pyrameis is able to distinguish the following substances from one another through its tarsal organs: distilled water, 1M saccharose, and 2M NaCl. It is also able to distinguish M/10 quinine hydrochloride from distilled water, from 2M NaCl, and probably from 1M saccharose. 5. The efficiency of distilled water in evoking the proboscis response is directly dependent upon the nutritional condition of the animal. CHEMICAL SENSITIVITY OF TARSI OF PYRAMEIS 81 6. The responsiveness to 1M saccharose, 2M NaCl, and M/10 quinine hydrochloride shows no apparent relationship to the nutritional condition of the animal. 7. According to the scheme of measurement employed in the present paper, the total response of all animals together was 100 per cent to 1M saccharose, 84.7 per cent to M/10 quinine hydro- chloride, and 51.6 per cent to 2M NaCl. 8. Since the organs of the tarsi are contact chemoreceptors, and since they are concerned with the discrimination of substances to be taken as food, they may be appropriately termed organs of taste. BIBLIOGRAPHY Browne, C. A., Jr. 1899 A chemical study of the apple and its products. Bull. Penn. Dept. Agriculture, no. 58, 46 pp. Minnicu, D. E. 1921 An experimental study of the tarsal chemoreceptors of two nymphalid butterflies. Jour. Exp. Zoél., vol. 33, pp. 173-203. Pirrer, A. 1911 Vergleichende Physiologie. Jena. ve ‘ = wae A ve: ciel AMR diy, | f ee elas 8 wre me geen os pr ty aeeaaey) “aaa et | ee Herero: rondindaa ning ots iyo: hocdaieain “ ar Pa nseceleitins Wi sepia abiaeaty oct! pine a ‘ ee id uasayaey hotnateystuaqorgeyy ac. yale LT biota ‘ ia I Dita ad \c Ma. ee “ ee aye vale rt Sep: ibe , ; hd sé F * payed Rev A ‘' OlaedABO! eee aul ea YM, ie r Pane See ee sulk axel Brus aeacheray elites Me: ia ne Tol bbn by. in Ca A, MEG ™ h' Y ier ) fy a + kad pares ory 1e sea, , aft fis # Cae ‘ Pry bs ' ‘A \ ned ers vleneail5 loasl Yl th rete fey aca eea * ae ROSS ST vite: CODD Cac aL. tHa's) HA) hitodeshiat Deri a i rab Meee ay lpnay antl? ARID ieee big ota lene Mil De ee 4 haw alk: wee he Al Ly J ‘ iu Stee) 1o0)') Tiel, TR ae... - é - ri il SAT al * é mata a! i * [ ae a Aa ] He A? ee > oe AUTHOR’S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, JANUARY 16 THE TEMPERATURE SENSES IN THE FROG’S SKIN ANN HAVEN MORGAN Mount Holyoke College, South Hadley, Massachusetts ONE FIGURE CONTENTS CUAL DOSE TIC pees AS 7 eR a oa SON ne Pe aE CO Sh PE gS HL A 83 eterna ea ta eae he ee erate a ee REL sl statecBidals o beveee weg tieeeiee dem 84 Mert | Oa eee on a A eins oct eee Rae) a eee RRC oe TC Emo e 88 I ERAT GSTS MPa Ne ae 35 Ree, 808 Cech yn: Soles goo eee ss ws rae 8 89 MP HICSDONSCSALOLMCA LEN eee anne teen tet chr eee oie, HOtiels snculeitee 89 a. Independence of receptors for touch and heat.................. 96 b. Independence of receptors for pain and heat.................... 100 c. Independence of receptors for acid and heat.................... 102 Pe OP SSCHI TO) COMM eres nee sa rc uct caw otuied- oieeine meas ae 103 a. Independence of receptors for cold and heat.............--..... 107 b. Independence of receptors for touch and cold................... 107 c. Independence of receptors for acid and cold.................... 108 d. Independence of receptors for pain and cold.................... 108 BIR CUS SIOTIM Etna erste Nee oir cists cies cP ee oie, ey eae ene canoe eyes nON Seheusie ONS eee 111 SOCLED TINTS YS as fotena tcp MMM ges LS" Ol A A SARI di 2 a ge to NO 7 SP 112 eee SETS re ee te Sh hah Nuie, vache enn Zh aera tatalaial Se ae pads aaa 113 INTRODUCTION Is a temperature sense present in the frog’s skin? Canit be isolated from the chemical and tactile senses which have already been shown to be there?’ Is it separable into the elementary senses of heat and cold.as in the human skin? If so, what is the limitation and the nature of the responses to heat and cold? It is in common knowledge that frogs go down into the mud in the winter and come up in the spring and an expectation of March and April that they will be heard on the warm nights. Their thermic susceptibility has long been known, both in nature and in the laboratory. Brown-Séquard (’47) alluded to the possible effect of temperature upon his reflex frogs which lived longest during the months from June to September, and Kunde 83 84. ANN HAVEN MORGAN (60) recorded that reflex frogs dosed with strychnia were seized with spasms in a warm room, but became quiet when placed upon ice. A series of investigations followed these early suggestions, actual experiments on the direct and indirect effects of tem- perature upon the central nervous system (Tarchanow, 771; Archangelsky, ’73; Freusberg, ’75; Wundt, ’76), the sensory nerve endings (Heinzmann, ’72; Foster, ’73; Rosenthal, ’75; Sedgwick, ’82), and the general behavior of the frog under stimulation by heat and cold. Recent workers upon its responses to light, electricity, sound, chemicals, and temperature have either shown something of its thermic sensitiveness (Kordnyi, ’92; Parker, 03; Torelle, ’03; Yerkes, ’06; Pearse, ’10) or have suggested the presence of a temperature sense in the skin. But its exclusive presence there, its existence as a separate sense, and the nature of its responses have not been adequately shown. It was with the hope of doing this and of answering the questions already suggested that the present study was undertaken. The problem was suggested to me by Prof. G. H. Parker, and it gives me pleasure to express my appreciation of his friendly criticism and constant help. HISTORICAL Temperature studies upon the frog have covered a wide range of attack. Information regarding the effect of temperature upon the skin has come into the literature indirectly, usually in connec- tion with special studies of a system of organs or the behavior of the whole organism. In the hope of showing these different aspects with greater clearness, I have discussed them by topics rather than in historical sequence. Frogs respond to variations in temperature by visible motor reac- tions. ‘This was established experimentally by Kunde (’60), Cayrade (’64), Goltz (69), Tarchanow (’71, ’72), Archangelsky (73), Rosenthal (’75), Freusberg (75), and Wundt (’76). The means of stimulation were partial or complete immersion in warm or cold water, dipping in warm or cool dilute acid, and ice packs and hot sand baths. From treatment of these kinds one group of workers (Kunde, ’60; Richardson, 67; Weir-Mitchell, ’67; TEMPERATURE SENSES IN FROG’S SKIN 85 Rosenthal, ’75) maintained that cold caused a depression in reflex excitability, except in the case of ice packs, which in- creased it (Richardson, ’67; Weir-Mitchell, 67; Wundt, ’67), and another group were of the opinion that heat properly ap- plied also caused excess excitability (Cayrade, ’64; Goltz, ’69; Tarchanow, 771, ’72; Freusberg, ’75). When cold or heat are applied very gradually to a frog the reactions decrease in extent and tigor. ‘There has been a good deal of dis- agreement in the literature, regarding reactions to gradually applied stimuli. The question was opened by Goltz (’69), who immersed normal and reflex frogs in water of gradually increasing temperatures. When slowly stimulated up to 30°C., the normal frog became violent, but the reflex frog remained inert. Goltz’s main purpose had been to show the difference between the two conditions in the animal, and he immediately declared the lassi- tude of the reflex frog due to its brainless state. By the same method, Tarchanow (’71, ’72) secured similar results on normal frogs. The next year Heinzmann (’72) continued similar ex- periments from the point of view that the sensory nerves might be affected by a stimulus increasing in intensity so slowly that destruction of the nerve would result before a reaction could occur. Normal and reflex frogs were heated with the expected results to both of them, and these were explained as due to the very gradual succession of the stimuli. In 1875 Fratscher re- peated these experiments with identical results. The quiet normal frogs of Heinzmann and Fratsecher were thus pitted against the violent normal frogs of Goltz and Tarchanow, but the main conclusion seems to have been that no reaction would result if stimulation were applied with sufficient gradualness. Foster (’73) had previously questioned Goltz’s -tatement that brainless frogs would give no reaction to stimuli to which normal frogs reacted so vigorously. He immersed reflex frogs ‘locally and totally’ and obtained very different results in the two cases. When large areas of the body were immersed there was no re- sponse, but when only the toes were dipped, no matter how gradually the heat was increased, they were always withdrawn at about 35°C. This peculiar result was explained by Foster 86 ANN HAVEN MORGAN on the ground that immersion of the larger areas heated the blood, which in turn warmed the spinal cord and reduced its irritability. With the stimulation of the small area no such general warming could take place, and hence the normal irritability of the cord was retained and the vigorous response followed. Certain puzzling phases immediately presented themselves, and Sedgwick (’82) repeated the experiments upon which this explanation was based. He suspended the reflex frogs in the manner described by Foster and at once discovered that in this upright position the heart was practically empty and could not possibly circulate the blood as stated by Foster. From this tangle of statements the best evidence seems to show that the reflex frog will respond to heat at certain degrees, no matter how gradually it is applied, but that the extent and vigor of these responses may be reduced by the graded application of the stimulus. Effect of heating and cooling the spinal cord. With the object of stimulating the spinal cord, Archangelsky (’73) suspended reflex frogs with their trunks surrounded by a jacket of hot air which produced a rise of excitability, and by a jacket of slowly heated air which produced no change. Tarchanow (’71, ’72) stimulated the cord directly with an ice pack, thereby causing a depression of reflexes. Frogs can withstand a temperature as low as 6°C. The body temperature of frogs was recorded by Milne-Edwards (’68) and by Rogers and Lewis in 1916. Knauthe (91) and Miiller- Erzback (’91) froze frogs in water and exposed them to tempera- tures of —4°C. to —6°C. for several hours. Maurel et Lagriffe (00) studied the effect of temperatures from —4°C. to 41°C. and maintained that a frog may survive a temperature of 0°C. or even —3°C. Respiration is quickened under stimulation by heat. When Babak (’13) warmed the skin of a reflex frog, the speed of respira- tory movements was quickened, and when he cooled it, corre- spondingly the speed was decreased. Frogs which are immersed in cold water will swim downward and will remain at the bottom a greater percentage of the time as the cold TEMPERATURE SENSES IN FROG’S SKIN 87 is increased. Frogs which Torelle (’03) placed in water of 10°C. immediately swam down and remained below, usually with legs stifly outstretched. In 1918 Brooks corroborated this by a series of detailed observations on frogs which were placed in water of decreasing temperature. As the water was cooled the frogs remained for a shorter and shorter time at the surface till at 5°C. they settled to the bottom and remained there. The skin is sensitive to variations in the temperature of air, of water, and of acid solutions. Comparing the sensibility of the skin and afferent nerves by treatment with warm and cool acid solutions, Tarchanow (’72) was the first to point out that the thermal end-organs must be in the skin and that the quicker response to the warmer acid solutions was due to an increased irritability in the nerve endings, agreeing in this with Archangel- sky (’73) who had used the same stimulus. This sensitiveness of the skin has been mentioned or investigated by recent workers in connection with studies of other sense organs, and Kordnyi (93) and Pearse (710) found the frog’s integument responsive to both light and heat. Pearse secured responses from frogs whose feet were dipped in water at 40°C. and 45°C. and Reese (06) obtained similar results from Cryptobranchus, while Parker (03) and Yerkes (’06) both alluded to the susceptibility of the skin to changes of temperature. Warmth produces a positive and cold a negative response to light. When frogs were placed in warm air or water they moved toward the light, but in the same media at 8°C. they moved away from the light (Torelle, ’03). L. J. Cole (’07) secured similar results when he placed a frog in a dark box between a large and a small illuminated area at the opposite ends. When the frog was cooled to 6°C. and 10°C. it would move toward the smaller area, but when warmed it would immediately move toward the larger one. In order to compare the relative effects of light and heat, Pearse (10) arranged a series of tubes, with a measured heat radiation upon the sides of a totally dark box. Another box contained a light whose heat output was one-half that of the pipes. LEyeless toads placed in these boxes proved to be almost totally indifferent to the heat, but were strongly phototropic, showing that light 88 ANN HAVEN MORGAN and heat were unlike in effect and that the photoreceptors were much more easily excited than the receptors for heat. A slight difference in light, on the other hand, made no impression on frogs with which Torelle (03) worked. They swam up and down in the jars regardless of adjustments of light and dark. Frogs are stereotropic in temperatures between 10°C. and 4°C. When Torelle placed frogs in water cooled to 10°C. or below, they flattened their bodies against the bottom or crept under rocks placed on the floor of the aquarium. Effect of temperature on responses to electricity. An electric current which produced tetanic movements on a warm frog showed retardation when the frog was cooled (Kunde, ’60). METHODS The experiments which follow were performed upon green frogs (Rana clamitans) and leopard frogs (Rana pipiens) in a laboratory the temperature of which varied between 18°C. and 23°C. The work was done between October and January upon animals which were kept in a basement tank and brought into the laboratory at least two days before they were used for experimentation. For all except one experiment, the front part of the head was removed by a single transverse cut made just in front of the eardrums. Through the lower jaw thus left intact a loop of silk was drawn, and by this the frog was suspended, thus avoiding the irritation caused by the repeated use of a metal hook. Frogs were hung from an extension bar, attached to a standard; the bar could be easily raised and lowered. They were com- pletely immersed in a bath of water at the beginning of each experiment, and at certain intervals during treatment in order to keep the temperature normal, the skin moist, and free from particles of dust. At the beginning of an experiment the tem- perature of the room, bath water and frog were taken, the latter being secured by putting a thermometer through the mouth and down into the stomach. Records of these temperatures have been given with each experiment recorded in this paper. The experimental frogs were easily kept in good condition and usually lived from four to five weeks. TEMPERATURE SENSES IN FROG’S SKIN 89 The surface of the foot was the only area treated. Sometimes one foot was stimulated and the other kept as a check, but in most cases there was an alternate stimulation of the normal feet, or of the normal and the treated foot. Baths of water and ap- plications of stimuli were given at definite intervals which were kept uniform through each experiment. Preliminary experi- ments were made with each different kind of stimulation in order to find out what reaction might be expected. At each test a definite allowance of time was given, and if the reaction did not occur within that period the stimulus was regarded as producing no reaction and recorded as ~. In the tables the period just described is termed the reaction allowance. The interval which actually elapsed between the application of the stimulus and the reaction was taken in seconds with a stop- watch and recorded with the description of the response. No periods less than half a second were recorded. Intervals which separated stimulations sufficiently to prevent exhaustion were also selected by experiment. These have been designated the stimulation intervals. During the experiments the normal feet were kept in ‘bath water’ unless actually undergoing stimulation. In experiments made under cocaine treatment the foot was always returned to the cocaine solution after it had been immersed in the stimulant. In the preliminary part of this work a good deal of trouble was experienced by the washing out of the cocaine, so this procedure was found necessary. A solution of 1 per cent cocaine was the only anaesthetic used. OBSERVATIONS Responses to heat The first experiments of this series were made in order to deter- mine whether the frog’s foot would regularly react to heat, and if so to what degree of heat. A typical heat response was also looked for, a position or movement which should recur in many different individuals. Both feet of the experimental frog were kept in normal condition. In the first experiments considered 90 ANN HAVEN MORGAN (table 1) only the right foot was stimulated and the left served as a check. The. right foot was first immersed in water at 30°C. and at intervals of two minutes after that in baths increasing each time by 1°C. from 30°C. to 50°C.: no responses occurred below 39°C., and in some instances none below 43°C. As the heat was increased the vigor of the response was also increased TABLE 1 Reaction intervals in seconds of frogs’ right feet subjected to temperatures ranging at one degree intervals from 30° to 50°C. No reactions (indicated by ©) were obtained at temperatures of 388° or lower, hence this part of the table is condensed. Feet normal. Reaction allowance, 30 seconds. Stimulation interval, 2 minutes Number of individual................ 2 4 6 7 13 8 9 10 ll 11 3 Number of experiment .............. eat Sa Ta nae fre leaae He are 2 gas ee Temperature of room................ Po “19° ro “942 Po Pu 25° “99° “one “242 | 19°) Temperature of bath water.......... eo a7 18° “ae “18° | 20°, “20° “19° 18° “ore oe Temperature Of 1TOL.:...cces geese 22° “48° “19° “922° “19° “20°” 93° 20°, “19° “one “18° co @ co loo) je2) oO foe) co foo) Bi Ce ane 2 ene co | co | © | @ | @ | wo | wo | ©} o | 10] © FOG 2 aia oie a Sores sis Hn wo | 24) ao} wo] ao}! oo] oo} 15)17} 5} 12 AD! O95 Seen eeepras OS eg 9°| 14] @ | 22) © | 20) 91) 22 1 8) Gay Ba aera Raha eee ues 9} 11] ©} mo | ~ | 14] |} 15}] 10} 4] 13 AD ee Bs Was Ss aialcha Sune are 6} 10%) Ta) 15 eS SAS: |e) eee WAN 25G ae. ae Shei Se 4). Gp Diled2 Se SDE AL Gr seer FONG Soi ete a/cicts cis oleae B} 2 On 9 Teer o.|) | Ser 1D Be nah Ne Sea ele 2) 34 4) °8°| Suirolakan| sor Ay aan AEM sheets Sse sel es 2 EW BC Sh SMP Oot, OE ie: sea tate ASH. oJ desis ea dealels Wale 2h [2 ole Bab eee al od * Sah pees Mies CAS, Orc onan, Benen iors 24 2) 2) 425) Als) ode onl. act aces HOO ceramic Oe hur Cee ere ante 2 A DL 2a a Se ae al ee and the reaction interval became shorter and shorter. The heat response was a vigorous upward jerk of the foot, so uniform that usually no attempt to describe it has been made except by the word ‘jerk’ and the statement of the length of the reaction in- terval, i.e., the time between the application of the stimulus and the reaction itself. The next step was to discover whether there would be different results if the heat was applied with differences of temperature TEMPERATURE SENSES IN FROG’S SKIN 91 ereater than 1°. It will be remembered that some of the early workers (Goltz, 69; Tarchanow, ’71, ’72) maintained that if heat were increased slowly enough a frog might be subjected to an extreme degree without making responding resistance. No such results were secured in this investigation even when the heat was increased very slowly. A frog’s foot was placed in a beaker of water at 2°C., and a stream of warm water allowed to flow into this which brought it up to 45°C. with almost imperceptible slowness. Although care was taken to prevent the foot from TABLE 2 Responses in seconds to heat increasing by 5°C. at each stage of stimulation. Feet normal. Reaction allowance, 30 seconds. Stimulation interval, 2 minutes for nos. 8, 9, 10, and 5 minutes for no. 11. © = no reaction Number of individual.......... 8 8 8 8 10 10 9 9 11 11 11 11 ioe aan Iga aauedle a | Sh lap ee bos eel each Baud Foot stimulated................ ba ae RD vi: oR ae seeeliiay pes Pie aa ia Temperature of room.......... 24° ro ro ro “20°” “20° “25° 252 “29- 99° | 25° “25° Temperature of bath water.... “20° “20° 21° “2 “48° | 18° “20° “20°, “99° “99° 20°, “20°, Temperature of frog........... 93° | 93° “99° 9c° | 18° | 18° “93° | 93° | 99° | 99° | 23° oo [oo] co Loe) co foe) co foo} co co co Sn Osea ae oe CoM |fcows|(etcons |tlsy a |Rnconumcon | cco encom 25S 2Q ii etelal aL? ANT Oh Se eae ee Uf || 2% 4 yi AON AS | 122 df || 1B 4 4 AE Oe Oni aR ee BUA Se AL ee Ca] TS ra | Sa Carel rete ped Lay) Pa ONG Sa ee ei aete Ae be ero 2 2 1 i 3 4 4 3 being affected by this stream, the feet were invariably lifted before the heat had reached 45°C. Variations in the stimulation time and in the heat increments were also tried (table 2). Heat was increased by 5°C. at each stimulation and the right and left feet of the frog were dipped at one-, two-, and five-minute intervals. Of nine frogs used only four reacted at a degree lower than that to which the frogs responded which were subjected to 1°C. increases. The results showed the tendency toward the later reaction with gradually applied stimuli, but also suggested that individual idiosyncrasies 92 ANN HAVEN MORGAN were a factor and that certain frogs were especially sensitive to heat. This opinion was further justified by a series of tests repeated over and over on particular individuals. The right foot of each of these frogs was stimulated by hot water whose temperature ranged from 25°C. to 50°C. and at intervals of 5°, the experiment being repeated seven times consecutively (table 3). This was done to find out whether each frog would preserve its individual eccentricities toward heat in successive tests and TABLE 3 Responses in seconds of two frogs to hot water whose temperature was increased by 5°C.. at each stimulation. Each series repeated consecutively seven times. Feet normal. Reaction allowance, 30 seconds. Stimulation time, 2 minutes. © = no reaction Number of individual........... 11 11 14 Number of experiment........... 1 2 1 Hoot stimulated sn)... eit sersiers R R R Temperature of room............ 22° 22° 20° Temperature of bath water...... 20° 20° 18° Temperature of frog............. 22° Ppa Pie Stimulated by water at 25°C Deer Oud) OUI ROS! O SOs ©} CO} CO} CO} CO} GC} CO} GO| CO} CO} CO} CO} CO} CO] CO} CO] GO} CO} CO! CO} CO BORG Has Oke. reer Os 0 | co} co] co} co] co} co] co} co} co] co} oo] 00} co] co] a] co} c| 00} co] 0 SON, em MeN en eee @9 | 00 | co} a9} co} co} c9| co} co] co} co] co| co] co} 6) 7| 4/a]| 4) 3/12 AQR OLY er ee a Pe pee oo} 9/10) 5] 5} 7} 5} 5}/22} 0}19} 6)15} ©} 3} 5} 4] 3] 4) 2) 3 7 OURO fies Sere ai, 19° 120°) | 45°C. 20 4 jerked 2 jerked keeping the skin from being exhausted by tactile stimulation before the cocaine treatment was finished. Frogs which strug- gled had to be repeatedly adjusted, and this could hardly be accomplished without touching the frog somewhere. It was necessary, therefore, to perform a good many experiments and to select quiet frogs. Fifteen such frogs were used, and from their records the five in table 8 were selected. In these fifteen frogs 100 ANN HAVEN MORGAN the normal foot never failed to react to touch except in two cases, clearly caused by a faulty technique. The cocained foot failed to react to the strike of the falling shot at any time (table 8), but both the normal and cocained foot reacted regularly and with the same retardation which had been affected by the cocaine in previous experiments (tables 4, 5). The conclusion is that independent heat receptors are present in the foot of the frog and that a complete separation of the touch and the heat sense had been affected. TABLE 9 Responses in seconds to pain and heat. Pain stimulation by pricking outer side of fifth toe. Feet normal. Reaction allowance, 30 seconds. Stimulation time, 2 minutes SUES ee MU DEGREES C. a STIMULATED on mas Oe EXPERI= |———— = || BY NEEDLE BY MENT | Room eae Frog was Right foot Left foot 10 10° 202) 9? fetse Pricking 1 jerked 1 jerked 10 10 20 cae als Pricking 1 jerked 1 jerked 10 £0, | AZOP | TOS Ase 40°C.| 6 jerked 7 jerked 50 il PA ie 20? ie TSE: Pricking 1 jerked 1 jerked 50 a PA | ADS || “ksh Pricking 1 jerked 1 jerked 50 1 PN 1) PANS NP alte) 40°C.) 6 jerked 7 ‘jerked Independence of the responses to pain and heat. The method used in this separation was dipping the foot in water at 40°C. and pricking the skin on the outer side of the fifth toe. No degree of heat higher than 40°C. was used, because of the possibility that the higher degrees of heat might be painful and the two responses thus confused. Pricking the web between the third and fourth toes was first tried, the particular web being quite arbitrarily selected for stimulation. In some cases the foot would react to pricking done anywhere on this web, in other cases it would react to it in certain areas only, and in still other cases the foot would fail or almost fail to give any reaction at all to pricking anywhere on this web. Other webs were after- ward tried with much the same result. When the skin on the TEMPERATURE SENSES IN FROG’S SKIN 101 side of the toe was pricked, the normal foot never failed to react, though care was taken that the needle did not pass through underlying tissue. TABLE 10 Responses in seconds to pain and heat. Pain stimulations by pricking the side of the fifth toe. Right foot cocained 50 minutes. Left foot normal. Reaction allowance, 30 seconds. Stimulation time, 2 minutes. © = no response TEMPERATURES IN ne tae oF peewee STIMULATED Gharean rea pee EXPERI- BY NEEDLE BY MENT | Room peer Frog Meee oe Right foot Left foot 47 1 AREA IGs. 1-165 40°C co 1 jerked AT 1 IGE SIGS Mh OIG? Pricking 1 jerked 1 jerked 47 1 1B2. siGS al Ae 40° 10° co co 31 1 BRAD OW TESA That xe re) 1 jerked as toe touched water SL 1 Zee lo. elon is ro) 1 jerked as toe touched water 1 11 20° iG 16° 155 co co 17 11 20a) tos lor? 10" © 11 foot lifted, web spread 17 11 200i. o15° |G? 5° co | 18 foot lifted, web spread 17 11 20° 1G 16° iy co oo 17 12 DAO 25 14° 152 co oe) iy 12 ee eteoes el Oe © 1 toes extended, web spread 17 12 PA {PAS 14° NS oo) 1 jerked iV 12 PONE SN Ee ee i iy © 1 toes extended, web spread 45°C. (table 17). Responses to cold entirely ceased; those to heat remained vigorous with a lengthened reaction time. Independence of responses to touch and cold. The ease with which the sense of touch disappears from the skin has already been mentioned in connection with touch and heat. In that case responses to touch failed after the foot was im- mersed in cocaine 10, 15, 20 and 25 minutes, but reaction to water at 45°C. remained retarded but vigorous. In the same 108 ANN HAVEN MORGAN manner as with touch and cold, touch was very easily eliminated, and by reducing the cocaine treatment the cold sense could be preserved (table 18) in a very effective condition. Independence of responses to acid and cold. The sensation of cold was separated from the chemical sense in the skins of twelve different frogs. After thirty minutes of treatment with cocaine on the right foot the feet were alternately stimulated as usual. TABLE 17 Responses of the foot to heat and cold. Right foot cocained 20 minutes. Left foot normal. Reaction allowance, 30 seconds. Stimulation time, 2 minutes. ~ = no response TEMPERATURES IN DEGREES C. REACTIONS NUM- | NUM- BER OF | BER OF INDI- |EXPERI- Bath Stimu- 4 VIDUAL | MENT | Room z Frog lus Right foot Left foot water ms water 5 15 ToS) |, 15S 4) Ae De 20 1 toes stretched 5 15 15° | 15° | 14° | 45° | 13 jerked] 13 jerked 17 6 POR AAG eS cig Pe <) 1 toes stretched, webspread 17 6 | 19° | 19° | 19° | 40° | 25 jerked] 5 jerked 17 6 19° | 19° | 19° | 45° | 10 jerked] 6 jerked 17 6 19a AOE Ge 2° 00 1 stretched, web spread 17 6 | 19° | 19° | 19° | 40° | 20 jerked| 17 jerked 21 5 102) ge 7) AGS 2° 0 1 web spread 21 5 19° | 16° | 16° | 40° | 17 jerked} 10 quiver, jerked 24 Ges Or. | Geo) Tee Ze co 3 stretched 24 6 20°) | IGE. Gr 40" 9 jerked| 6 jerked The cold response was easily obliterated by the cocaine, but the acid caused a sharp upward jerk at long or irregular reaction intervals. This response itself seemed to differ little from that of the normal foot. Independence of responses to pain and cold. Pain responses were produced by pricking the skin of the frog on the side of the fifth toe—a procedure that produced more regular results than when the web was similarly stimulated. Care was always taken that the needle did not go into the deeper tissues. The foot was TEMPERATURE SENSES IN FROG’S SKIN 109 TABLE 18 Responses in seconds of the foot to touch and cold. Right foot cocained 10 to 25 minutes. Left foot normal. Reaction allowance, 30 seconds. Stimulation time, 2 minutes. © = no response TEMPERATURES IN DEGREES C, |LENGTH REACTIONS NuM- | NUMBER OF co- BER OF ORM ae | eee eS inca || carne INDI- EXPERI- USED | TREAT- ; Stimu- a2) MENT i \cRoom path. Frog “hus. Soe MENT Right foot Left foot 23 18 Bie 18° ig 5 25 co 1 lifted 23 18 Doe LS SLES we 25 | 6 lifted 2 lifted 38 20 Pile 16° rises 5 15 co 1 lifted 38 20 pile 16° ibys 5 15 fe) 1 lifted 38 DOR eo | rebte- ener dy < Zee is 15 | 1 lifted 10 lifted 39 32 Pile ie 18° 5 25 20 1 jerked 39 32 Pe lhecgel ica (pel oo oe 25 | 1 stretched 1 jerked 40 1 20° yf 19° 5 15 oe) 1 lifted 40 1 20° 172 19° Ae 15 | 1 shiver 1 lifted 41 1 DNS 18° 19° 5 10 o0 1 jiggle 41 1 22° edae | 19° 2° 10 |1web spread} 1 jerked TABLE 19 Responses in seconds of the foot to acid and to cold. Right foot cocained 30 minutes. Left foot normal. Reaction allowance, 80 seconds. Stimulation time, 2 minutes. © = no response TEMPERATURES IN DEGREES C. REACTION ACID : STIMU- Stimu- ifr Lus HCl NUMBER | NUMBER OF OF Lad nal a == sy INDI- EXPERI- Bath VIDUAL | MENT | Room epiter Frog Right foot Left foot water per cent 24 of Dec. |e elGrs | 18° 2° ©0 3 body twisted 24 7 2g elon 18° 0.5 | 15 jerked | 2 jerked 24 8 Seen uelon 16° 2S © 1 toes extended jerk 24 8 iS || TIS |] alas 0.5 1 jerked | 1 jerked 28 3 1925 eel Ogee Oe De 00 1 web spread 28 3 19° iG= 16° 0.5 | 20 jerked | 1 jerked 32 it Doe i Moa we eles ca Be co 1 web spread 32 il DREN OSS | TIES 0.5 | 15 jerked | 3 jerked 110 ANN HAVEN MORGAN supported when being pricked and by putting the needle only through the side surface it was thought that this was avoided. Twelve frogs were stimulated by pricking and by water at 2°C. Reaction to cold was eliminated by cocaine treatment in as short a time as ten minutes. No effect upon the pain response or its reaction could be discerned. TABLE 20 Responses in seconds of the foot to pain and cold. Right foot cocained 10 to 30 minutes. Left foot normal. Reaction allowance, 30 seconds. Stimulation time, 2 minutes. © = no response NUM- NUM- Maree Wolk i COCAINE NCIS BHR ORS BERION |= se Se STIMULATED TREAT- ENE EXPERI- Stim- BY NEEDLE MENT \ABSNEE EL || ALHOESEN VHT BY oyovaat Bath Frog | ulus Right foot Left foot water Rint. ie minutes 23 be a2 8 iN I en a 30 oo 1 spread, lifted 23 iL WAXD? |i ig’ |p ilkss Pricking 30 | 1 jerked | 1 jerked 43 LORS ay aia ae = ae 20 ro) 1 web spread 43 1 oe oe | ae Pricking 20 | 1 jerked | 1 jerked 23 DN DOR ASS Mas ct 28 20 <0 1 -web spread 23 Py || PAO Messy Tb Pricking 20 | 1 jerked | 1 jerked 4A) Dee 212 15°) 16° 4) 22 20 co co 44 Th PAS | S| a= Pricking 20 | 1 jerked | 1 jerked De 1 PANS AW aa Wee | Pe 15 co 1 web spread 23 Peoe2Ors dae Ge Pricking 15 | 1 jerked | 1 jerked 17, 1) PAS alyee Wax 10 ©0 1 stretched 17 1 a aah Ne aires Pricking 10 -| 1 jerked | 1 jerked TEMPERATURE SENSES IN FROG’S SKIN jul DISCUSSION AND RESULTS A temperature sense is easily demonstrable in the frog’s skin. There is a response to heat characteristic in form and reaction time. The lowest degrees of heat which stimulate the skin lie somewhere between 35°C. and 41°C. If the skin be stimulated by water increasing in heat by 5°C. at each stimulation from 30°C. to 50°C., the first response may be expected at 35°C. If the same series is followed except that the heat be increased by 1°, the first response may occur at 40°C. or 41°C. The skin responds to the higher degrees of heat with great regularity. As the heat is increased from 35°C. to 50°C. the reaction time decreases with more or less regularity from long intervals (25, 15, 12 seconds) to short ones (2, 1 second). It will be remembered that some of the early workers main- tained (Goltz, ’69; Heinzmann, ’72) that if reflex frogs were stimulated gradually enough with increasing heat they could be subjected to considerable warming without resistance. My investigations show only a slight agreement with them which has been mentioned. It has not been possible to stimulate the foot with increasing heat beyond 43°C. without response, even when the frog’s foot was suspended in a beaker of water at 20°C. and the heat almost imperceptibly increased by an inflow of warm water. The long reaction time of the heat response agrees with v. Frey’s contention that the heat receptors are in the deeper and the cold receptors in the more superficial layers of the skin. It has been possible to isolate the temperature sense from the tactile and chemical. This has been done by treatments with 1 per cent solution of cocaine. Crozier (16) used this method in separating the tactile and chemical senses, and by it Cole (710) eliminated response to pain, but preserved sensitiveness to taste. The separation of temperature from other senses gave the follow- ing results. Response to acid (0.5 per cent hydrochloric) per- sisted beyond response to heat. Pain persisted beyond heat; heat and cold beyond touch. With the thermal and chemical stimulations care has been taken to immerse the same amount of surface. It has of course not been possible to make any 12 ANN HAVEN MORGAN equivalence between chemical and tactile stimuli or degrees of heat and cold. Granting this necessary inaccuracy, 45°C. heat and 2°C. cold were selected as sufficient extremes to be set against each other. There is a definite cold sense present in the frog’s skin. When the foot was immersed in water of decreasing temperatures, the first responses occurred at 10°C. Contrasted with that of heat, the interval between stimulation and response was an inconsider- able period and could not be accurately taken with a stop-watch. The responses to cold were of two types, a sudden rigidity of the muscles of the leg, with a spreading of the toes and web, or an upward jerk instantly following the contact of the toes with the water. The latter action was less frequent and usually occurred after stimulation by severe cold or in unusually sensitive frogs. Such responses differed from heat responses only in the length of the reaction time. The sense of cold may be wholly eliminated by cutting the nerve, removing the skin, or by cocaine treatment. It can be shown to be independent of heat, and the tactile and chemical senses by the same treatment. In such comparisons sensation to cold disappears, acid remains; cold disappears and heat and pain remain, but cold remains while touch is eliminated. The frog’s skin is indifferent to temperature of 10°C. or 15°C. to 35°C., whether the stimulation be made by gradual increases or whether it be given suddenly at one selected degree. SUMMARY The skin of the frog contains well-defined receptors for heat and for cold. The heat receptors have a comparatively long reac- tion time. The heat receptors are stimulated by 39°C. to 48°C.; the cold receptors at 10°C. This response is immediate and becomes more vigorous as the cold is increased. The typical response to heat is an upward jerk of the leg. The typical response for cold is a rigidity and tenseness of the muscles, but there may be an upward jerk similar to that of the heat response. Responses to heat and cold may be separated from each other and from the tactile and chemical senses. TEMPERATURE SENSES IN FROG’S SKIN cw Ses BIBLIOGRAPHY ARCHANGELSKY, P. 1873 Ueber den Einfluss der Wirme auf das Nerven- und Blutgefiiss-system des Frosches. Original paperin Russian. Abstract in Jahresb. der Anat. und Physiol. (Hofmann und Schwalbe), Bd. 2, S. 555-559. Basdk, E. 1913 Zeitschr. Sinnesphysiol., Bd. 45.” Brooks, E.S. 1918 Reactions of frogs to heat and cold. Amer. Journ. Physiol., vol. 46, pp. 493-501. BrRowN-SEQUARD 1847 Note sur la durée de la vie des grenouilles en automne et en hiver, aprés l’extirpation de la moelle allongée et de quelques autres portions du centre nerveux cérébrorachidien. Comptes rendus de l’académie des sciences, Paris, T. 24, pp. 363-364. CayRADE, J. 1864 Recherches critiques et expérimentales sur les mouvements réflexes. Thése pour le doctorat en médecine, Paris. Corz, L.J. 1907 An experimental study of the image-forming powers of various types of eyes. Proc. Amer. Acad. Arts and Sci., vol. 42, pp. 335-417. Coir, L. W. 1910 Reactions of frogs to chlorides of ammonium, potassium, sodium, and lithium. Jour. Compar. Neurol. and Psychol., vol. 20, pp. 601-614. Crozier, W. J. 1916 Regarding the existence of the common chemical sense in vertebrates. Jour. Comp. Neur., vol. 26, pp. 1-8. Foster, M. 1873 On the effects of a gradual rise of temperature on reflex actions in the frog. Journ. of Anat. and Physiol., vol. 8, pp. 45-53. Also, Studies from Physiological Laboratory, Univ. Cambridge, 1873. FratscHer, C. 1875 Ueber continuirliche und langsame Nervenreizung. Jenaische Zeitschrift, Bd. 9, N. F., Bd. 2 (1875), S. 1380-138. Frevuspera, A. 1875 Ueber die Erregung und Hemmung der Thitigkeit der nervosen Centralorgane. Archiv. f.d. ges. Physiol. (Pliiger’s Archiv.), Bd. 10, S. 174-208. Goutz, F. 1869 Beitrige zur Lehre von den Functionen der Nervencentren des Frosches. Berlin, 1869, 127. HetnzMann, A. 1872 Ueber die Wirkung sehr allmidliger Aenderungen ther- mischer Reize auf die Empfindungsnerven. Archiv. fiir die ges. Physiol., Bd. 6 (1872), S. 222-236. KnaurtHe, K. 1891 Meine Erfahrungen iiber das Verhalten von Amphibien und Fischen gegeniiber der Kilte. Zool. Anz., Bd. 14, 8. 109-115. KorAnyt, A. v. 1892 Ueber die Reizbarkeit der Froschhaut gegen Licht und Warme. Centralbl. fiir Physiol., Bd. 6, 8. 6-8. Kunpe, F. 1860 Der Hinfluss der Wirme und Electricitit auf das Riickenmark. Archiv. f. path. Anat. und Physiol. (Virchow’s Archiv.), Bd. 18, S. 357-360. LANGENDORFF, O. 1877 Die Beziehungen des Sehorganes zu den reflexhem- menden Mechanismen des Froschgehirns. Archiv. fiir Anat. und Physiol. (Physiol. Abth.), 1877, S. 435-442. Lorsmr, W. 1905 A study of the functions of different parts of the frog’s brain. Jour. Comp. Neurol. and Psychol., vol. 15, pp. 355-373. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 35, NO. 2 Resumen por el autor, 8. R. Detwiler. Experimentos sobre la transplantacién de los miembros en Amblystoma. Nuevas observaciones sobre las connexiones nerviosas periféricas. El cambio de posicién del rudimento del miembro anterior de Amblystoma en un numero determinado de segmentos poste- riores 0 anteriores a su posicidn normal no afecta del mismo modo el cambio correspondiente de la contribucion del nervio seg- mentario al plexo braquial. Existe una marcada tendencia en el miembro transplantado a recibir inervacién del nivel del miembro normal en la médula. Cuando se transplanta el miembro en una positién anterior de dos o tres segmentos las porciones distales de los nervios del miembro normal crecen anteriormente, en contra de la oposicién mecanica de los miotomos en vias de desarrollo, con el fin de efectuar una conexién funcional con el apéndice heterotédpico. La posicién y extensidn del rudimento transplantado determinan solamente hasta cierto punto el numero de nervios segmentarios que contribuyen al plexo. Las pruebas acumuladas en estos y en previos experi- mentos (Detwiler, ’20) indican que existe una relacién, de de- sarrollo entre el miembro y sus nervios normales, la cual es mas intima en cardcter que cualquier otra asociacién semejante entre estos nervios y otras estructuras. La funcién de los miem- bros transplantados esté condicionada por cuatro factores prin- cipales: I) La estructura incompleta de la cintura escapular; 2) las deficiencias musculares; 3) la inervacién periférica de- fectiva, y 4) las conexiones defectuosas dentro del sistema ner- vioso central. La funcién mas completa de los miembros trans- plantados que reciben nervios del nivel tipico del miembro se atribuye al hecho de estar en conexién con mecanismo central de reflejos adecuado para la motilidad normal. Como apéndice al trabajo el autor incluye una consideracién tedrica de los estt- mulos que jJuegan papel en las conexiones periféricas normales. Translation by José F. Nonidez Cornell Medical College, New York AUTHOR’S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, JANUARY 16 EXPERIMENTS ON THE TRANSPLANTATION OF LIMBS IN AMBLYSTOMA FURTHER OBSERVATIONS ON PERIPHERAL NERVE CONNECTIONS 8. R. DETWILER The Anatomical Laboratories, Yale School of Medicine, New Haven, Connecticut, and the Peking Union Medical College, Peking ; THIRTY-TWO FIGURES INTRODUCTION In a recent paper dealing with the function and peripheral innervation of transplanted limbs (Detwiler, ’20), reference was made to the striking tendency on the part of the normal limb nerves to supply the appendage when the latter was transplanted a considerable distance away from its typical site. In these experiments it was found that shifting the anterior limb a given number of body segments caudal to its normal position on the same embryo did not effect a corresponding shifting of the seg- mental nerves contributing to its plexus. Limbs transplanted caudally the distance of four and five body segments received one or more nerves from the original limb level of the cord (Detwiler, op. cit., table 2, p. 133). The facts were also brought out that transplanted limbs which received innervation from the normal limb level of the cord functioned more perfectly than did those whose segmental nerves were derived from the post-limb level, and that the limbs showed a gradually increasing loss of function as they were trans- planted farther and farther away from the normal situation. This was attributed to diminution of connections within the central nervous system rather than to a corresponding decrease in effective peripheral innervation or to structural deficiencies within the limb and the shoulder-girdle. The cumulative evi- dence suggested that the more perfect function of limbs whose 115 116 S. R. DETWILER nerves were derived from the original limb level of the cord had its explanation in the fact that the appendage in such cases was connected with a central nervous mechanism adequate for normal motility. The remarkable increased growth of the limb nerves and the entirely new pathways which they followed to reach their normal terminal end organ (limb) when the latter was transplanted considerable distances caudal to its normal position strongly suggested the possibility that the limb might exert some directive influence upon the segmental nerves contributing to its plexus. Of considerable significance were the facts that the nerves of the limb level, especially the fourth and fifth, grew greater dis- tances to meet the transplanted limb than did those coming from the more caudal segments, and that the farther away from the normal level the limb was transplanted, the less was its liability to receive nerves from segments situated anterior to the position of the limb, and the greater its tendency to receive nerves from segments corresponding to the position occupied by the limb. Experiments on various anuran forms (Braus, ’05, and Harrison, ’07) have shown, in general, that when a limb is trans- planted to an abnormal (heterotopic) position it becomes in- nervated from that part of the central nervous system of the host corresponding to the position occupied by the implanted limb rudiment. In the majority of these experiments the rudiment was transplanted at a stage when the principal nerve paths were already laid down. Accordingly, when the wound in the host was made for the reception of the transplant, the terminal branches of the nerves in that region were severed and the im- planted rudiment was placed in close apposition to the cut ends. The possibility of the limb’s exerting any directive influence on the nerves could hardly be tested in such cases, since, as has already been pointed out, the limb buds were placed in the direct pathway of already formed spinal nerves whose ends were severed in preparing the wound, and it is to be expected that these nerves would continue their growth into the rudiment so placed. In the urodele, Amblystoma punctatum, the forelimb rudiment becomes localized quite early, even before the closure of the TRANSPLANTATION OF LIMBS IN AMBLYSTOMA O17 medullary folds (Detwiler, 717, 718), and it can be transplanted prior to the period when outgrowth of the spinal nerves begins. Here it is possible, therefore, to test whether or not the trans- planted end organ (limb) exerts any directive influence on the normal limb nerves at a time when the latter undergo their initial outgrowth. The fact that limb nerves made connections with their proper end organ when the latter was transplanted four and five seg- ments caudal to the normal site scarcely seemed explainable on purely mechanical grounds any more than proper selective periph- eral connections under normal conditions of development can be explained solely by mechanical agencies. ‘These nerves did not terminate at the limbless area as they did upon simple extirpa- tion of the limb, but they continued their growth caudally until the heterotopic limb was reached and connections were made. The caudoventral elongation of the myotomes undoubtedly acts as a mechanical factor in directing the nerves in this same general course, since they lie in grooves between the elongating muscle segments. We might conceive of the extended caudal growth of these nerves as due in part to an insufficient number of muscles in the general limbless region for the accommodation of all of their axones. The non-supplied fibers, being directed by the caudo- ventral elongation of the myotomes, would continue to grow until the limb muscles were reached and connections were made. An interpretation of this nature, however, would not explain why the nerves should finally enter the limb when they do reach it, nor why they should take priority over the nerves coming from segments of the cord corresponding to the position now occupied by the transplanted limb. This evidence, pointing to the direct- ive influence of the transplanted limb upon its normal nerves, suggested the experiments which are taken up in the present paper. 118 S. R. DETWILER ANATOMICAL Since these transplantation experiments have been carried out from the standpoint of the nerve connections and functional behavior of the hmbs, it becomes necessary, for proper discussion of the results, to consider briefly the anatomical factors involved in the regulation of the motility of the heterotopic appendage. Four such factors have been discussed previously (Detwiler, ’19, ’20) :—1) the completeness of the shoulder-girdle in the hetero- topic position, 2) the degree of differentiation of the shoulder and limb muscles, 3) the completeness of peripheral nerve connections with the above muscles, and, 4) the character of the connections within the central nervous system. 1. Shoulder-girdle Since the shoulder-girdle has the character of a mosaic (Detwiler, 718), its degree of development in the transplanted position is variable, depending on the size of the graft and the region from which it is taken. When a typical limb rudiment is transplanted (fig. 1), only the tissue normally developing into the more central portion of the girdle isincluded. The localized rudiments of the more outlying portions of the girdle (supra- scapula and the ventral portion of the coracoid) lie beyond the limits of the tissue typically included in the limb graft, and con- sequently undergo development in situ following excision of the transplant. The girdle which develops in the heterotopic position is always reduced in size and is qualitatively incomplete. There is considerable evidence in many cases, however, to show that compensatory hyperplasia from the dorsal and ventral portions of the reduced heterotopic girdle takes place as the larva matures, so that the final size of the girdle may almost equal that of the normal. Such cases will be referred to later. No conclusive evidence has been obtained from transplantation experiments in Amblystoma to show that the girdle rudiment in this form, in spite of its developmental intimacy with the lmb, constitutes a totipotent system such as the limb itself is. Braus (09, p. 271), however, maintains that in Bombinator the girdle, like the limb, constitutes an equipotential restitution system. TRANSPLANTATION OF LIMBS IN AMBLYSTOMA 119 Because of the mosaic nature of the girdle and the subsequent variability in its degree of development in the transplanted position, the motility of the attached appendage is markedly affected by developmental deficiencies in this structure. 2. Shoulder muscles In a previous description of the shoulder muscles in larval Amblystoma (Detwiler, ’20, p. 123) it was pointed out that of the thirteen shoulder muscles which typically develop, nine only connect the shoulder with the limb. The remaining four serve for anchorage of the shoulder and are of myotomic origin. The muscles connecting the shoulder with the extremity are of somato- pleural origin, as is the limb itself, and their rudiments are local- ized in the tissue included in a typical limb dise (fig. 1). Con- sequently, when the limb rudiment is transplanted, these muscles develop in the heterotopic position (Detwiler, op. cit., pp. 128 and 136). Their nerve supply is derived from the brachial plexus, and although all may be differentiated in the heterotopic position, their nerve supply is by no means constant and many cases develop in which these muscles receive defective nerve supply or are completely lacking in innervation—a condition which markedly affects the degree of motility of the transplanted ex- tremity. The incompleteness of their development secondarily accompanying the smaller area for attachment in the reduced gir- dle also serves to limit the extent of movements of the extremity on the shoulder. 8. Brachial plexus The brachial plexus is normally derived from the ventral rami of the third, fourth, and fifth spinal nerves (fig. 4). The develop- mental evidence indicates that, under normal conditions, these outgrowing nerves effect connection with the limb rudiment when it occupies its maximal extent (anterior border of the third myotome to the posterior border of the fifth, as shown in figure 1) and that convergence of the nerves and plexus formation results secondarily from the concentration of the limb rudiment 120 Ss. R. DETWILER into the definitive limb bud which centers ventral to the fourth myotome. From this it would appear that the number of seg- mental nerves entering the plexus would be determined by the extent of the limb rudiment at the time when initial connections are made. It was Firbringer (’79) especially that called attention to the fact that the nerve plexus from which a limb is supplied might in two cases have a different segmental origin, yet the distribution of the limb nerves arising from the plexus might be exactly the same in each. Both Firbringer (op. cit.) and Gegenbaur (’98), who closely held to the idea that muscle and nerve form an in- separable unit, admitted the difficulty of satisfactorily explaining such segmental variations. From the results of his own experi- ments, Harrison (’07) interpreted segmental differences as due to the position and extent of the limb rudiment at the time when initial connections of the nerves were made, this serving as an index of the number of nerves contributing to the plexus. From the identity of intrinsic distribution, regardless of metameric origin, Harrison concluded that the mode of segregation and the growth of the structures within the limb determine the specific intrinsic nervous pattern. From previous observations (Detwiler, ’20) and from the results obtained in the present experiments, considerable evidence has accumulated to show that the position and extent of limb rudi- ments lying beyond the confines of the normal position are not the only factors in determining the source of origin of the nerves contributing to its plexus. This question will be more fully considered in the discussion. 4. The first and second spinal nerves A consideration of the normal pathway of the ventral rami of the first and second spinal nerves is herewith given, since the hypobranchial region is involved in the present experiments. The course and terminal connections of the ventral rami of the first and second spinal nerves in Amblystoma punctatum are found to be very similar to those described by Norris (13) for TRANSPLANTATION OF LIMBS IN AMBLYSTOMA it Siren lacertina, and in general accord with the arrangement characteristic of urodeles (Coghill, ’02, ’06; Driiner, ’01, ’03; Kallius, 01, and Norris, 713). The ventral rami of the first and second spinal nerves, typically, unite to form the hypoglossal trunk which supplies the hypobranchial musculature. In Amphiuma, however, Norris (’08) described the hypoglossal trunk as being derived solely from the ventral ramus of the first spinal nerve. A résumé of the arrangement of the hypoglossal complex in both anurans and urodeles is given by Black (17). In Amblystoma punctatum the first spinal nerve is entirely motor. It divides into dorsal and ventral rami. The latter at first passes caudolaterally through the dorsal axial musculature to the lateral border of the pharynx. Then it curves laterally and ventrally around the pharynx, and at a point just caudal to where the ramus intestino-accessorius X breaks up into its component divisions, it unites with the anterior prolongation of the ventral division of the second spinal nerve. The common trunk thus formed (hypoglossal), after contributing a branch to the ramus intestino-accessorius X, passes anteriorly along the dorsolateral border of the ventrolateral musculature beneath the pharynx. More anteriorly, it runs slightly lateral to this muscle mass and supplies its component segments (m. sterno- hyoideus). The distal anterior prolongation of the nerve enters the substance of the m. geniohyoideus, in which it travels for a considerable distance, giving off branches. The terminal por- tion emerges from the m. geniohyoideus and finally breaks up in the m. genioglossus. The second spinal nerve arises by both dorsal and ventral roots and presents a small ganglion. The proximal part of the ventral ramus passes posteriorly for a short distance through the longi- tudinal trunk musculature, then passes laterally to the dorso- mesial border of the ventrolateral musculature. Here it curves anteriorly, and after supplying motor fibers to the ventrolateral musculature, it unites with the first spinal nerve. According to Norris (13), the ventral ramus of the second spinal nerve in Siren contributes a branch to the brachial plexus which in this form is made up principally from the third and fourth spinal 22 S. R. DETWILER nerves. In Amblystoma punctatum there is normally no com- municating branch to the brachial plexus which is formed from the ventral divisions of the third, fourth, and fifth spinal nerves (fig. 4). The above observations concerning the pathways and connec- tions of the first and second spinal nerves are based on a study of serial transverse sections of larvae ranging in age from fifty to seventy days after the closure of the medullary folds. The arrangement of these nerves is not exactly the same in all cases studied, but the variations are of only minor significance. EXPERIMENTAL The experiments here carried out consisted in transplanting the fore-limb rudiment varying distances anterior to its normal position, with the idea of testing whether the limb nerves could be induced to change their direction of growth and effect connec- tions with the displaced rudiment. Any effort on the part of the nerves to make connections with the limb so placed would necessitate a reversal of their course as compared with that taken by them when innervating limbs placed caudal to the normal site (fig. 5). Under such conditions the nerves, in order to reach their displaced end organ, would meet with considerable opposi- tion, since the differentiating myotomes tend to direct the seg- mental nerves in a caudoventral pathway. Normal connections made under these conditions would strongly support the idea that there exists a greater attractive influence for the nerves in the developing end organ than in extrinsic structures. The experiments were carried out upon embryos in the tail- bud state (fig. 1). The circle ventral to the pronephros (pn) indicates the position of the limb rudiment. ‘The slightly raised eminence just anterior to the limb rudiment constitutes the gill swellings from which develop the three external gills (cf. figs. 1 and 9). The position of the first and second myotomes with respect to the gill swellings is also seen. TRANSPLANTATION OF LIMBS IN AMBLYSTOMA £23 Series A In the first series of experiments (twenty in number) the limb bud was excised and reimplanted the distance of two segments anterior to the normal position. In preparing the wound for the Fig. 1 Drawing of Amblystoma embryo in the tail-bud stage (stage 29). The circle ventral to the pronephros (PN) indicates the position of the fore-limb rudiment. MY3, third myotome; MY1, first myotome; GS3, third external gill swelling. X 10. Fig. 2 Drawing of Amblystoma embryo, showing the fore-limb rudiment transplanted the distance of two segments anterior to its normal position (cf. fig.1). The ectoderm and mesoderm of the third gill swelling was removed prior to reimplantation of the limb. MY1, myotome 1; GS/, first gill swelling. X 10. Fig. 3. Drawing of Amblystoma embryo following excision of the ectoderm and mesoderm of the gill region and the reimplantation of the right fore-limb rudiment (LR) into the excavated area. Denuded limb area covered with indif- ferent ectoderm. X 10. reception of the transplant, an area of ectoderm with the under- lying mesoderm ventral to the first and second myotomes was excised. ‘This excision involved the tissues normally forming the third gill (fig. 2). The excavated limb area was cleaned of all free mesoderm cells, but was not covered. The results are 124 S. R. DETWILER given in table 1 A. Four cases of the eleven positive experiments yielded normal results. The limb bud, although originally two segments anterior to the normal and in the region of the third gill, gradually shifted caudally during differentiation and finally assumed the orthotopic (normal) position. The third gill, although late in making its appearance and at first very much smaller than its counterpart, gradually developed to full size. In this series of experiments no normal limbs developed in the heterotopic position. Those which did permanently occupy the TABLE 1 A. Showing the effects of removing the gill ectoderm and mesoderm from the posterior (third) gill swelling (fig. 2) and transplanting a limb into the excavated area, B. Showing the results of removing the ectoderm and mesoderm from the entire gill swelling (fig. 3) and transplanting a limb into the denuded territory . 2) ' EE n a og am Bye ra] a 1 86 ° 2 5 z z, 3 = B S ala| g a ag 2 > z Zest ||) os a ad bs | Bom] 2, goa | ho | sa SI am | Seo) Ba 2 a ag ae | Soe | Sia | "oO BN sak oie i I ea ea AR Ba ao | eae | sa | 2a |) os Ze a ac |ze.,/ 92 2S 5 5 E rs ao oe mOH ae a4 oH ae 5, Be DIKE Se Xe AG 520 11 11 41 ye 5 6 5 1 6 0 Bal soo 22 pepe 3 17 1 8 14 3 0 12 10 1 In all four cases the limb shifted caudally during development and finally occupied the orthotopic position. The above table does not include one series of forty operations which were made upon unsatisfactory material. The serial case numbers, however, have been preserved (table 2). transplanted position developed into abnormal appendages, of which 45 per cent were abortive. These limbs, in their initial development, assumed the posture shown in figure 10, but, curiously enough, they soon began to take on the typical posture of the third gill, which failed to develop (fig. 6), and they never fully differentiated into free motile appendages. The first and second gills in some cases developed quite normally, whereas in others they were incomplete and abnormal in appearance. In case 15 (fig. 7) the limb, except for an imperfectly redupli- cated hand, was normal and it functionated to a considerable extent. The third gill was wanting, although the first and second TRANSPLANTATION OF LIMBS IN AMBLYSTOMA $25 gills were practically normal in appearance. In case 19 an appendage with a reduplicated forearm and hand developed in place of the third gill. A complete limb which was normal in function regenerated at the original site. The reduplicated appendage exhibited only limited movements in the forearm and hand. In cases in which complete regeneration of the gills did not occur, the limb retained its transplanted position. Such rudi- ments, although having a normal aspect at the beginning, soon took on the typical posture of a developing gill, and in all cases distorted and abortive development followed. In connection with the question of gill development, Ekman (713, 714) has shown that the factors for the outgrowth of the external gills in various anuran forms (Rana fusca, Rana esculenta, Bombinator, and Hyla) reside entirely in the ectoderm. His experimental results have proved that this ectoderm, which becomes localized before the closure of the medullary folds (stage 1), possesses the properties of self-differentiation. He has also shown that this gill-producing faculty resides not only in the ectoderm of the immediate gill region, but that prospective gill forming potencies extend for a considerable distance beyond the limits of the immediate gill region, particularlyin theregion of the heart and the pronephros. According to this author, the capacity of the outlying ectoderm to regenerate gills is not the same in all forms—it being higher, for example in Bombinator than in Rana fusca. The equipotential properties of the gill ectoderm in Amblystoma is secondarily brought out in the present experi- ments, particularly in series A (table 1), in which it has been shown that after complete extirpation of the tissue of the third gill swelling, complete gill formation occurred in more than 50 per cent of the cases, and that the limb rudiment, in these cases, which originally occupied the region of the third external gill, underwent a caudal displacement so as to lie eventually posterior to the normal gill region. The difficulty of making successful limb transplantations was likely due to the presence of the inherent gill-producing property 126 Ss. R. DETWILER of the ectoderm in the general region surrounding the trans- planted rudiment. Developmental conflicts arose between the gill-forming ectoderm and the limb mesoderm, producing not only various distortions in both gills and the limb, but frequently resulting in almost complete suppression of the latter. Further experiments showed that complete limb differentiation in the gill region could not be expected with any certainty unless complete suppression of the gills was first accomplished by ex- cising much larger areas prior to making the transplantation. Several cases were obtained, however, in which normdl limb differentiation did take place, even though one or two atrophic gills did develop. Series B In the second series of experiments the ectoderm and mesoderm from the entire gill swelling was removed (fig. 3). The trans- planted limb rudiment in these cases occupied a position ap- proximately the distance of three segments anterior to the normal (fig. 3). The excavated limb region in all cases was covered with indifferent ectoderm taken from the caudal portion of another embryo. The results of these transplantations upon the develop- ment of the limbs and the gills are summarized in table 1 B. Although the ectoderm and mesoderm from the entire gill swell- ing was removed, the tabulation shows that 55 per cent of the cases developed abortive and abnormal gills. That these struc- tures develop from tissue lying beyond the confines of the immedi- ate gill swelling is without doubt, and this observation confirms that of Ekmann (op. cit.).. Because of the inherent ability of the surrounding ectoderm to produce gills, the percentage of ab- normal limbs was high. Even when only one or more abortive gills were formed, the ectoderm surrounding the base of the developing limb frequently migrated out upon the appendage, causing it to become secondarily fused to the side of the body wall (fig. 14). Such limbs, although structurally complete and with considerable nerve supply, were unable to enjoy freedom of movement. TRANSPLANTATION OF LIMBS IN AMBLYSTOMA sear All of the operated animals were kept under daily observation during the first twenty days. A large number in which the limb and the gills showed marked abnormalities were fixed and only the more important cases were kept. Serial transverse sections were made of eight cases ranging in age from twenty-six to sixty days after the operation. The sections were cut 10 » thick and stained with Ehrlich’s haematoxylin and erythrosin. Fig. 4 Graphic reconstruction of the normal left brachial plexus of Ambly- stoma larva AS40.6, preserved sixty-eight days after the operation. X 20. (Det- wiler, ’20, fig. 13.) Fig. 5 Graphic reconstruction of the right brachial plexus of case AS426, showing the segmental nerve supply to the right anterior limb when transplanted the distance of four segments posterior to the normal position. 20. (Detwiler, 20, fig. 10.) Description of cases Case AA2S,;. This experiment was carried out under the conditions described in table 1 A. There was considerable delay in the growth of the limb bud. Only the first and second gills developed, both of which were rather smaller than normal. At fifteen days after the operation the limb rudiment, which oc- cupied the position of the third external gill, took on the charac- 128 S. R. DETWILER teristic posture of a developing gill and pointed almost dorsally as is seen in figure 6. The limb later changed its orientation and developed into an appendage with considerable freedom of movement, and which was normal except for a reduplication of the digits (fig. 7). Examination of sections of the above case showed that the glenoid fossa was situated slightly dorsal and approximately the Fig.6 Lateral view of case AA2S;; drawn fifteen days after operation, showing typical gill posture of the developing appendage (LB) which occupies the position of the third external gill (fig. 2). Normal limb posture at approximately same stage is seen in figure 23. X 10. Fig.7 Dorsal view of case AA2S;; drawn fifty-three days after the operation. The transplanted limb occupies the region of the third external gill (fig. 2). bas distance of two segments anterior to the normal position. The shoulder-girdle was well formed. All of the shoulder muscles which typically develop in the heterotopic position from a typical limb-bud transplantation (Detwiler, ’20, p. 136) were present. The m. coracobrachialis longus and the m. coracobrachialis brevis were small and abnormal. The remainder of the shoulder mus- cles were typical. The musculature of the arm was somewhat deficient on the extensor surface. The limb and shoulder mus- TRANSPLANTATION OF LIMBS IN AMBLYSTOMA 129 cles were supplied by a plexus derived from the ventral rami of the second, third, and fourth spinal nerves (fig. 8), the latter two of which are normal limb nerves. It is seen that the distal portions of the nerves have grown anteriorly to effect connection with the heterotopic limb. The typical course of the third and fourth nerves is shown in figure 4. The ventral ramus of the second spinal nerve, which typically unites with that of the first to form the hypoglossal, is contributed almost entirely to the brachial plexus. The completeness of peripheral connection with the shoulder muscles is given in table 3. Fig. 8 Graphic reconstruction of the ventral rami of the second, third, and fourth spinal nerves, showing their direction of growth and contribution to the limb plexus in case AA2S;;. Arrow A indicates position of transplanted limb; arrow B designates approximate level of normal limb. For normal pathways of third and fourth spinal (limb) nerves (fig.4). X 25. The ventrolateral musculature in this case is almost entirely wanting. The ventral ramus of the first spinal nerve supplies the fragmentary m. sternohyoideus, the anterior prolongation of which is entirely wanting as is the entire m. geniohyoideus and the m. genioglossus. The hyoid cartilage on the operated side was also wanting. The deficient development of the m. sterno- hyoideus and the complete absence of its derivatives, the m. geniohyoideus and m. genioglossus, indicate that the myotomic rudiments of these muscles were excised in preparing the wound for the reception of the transplant. Lewis (710) has experi- mentally shown that the sternohyoid portion of the ventrolateral musculature in Amblystoma is derived from the ventral processes of the first three myotomes, and that, after extirpation of the {30 S. R. DETWILER first myotome, the anterior portion of this muscle is entirely lacking as is its derivative, the m. geniohyoideus. No mention was made of the genesis of the m. hyoglossus, but from the present experiments, the evidence suggests that this muscle also arises from the anterior segment of the m. sternohyoideus, for when the anterior segments of this muscle are lacking, the m. hyoglossus is also absent. The ventral portion of the first myotome in Amblystoma is in intimate relation to the gill mass (fig. 1). The deficiencies Ee 0 EG! Fig.9 Left lateral aspect of case AA2S»: drawn seven days after the operation. ~ 10. Fig. 10 Right lateral aspect of case AA2S», drawn seven days after the opera- tion. The limb bud occupies the region of the second and third external gills. The first gill (HG1) is considerably smaller than the normal (fig.9). > 10. in the hypobranchial musculature on the operated side indicate that in preparing the wound for the limb rudiment, the ventral portion of the first myotome, as well as a portion of the second, was excised. Case AA28,,. In this experiment, the ectoderm and mesoderm of the entire gill region were excised and the limb rudiment was transplanted into the excavated area, as seen under conditions indicated in figure 3. The denuded limb area was covered with indifferent ectoderm. Seven days after the operation the ap- pearance of an external gill was seen between the heterotopic limb bud and the balancer (fig. 10, ef. fig. 9). At fourteen days TRANSPLANTATION OF LIMBS IN AMBLYSTOMA fat the developing limb bud had rotated dorsally under the influence of ectoderm dorsal to the limb, simulating a condition shown in figure 6. It later assumed typical orientation and developed into an appendage which, although somewhat small, was ex- ternally normal in appearance (fig. 11). During later develop- ment, the limb gradually shifted caudally so that its final position was scarcely the distance of two segments anterior to the ortho- topic position. Movements of the limb on the shoulder were quite defective, owing to incomplete muscular differentiation and imperfect innervation (table 3). The musculature of the Fig.11 Ventral aspect of case AA2S.; drawn forty-eight days after the opera- tion. Only one small external gill has developed, and the limb, which earlier occupied the region of the second and third external gills (figs. 9 and 10) has migrated caudally to its final position. & 5. limb itself was well differentiated and received innervation from two nerve trunks (fig. 12), one on the flexor surface and the other on the extensor surface. The shoulder-girdle was typical in shape, although the ventral zone was much shorter than normal, with a correspondingly curtailed development of the coraco- brachial and the pectoral muscles. The limb and shoulder were innervated by the ventral divisions of the second and third spinal nerves (fig. 12). The distal portion of the latter, which contrib- uted the main bulk of the nerve supply to the limb and the shoulder had elongated a considerable distance anteriorly. The ventral ramus of the second spinal nerve passed at first caudally 132 Ss. R. DETWILER through the dorsal trunk musculature and then almost directly laterally. Before uniting with the third nerve it supplied several branches to the abdominohyoideus musculature. No union was found between the ventral rami of the first and second nerves. The former was found to pass somewhat caudally, then laterally, terminating in the posterior segments of the m. sternohyoideus. 13 Fig. 12 Graphic reconstruction of the segmental nerve contribution to trans- planted limb in case AA2S», (fig. 11). xX 25. Arrow designates level of trans- planted limb. Fig. 13 Graphic reconstruction of the segmental nerve supply to trans- planted limb in case AA2S.,. Limb, transplanted into the gill region (fig. 2), has migrated caudally during development, finally occupying a position approxi- mately one segment anterior to the normal position. 25. Arrow A indicates position of transplanted limb; arrow B designates normal limb level. A typical anterior prolongation of the hypoglossal nerve was lacking in this case, as were also the anterior segments of the m. sternohyoideus and their derivatives (m. geniohyoideus and m. genioglossus). Case AAIS.,. Although in this experiment the limb rudiment was transplanted into the gill region, as is indicated in figure 3, its final position was only one segment anterior to the normal, TRANSPLANTATION OF LIMBS IN AMBLYSTOMA 1383 and it is hence classified under series AAIS (table 2). Owing to the early caudal migration of the rudiment in this case, develop- ment proceeded normally. Although the limb occupied a posi- TABLE 2 Showing the segmental nerve contribution to the right fore limb when transplanted. A. One to three segments anterior to the normal position (series AA1S, AA2S). B. One to five segments caudal to the normal position (series AS1, AS2, etc.) POSITION OF LIMB SEGMENTAL NERVE CONTRIBUTION Number of|Number of SIRS eAnue segments | segments anterior to | caudal to 1 2 3 4 5 6 7 8 | 9 normal normal position position INormailzeneey: 1 3/415 fle 43 1 3/415 (NAN Sieh ae o4 1 ae ie 15 2, Peay || 2 21 2, PAN BS || Ze BOS. Wet | Oh 2/3 |4 73 2 Dita |) 78 23-3 Je AVS Teepe iy ak 12) 1 3] 45 5 2 Sy | SS 2 ee is : ae : 9 3 415/16 WSS odo ees & 5 Alcea B 12 3 Ab ly} CoN |Z ASA Ste stele 24 4 Ny Xa) || @ 30 4 WG || 7 25 5 Dil On| edie mena DO NSD es ae 27 . 5 Gales 9 30 5 (ie | NS | | 1 The figures in ‘B’ are taken from a previously published table (Detwiler, ’20). tion caudal to the gill region, gill formation was almost entirely suppressed, and only one small external gill developed, which, from its position, was taken to be the first. Ss. R. DETWILER 134 ‘uoryvorjdnpoer Jo coquiowt 1ol4e4ysod = ‘uoIyBoTTdnpoad JO JOquIeW IOT10yUe “UOTPVAIOUUT SYOV] OPOSNUT 4YvYY SOYIUSIS — {poyRAIOUUT SI VpOsnu yey} soytusts + +] JUsserg |+| Juesorg | ({)+ | Juosorg [+] JUeserg |—| yuosorg |+| quesarg |+ juosoig |—| jJueserg +] JuesoIg |-++| JUSSeIg | (4)+ | JUaserg |-+| JUeSeIg |+] yuOserg |+ yuosolg |-+| JUdSeTg |+] yuesorg +} JUeserg |+|JUeseIg | + | JUeSeIg |—| JUesSqy |—| yuesqy |—| yuUesqy |—| Jyuesqy |—| quesqy (qtoys +| queserg |+|juesoIq | 3 yuosaig |+) JUeseIq |—| JUosqY |—| JUesqy |+) Aro9A) |—| QUOSeIg (g10ys ({yeus 3 q104g |—| AeA) | ¢ yoy |—| JUesqy |—| Juesqy |—| guesqy |+] yuesoIg |Z Aid A) (jeus | ({yeurs +] quesarlg |Z A109 A) | — qyuosqyV |+| JUesoTq |—| ({[vaIg) |+]| yUeserg |+ | JuoSerig |—| A19,) +| }uesorg |+/ juesotg ) + quoselg || JUeseTg |-+| JUesSeIg |+| JUOSeIg |-+| Juoserg |+| yuasorg snsuoy ‘ StTPBrpout ay 7ndGas S Sera avalcok St[eiopood “PL se sonics 109 “JW Saea abt acres, “W Bieta W Prroneiard “W Serna a Sa TOSoOW yijddns a. quosol quosod quesqy quosatg quesolg quosolg quosolg + of -|- ++ sIper9tan Y -00B10001d * Jy G =i I 82] “StVV gL\"* V Go| “SZVV 1Z| “SCVV GT| “ScVV ¥Z| “STVV 2 a Bq | saras a cB aalau aayoadsat tay? fO adUasqn wo aduasaid oy? pun ‘quin pazunjdsunsy 947 YIN padozaaap YoY sajzasnw Lapynoys ay) Buinoys € HIaVL TRANSPLANTATION OF LIMBS IN AMBLYSTOMA 135 Movements of the limb were practically normal. The shoulder and limb muscles were completely differentiated and were well supplied by nerve fibers from the ventral rami of the second, third, and fourth spinal nerves (fig. 13). From this figure it is seen that although the course of the proximal portions of the nerves is very similar to that of the normal limb nerves (fig. 4), their distal portions are continued anteriorly for some distance to the heterotopic limb. The ventral ramus of the second nerve had no connection with that of the first and passed directly to the brachial plexus. The ventral ramus of the first nerve was distributed to the fragmentary m. sternohyoideus, the anterior portion of which was wanting, as were also the m. geniohyoideus and the m. hyoglossus. In comparing the muscular differentiation and innervation of this case with that of former and subsequent cases (table 3) in which the limb developed in the region of the gills, it is seen that the immediate organic environment of the limb distinctly favors more normal development. Cases in which the transplanted rudi- ment remained in the gill region and there underwent differentia- tion showed considerable deficiencies in the development of the shoulder muscles as well as defective peripheral nerve connec- tions (cases AA2Si;, 21, 25, etc., table 3). The large percentage of defective limbs is likewise shown in table 1 B. The above shows that even though the limb constitutes an equipotential, self-differentiating system, the character of its development is markedly affected by developmental potencies of the region into which it is transplanted. Although the transplanted limb rudi- ment possesses complete intrinsic potentialities to develop into a normal appendage, it is clearly evident in these cases that the high percentage of abnormalities in the gill region is due to an inhibitory influence resulting from the more potent inherent gill-producing properties of the tissues in the immediate organic environment of the gills which offer a very unfavorable environ- ment for normal limb differentiation. The abnormalities on the part of both gills and limb clearly show developmental conflicts between the two systems. 136 S. R. DETWILER The general results of these experiments indicate that when the major portion of the gill tissue is removed and replaced by an entire limb rudiment, the remaining unremoved portion of the gill tissue, being in its normal organic environment, possesses a relatively greater potency to produce normal gills than does the entire limb system to produce a normal appendage—the latter being in the heterotopic position. When a limb is transplanted into a more passive region of the embryo such as that caudal to its normal position, practically normal differentiation results. For example, limbs reimplanted the distance of three body seg- ments caudal to the normal location yielded 75 per cent normali- ties (Detwiler, ’20). When transplanted the same distance anterior to the normal site, into the active self-differentiating gill region, less than 15 per cent of the cases developed normally (table 1). Case AA28,;. In this experiment (fig. 14) the initial develop- ment of the limb was normal. Only one rudimentary gill de- veloped, but the ectoderm in the vicinity of the base of the limb wandered out over its flexor surface forming a permanent ridge (figs. 14 and 15). The caudal border of the proximal portion of the limb was also fused with the ectoderm of the gill, which greatly restricted its motility. Function of the forearm and hand were practically normal. | Serial transverse sections showed that the glenoid cavity was approximately two and one-half segments anterior to the nor- mal position. The shoulder-girdle was abnormal. Its coracoid portion was continuous with the fragmentary coracoid which developed in the orthotopic position from unremoved portions of its rudiment. The suprascapular portion was short and quite thick. The shoulder muscles were poorly developed and were some- what defective in nerve supply (table 3), which was derived from the ventral rami of the second, third, and fourth spinal nerves (fig. 16). Although there was an abundant nerve supply from the cord, a survey of the sections shows clearly that the incom- pleteness of function in this case is due to the abnormal develop- ment of the shoulder-girdle, deficiency of its corresponding mus- culature, and the fusion of its proximal portion with the gill. TRANSPLANTATION OF LIMBS IN AMBLYSTOMA sve The anterior elongation of the distal portions of the limb nerves is seen in figure 16. It is interesting to note that the fourth nerve did not unite with the third until the periphery had been reached by an extended anterolateral growth. The ventral ramus’ of the second spinal nerve showed no connection with rh ene 2s) MGORBR LG 6: 14 15 Fig. 14 Ventral view of case AA28,;. Limb transplanted into the gill region as indicated in figure 3. Final position of limb, two and one-half body segments anterior to normal level. Animal preserved fifty-one days after the operation. x os Fig. 15 Transverse section of right brachium of case AA2S:5, at the level A-A, figure 14, showing fusion of abnormal gill with radial side of the appendage. < 58. No vascular elements have developed in the anomalous gill. H, humerus; M TR BR, m. triceps brachii; M COR BR LG., m. coracobrachialis longus; MB BR, m. biceps brachii; PNT, flexor nerve trunk; ENT, extensor nerve trunk; EG, ectoderm of gill. that of the first, and its entire trunk was contributed to the plexus. The hypoglossal nerve was formed solely of the first spinal nerve which, after supplying branches to the somewhat reduced m. sterrtohyoideus, continued anteriorly to end in the m. genio- hyoideus and the m. hyoglossus. 138 Ss. R. DETWILER Case AA2S,;. In this experiment, which was carried out under conditions indicated in table 1 B, the development of external gills was completely suppressed and two limbs of the same laterality developed in the heterotopic position (fig. 17). From external observations on the developing larva it was impossible to determine whether both members of the double limb developed 16 17 Fig. 16 Graphic reconstruction of segmental nerve supply to transplanted right anterior limb in case AA2S8,; (fig. 14). Limb occupies a position approxi- mately two and one-half segments anterior to the normal. Arrow A indicates level of transplanted limb; arrow B designates normal limb level. X 25. Fig. 17 Dorsal view of case AA2S;73, showing absence of gills and presence of two limbs of same laterality developed from anomalous reduplication of the transplanted rudiment. Experiment carried out under conditions indicated in table 1B Gig. 3). xX 5. from the transplant or whether the caudal member developed from an unremoved portion of the original rudiment. The limb bud which developed into the anterior appendage was the first to appear. This was soon followed by a second bud which de- veloped into the caudal member. The fact that the final posi- tion of the latter was approximately two segments anterior to the normal position suggested that it developed from the transplanted rudiment. Moreover, examinations of sections showed that its TRANSPLANTATION OF LIMBS IN AMBLYSTOMA 139 shoulder-girdle was developed anterior to the isolated distal portions of the intact suprascapula and coracoid, which always develop in situ after a typical limb-bud excision. Further, the fact that the anterior member lacked a girdle, except for a very fragmentary coracoid, also indicated that the well- developed girdle of the caudal member was developed from the girdle rudiment included in the transplant. Previous limb experiments on Amblystoma have shown that when supernumerary limbs develop from a single transplant under conditions of normal orientation with respect to the sur- rounding tissue, the one is always the mirror image of the other (Harrison, 717, ’21; Detwiler, ’18), i.e., a disharmonic appendage. In the light of these results, the anomalous reduplication in the present experiment, if resulting from a single transplant as the evidence favors, is an unusual phenomenon. Harrison (’21), however, describes several cases with anomalous double-limb formation, both of the same laterality, resulting from a composite rudiment in the orthotopic position. The two limbs in these cases are shown in Harrison’s paper (l.c.), 79, figs. 131 and 132. The combinations producing them are shown on page 70 of his paper (combinations 10 and 11). In the case under consideration, the development of a well- formed branchial bar between the bases of the two limbs might have effected cleavage of the original transplanted rudiment, resulting thereby in the development of the two limbs. It has been shown by Harrison (718) that, under normal conditions, splitting the limb bud by a vertical or horizontal incision never produces reduplication. There is no apparent reason, however why two portions of a transplanted limb rudiment, if permanently separated, should not develop into two limbs of the same lateral- ity. Such a result is normally produced when a portion of the rudiment is transplanted: one limb will develop at the original site and the other at the heterotopic position. In the present case the anterior limb developed in the posterior - region of the otic capsule. The shoulder-girdle consisted of only a fragmentary coracoid. This was fused dorsally with the cartilaginous capsule of the ear, which, together with the small 140 S. R. DETWILER coracoid, was molded into a typical shoulder-joint (fig. 30). This figure represents a transverse section at the level A—A in figure 17. A muscle mass connected the coracoid with the hu- merus, but theindividuality of themuscles could not bedetermined. The extensor musculature of the limb was well developed and supplied by a small nerve trunk. The flexor musculature was sparse, but functional. A nerve trunk of considerable size was also found on this surface of the appendage. Considerable Fig. 18 Graphic reconstruction of the segmental nerve supply to the anterior and posterior members of anomalous limb reduplication in case AA2S73. X 25. Arrow A indicates position of anterior member; arrow B of posterior member; arrow C designates normal limb level. movements in the forearm and hand were observed. The nerve contribution to this appendage was derived from the main por- tion of the ventral ramus of the second spinal nerve, which had elongated anteriorly a remarkable distance to effect this connec- tion (fig. 18). The base of the caudal limb was situated ap- proximately the distance of two body segments anterior to the normal position. The shoulder-girdle was well formed (fig. 31). Its extreme dorsal portion (suprascapula) was connected with the fragmentary coracoid of the anterior limb by means of a bar of cartilage which was taken to be a modified branchial bar TRANSPLANTATION OF LIMBS IN AMBLYSTOMA 141 (fig. 31). The shoulder muscles were typically developed and were supplied by nerves (table 3). The muscles within the limb were normally differentiated and were supplied with nerves of typical intrinsic distribution. The segmental nerve supply to this limb comprised the ventral rami of the second, third, and fourth nerves (fig. 18). The greater part of the nerve supply was derived from the latter two, although a good-sized branch from the second contributed to the plexus. The second nerve, in addition to supplying the anterior limb and a contributing Fig. 19 Dorsal view of case AA2S7s, showing absence of gills and the position of limb developed from rudiment transplanted into gill region (fig. 3). Animal preserved fifty-nine days after the operation. X 5. trunk to the posterior appendage, supplied fibers to the m. abdominohyoideus. No communication with the first spinal nerve could be found. In this case the hypoglossal trunk was formed solely from the latter, which exhibited a normal course along the dorsolateral border of the m. sternohyoideus in which it finally terminated. The m. geniohyoideus and the m. hyo- glossus were absent on the operated side. Case AA2S;;. In this experiment (tables 1 B and 2 A), gill development was also entirely suppressed and the transplanted rudiment developed into a normal appendage (fig. 19). Its 142 S. R. DETWILER location was somewhat dorsal and approximately the distance of two and one-half segments anterior to the normal site. Ex- amination of sections revealed a well-developed shoulder-girdle. The shoulder muscles were normally present, but within complete nerve supply (table 3), which placed considerable restrictions on the shoulder movements. The muscles of the limb were typically developed. Only one nerve trunk entered the limb, however, and this was situated on the flexor surface. The extensor paraly- sis which this limb showed and the absence of an extensor nerve trunk have been observed in a considerable number of trans- Ue Fig. 20 Graphic reconstruction of segmental nerves supplying transplanted right anterior limb in case AA2S7s (fig. 19). X 25. Arrow indicates position of limb and level of section shown in figure 32. j planted limbs. The reasons for this specific deficiency in the presence of muscular differentiation are not yet clear. The nerve supply to the shoulder and limb was derived from the ventral rami of the first, second, and third spinal nerves. The pathway of these nerves in their contribution to the limb plexus is seen in figure 20. The entire ventral rami of the first and second spinal nerve were contributed to the brachial plexus. The ventrolateral musculature, which was sparse in this case, lacked innervation. A small m. sternohyoideus was present, but the m. geniohyoideus and the m. hyoglossus were entirely lacking. Figure 32 shows the internal configuration of a trans- verse section at the level A—A, figure 19. The ventral ramus TRANSPLANTATION OF LIMBS IN AMBLYSTOMA 143 of the first spinal nerve is seen in cross-section. The distal nerve trunk in the vicinity of the limb region is indicated by the line a—a in figure 20. This nerve trunk represents a fusion of the distal portions of the ventral rami of the first, second, and third spinal nerves, all of which have grown considerable distances anteriorly to reach the transplanted limb. Two cases in which normal limbs developed, but which have not been sectioned, are shown in figures 21 and 22. In case 2] 22 Fig. 21 Dorsal view of case AA2S;0, showing presence of normal functionless limb, developed from rudiment transplanted into gill region (fig. 2). Animal preserved forty-four days after the operation. X 5. Fig. 22 Dorsal view of case AA2S;2, showing transplanted and regenerated limbs. The base of the former is fused with a strip of gill ectoderm which has grown out along its radial border. Incomplete reduplication of the hand is seen in the regenerated appendage. Animal preserved forty-four days after the operation. x 5. AA28;0 (fig. 21) the limb, although normal in appearance, was practically devoid of function. One small gill developed which was situated farther ventral than normal. In case AA28;. (fig. 22) the transplanted limb was quite normal in appearance. A fold of ectoderm extended out over its base and was fused along the radial side, thus binding the limb close to the side of the body. Just posterior to the transplanted limb there re- generated another limb with reduplication of the hand. Gill development was entirely suppressed. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 35, NO. 2 144 Ss. R. DETWILER Histories of characteristic anomalous cases Inasmuch as quite a large percentage of the experiments resulted in gill and limb abnormalities of similar character, the record of only a few such typical cases are appended. Case AA2S8o7 (figs. 23 and 24). March 3, 1920. Operation. Rt. limb transplanted into gill region. Conditions as described in table 1 B. Wound covered. Fig. 23 Drawing of left lateral aspect of case AA2S27. Animal preserved twenty days after the operation. X 10. Fig. 24 Drawing of right lateral aspect of case AA2Se7 (ef. fig. 23). The abnormal limb (L) has developed from rudiment transplanted into the gill region (fig. 2). A club-shaped external gill (@) has developed dorsal to the appendage and has fused with its lateral side. Proximal portion of arm is fused to gill ectoderm and points dorsocaudally. Two small gills are seen ventral to the limb. X 10. March 29. Completely healed. April 2. Transplanted limb bud points caudally, as in figure 10. April 7. Two small gills developing ventral and slightly anterior to limb rudiment. April 12. Limb points dorsally and is closely applied to the ecto- derm of the gill region, as in figure 6. April 16. Distal portion of the limb has rotated laterally and is directed caudally. A flange of ectoderm dorsal to the limb has extended out along the dorsal border of the arm. Two abortive gills situated ventral to the limb. - = TRANSPLANTATION OF LIMBS IN AMBLYSTOMA 145 April 18. Proximal portion of the limb fused to the body wall. Distal portion is free and points caudally. The ectoderm dorsal to the limb has developed into an abortive gill which is fused to the lateral side of the arm (fig. 24). Case AA2Sz;. April 3, 1920. Operation. Conditions of experiment as in AA2S,;. Wound covered. April 8. Limb bud developing with normal orientation. April 12. Developing limb bud points dorsally and is fused to the side of the body. Gill ectoderm extends out along the dorsum of the Pen Fig. 25 Drawing of left lateral aspect of case AA2Ss,. Animal preserved eighteen days after the operation. X 10. Fig. 26 Drawing of right lateral aspect of case AA2Ss4 (fig. 25). Anterior limb transplanted into the gill region as described in table 1 B. The limb (6) is abortive and is fused dorsally with an extension of the gill ectoderm which has migrated out over the extremity. A small atrophic gill has developed ventral to the limb. X 10. free distal portion.” One small gill developing anterior and ventral to the limb. April 18. Gill ectoderm has completely fused along the dorsum of the limb. Limb very abnormal and reduced to a club-shaped appendage. Small atrophic gill ventral to the limb. April 22. Animal preserved (fig. 26, ef. fig. 25). Case AA2S;;. April 3. Operation. Experimental conditions as in case AA2S.7. Wound covered. April 4. Completely healed. April 9. Limb bud developing with normal orientation. 146 Ss. R. DETWILER April 12. Two small gills developing anterior to limb. Limb abor- tive and is rotated dorsally. April 20. Posterior (second) gill fused along anterior border of atrophic limb. First gill very small. April 29. Animal preserved (fig. 27). Section of the above case shows the presence of several abortive and abnormally developed branchial bars. There is a small limb girdle, the glenoid cavity of which is deeply situated. Only the distal portion of the arm is free from the body. The appendage lacks muscular differentiation. The ectoderm of the fused external gill (fig. 27) is Fig. 27 Dorsal view of case AA2S;5. Right anterior limb transplanted into gill region as under conditions described in table 1 B. The limb (LZ) is abortive. Two abortive gills are seen anterior to the limb. The second (£G@2) is fused along the radial border of the limb. Animal preserved twenty-six days after the operation. X 10. continuous with that covering the extremity. No vascular elements are differentiated. Case AA2Sgs. April 4, 1920. Operation. Experimental conditions as in case AA28o7. April 5. Completely healed. April 10. Limb bud developing with normal orientation. No evidence of developing gills. April 15. Limb points dorsally and the proximal part is fused to side of body wall. Gill ectoderm fuses along dorsum of the limb. Very small, abortive gills developing ventral to the typical gill region. April 20. Distal portion of limb club-shaped. Digits fail to develop. Gill ectoderm is connected with limb along its dorsomesial border. Small spur developed on lateral border of arm. TRANSPLANTATION OF LIMBS IN AMBLYSTOMA 147 April 25. Animal preserved (fig. 29, cf. fig. 28). Very abnormal, functionless limb which is fused dorsally with an extension of gill ecto- derm. Three abortive gills situated ventral to the gill shelf. The gill-producing potentiality of the outlying ectoderm is well illustrated in the above cases. In all of these experiments the entire ectodermal and mesodermal tissues of the gill swellings were removed prior to implantation of limb, yet one or more gills MN — 28 cas) EG Fig. 28 Drawing of left lateral aspect of case AA2S¢,. Animal preserved thirty-six days after operation. X 7. Fig. 29 Drawing of right lateral aspect of case AA2Ses (fig. 28). Fore-limb rudiment transplanted into gill region as described in table 1 B. Gill ectoderm dorsal to abortive limb (Z) has migrated out on the limb and is fused along its dorsomesial surface. Three very small external gills (HG) have p sevelaned out of the tissue ventral to the normal gill region. X 7. developed in all four cases. From the results of these cases, as well as from others not reported, the evidence indicates that the ectoderm lying ventral to the typical gill region possesses a rela- tively stronger gill-forming capacity than does that lying an- terior and dorsal. In the majority of cases the gills developed ventral to the gill region in approximation to the general heart area as illustrated in figures 24, 26, and 29. This observation supports the results obtained by Ekman (713), p. 578, vide supra, p. 125). ‘snoprof£youie4ys ‘wi ‘qf H LS WW ‘eAdieu jessojsodcéy ‘dXAH N ‘teq [eryoueiq ‘gg feAreu quit ‘ArT fepnsdeo 01440 snourlsepyae9 ‘909 {ploovs09 ‘QQ {AqIAv plousls ‘95 ‘sniowny ‘YT "FZ XK “LT omnsy ‘y-y [eae] oy} ye “QzyV osvo Jo (FEE ‘OU) UOTZ00S OSIOASUBI} JO SUIMVIG CF “SI DETWILER Ss. IRAs 148 ADO ONI CED, Ler EN —— 2 ‘a we py 4 4 ees milk.——— Malted milk. 1 wk. — 2 wks. 2 wks. 1% ethyl alcohols 4% Seat alcohol. 33% alcohol. 3 1 wk. : 2 wk 2 wks 4%, Beng “alcohol ih slcoheld 25 2 wKS.e 2 wks. 9,10" wks. 11” wk. 12,13" wks. 14" wk.15,16" wks. 17" wk. 18,19" wks, 20" wk > Malted milk. > Malted milk. ————> Malted milk. ————> Malted milk. 2 wks. 2 wks. Bess 2 wks. alcohol s #% alcohol. 2%, alcohol. ————> si ee 5 2 wks. 2 wks. 2 wks. 4%, alcohol. ———-3 3% alcohol. ee ae Ono. ===> 6. aloohol. ey 2 wks. 2 wks. SY 2 wks. wks x 1% alcohol. , .2 Wks. 2% alcohol. 2 wks. 21,22" wks. 235" "wks 24" wk. 25" wke 26" wk. ee Wise Malted milk. ——> Malted milk. — Malted milk. 2 wks. 1 generation. 1 generation. $% alcohol. ————}Malted milk. ————_> Malted milk. 2 wxs. 1 generation, 1 generation. 24, alcohol. + Malted milk. ————> Malted milk. 2 WKS. 1 generation. 1 generation. seemed to warrant. At the end of this time (five weeks) 100 individuals from each line were isolated and the egg production and length of life determined. In the 4 per cent alcohol the range of egg production was from 4 to 22, with an average of 14,25; the range of life from 3 to 7 days, with an average length of life of 5.83 days in the 4 per cent alcohol the range in egg pro- duction was from 1 to 10, with an average of 3.98; the range of of life from 4 to 7 days, with an average of 6.09 days; in the controls the egg production ranged from 14 to 30, with an average of 14.10 eggs; the length of life from 3 to 7 days, with an average LIFE-HISTORY OF ROTIFER 249 of 6.17 days. When these percentages of alcohol are employed, the reduction in the number of eggs deposited is not so marked as in higher percentages, but the difference between the alcohol lines and the line reared under normal conditions is still great. There is little reduction in the average length of life in either percentage of alcohol, just as was the case where 1 per cent and 2 per cent were used. The alcohol experiment was carried on for twenty-seven weeks in the way just indicated; that is, 100 individuals each subjected to ¢ and } per cent alcohol solutions, and 100 controls, were allowed to reproduce for a period of two weeks, when isolation of 100 specimens was made from each line and the egg deposit and length of life for all the individuals determined under each ‘of the three conditions; reproduction continued for another two-week period, then another isolation made, ete., until the end of the twenty-third week. At this time the alcohol cultures were discontinued and progeny from both alcohol lines were returned to malted-milk solution without alcohol. In this way it was determined whether the effects of + and 4 per cent alcohol, which had been acting continuously on the progenitors of both lines for twenty-one weeks, had any lasting effect when the progeny were returned to malted milk. At the beginning of the fifteenth week progeny of the line reared in 4 per cent alcohol were isolated and reared in 1 per cent and 14 per cent alcohol for a period of two weeks and an isolation made at the end of this time to determine if under the continuous action of 4 per cent alcohol the organism had developed any degree of resistance to higher percentages. The effect of the alcohol upon the organisms themselves was marked only in the higher percentages. In many cases in these percentages the increase in size, which usually continues until near the close of the life-cycle, never took place and the indi- viduals were thin and attenuated; movement in many cases was reduced even in the young individual, and the adults became very sluggish. In the experiments where a low percentage of alcohol (4, = per cent) was employed, the treated organisms resembled the untreated checks except in the number of eggs deposited. 250 BESSIE NOYES An isolation made at the end of the tenth week showed that in the + per cent alcohol line the range of egg production was from 1 to 19, with an average of 11.07. This is slightly lower than the results in the fifth week, where the range was from 4 to 22, with an average of 11.25. In the eleventh week the range of life was from 3 to 8 days, with an average of 6.46. In the fifth week the range of life had been from 3 to 7 days, with an average of 5.83 days. In the eleventh week in } per cent alcohol the average and maximum production of eggs had fallen below that in the fifth week, but the maximum length of life at this time was the : greatest attained by any individual throughout the entire study. A greater length of life under treatment with alcohol was noted by Stockard in one of his guinea-pigs, which lived seven years—an unusual time in that organism. Individuals subjected to $ per cent alcohol showed at this time a range in the egg deposit of 1 to 10, with an average of 5; this was the same as the range in the fifth week, but at that time the average was only 3.98. The range of life for this period was from 3 to 7 days, with an average of 5.70 days, as compared with a range of 4 to 7 days and an average of 6.09 days at five weeks. In malted milk under normal conditions the range of egg production at the eleventh week was 8 to 29, with an average of 19.70, while the range in life was 3 to 7 days, with an average of 6.48 days. At this isolation, just as in the previous one, there is a decided reduction in the number of eggs produced in both the alcohol lines and very little reduction in the average life-period. At the beginning of the fifteenth week two new isolations of 100 individuals each were made from the 4 per cent alcohol line; one of these was reared in 1 per cent alcohol; the other in 13 per cent. At the beginning of the seventeenth week isola- tions were made for the determination of the egg production and the length of life from the five lines, malted milk, + per cent, 3 per cent, 1 per cent, 14 per cent alcohol. At this time the line in ; per cent alcohol showed a range in egg production of 4 to 22, with an average of 10.80; a range in the length of life of 3 to 7 days, with an average of 5.88 days. The line in 3 per cent alcohol showed an egg production ranging from 1 to 10, with an LIFE-HISTORY OF ROTIFER 251 average of 4.55, and a range of length of life of 2 to 7 days, with an average of 6.52 days. The line previously reared in } per cent alcohol and transferred to 1 per cent only two weeks previous to this isolation showed a range in egg production from 1 to 7, with an average of 2.98, a range in length of life from 3 to 7 days, with an average of 5.22 days. The line recently transferred to 14 per cent alcohol showed a range in egg production of 1 to 6, with an average of 2, and a range in length of life of 2 to 7 days, with an average of 4.91 days. The line reared continuously in malted milk under normal culture conditions showed a range for egg production of 1 to 28, with an average of 19.26, and a range in the length of life of 3 to 7 days, with an average of 5.81 days. In general during the twenty-one weeks in which individuals of a line of Proales were subjected continuously to the fumes of 1 and 3 per cent ethy! alcohol the maximum and average number of eggs produced was reduced in proportion to the percentage of alcohol used, but the average length of life was very little in- fluenced. Table 5 gives a summary of the range and average egg production and the range and average length of life of all the individuals studied throughout the experiment. In neither of the lines subjected to + and 3 per cent alcohol is there a con- tinual decrease in the average of either character studied through- out the successive generation. Tests for inheritance of the effect of alcohol. At the beginning of the twenty-fourth week both alcohol lines were discontinued and their progeny returned to malted milk only. At the end of the first week in normal conditions (twenty-fifth week of the experiment) isolations of the second generation individuals were made for determining the egg production and average length of life. In the line descended from the progeny of the { per cent alcohol group the range of egg production was from 9 to 27, with an average of 17.77, as compared with a range of 3 to 23, with an average of 12.33 eggs in the last generation in } per cent alcohol. The range of life at this time was 3 to 7 days, with an average of 5.58 days, differing little from the range and average in the last isolation in + per cent alcohol. In the line descended from individuals reared in 3 per cent alcohol the range in egg produc- . ISOLATION WAS 252 BESSIE NOYES tion was from 5 to 25, with an average of 13.57, as compared with a range of | to 11 and an average of 3.48 in the last genera- tion in alcohol. The average length of life, 5.70 days, had increased very little over the average of 5.62 days in the last isolation. Individuals reared continually in malted milk showed at this time a range in egg production from 4 to 29, with an average of 18.56, and a range in length of life from 2 to 7 days, with an average of 5.56 (table 4). TABLE 5 A comparison of the egg production and length of life of all isolations made of lines subjected to { per cent and } per cent alcohol and controls LINE IN } PER CENT ALCOHOL|LINE IN } PER CENT ALCOHOL CONTROLS NUMBER OF || Oe WEEK Egg . Length of Egg Length of Egg. Length of AT WHICH production life production life production life MADE a aa a a a Aver- Aver- Aver- Aver- Aver- Aver- Range age Range ase Range age Range age Range Range age age 4—22|14. 25] 3-7 | 5.83) 1-10] 3.98) 4-7 | 6.09] 4-30/24.10) 3-7 | 6.17 3-20/12.63) 2-7 | 5.60} 1-10} 4.44) 2-6 | 5.08) 4-29/20.95) 2-7 | 5.16 1-19]11.07} 3-8 | 6.46} 1-10] 5.00) 3-7 | 5.70) 8—29/19.70} 3-7 | 6.48 4—22)12.67| 3-7 | 5.10] 1-11) 4.99} 3-6 | 5.16] 5-28]15.84| 3-7 | 5.99 4-22)10.80} 3-7 | 5.88} 1-10} 4.55) 2-7 | 5.52) 4-28/19.26) 3-7 | 5.81 4-23/14.73] 3-7 | 5.95} 1-11] 4.39) 2-7 | 5.45) 5-28/20.59) 3-7 | 5.80 3-23/12.33] 2-7 | 5.95} 1-11) 3.48] 2-7 | 5.62) 5-28/18.88) 2-7 | 4.75 9-27/17.77| 3-7 | 5.58] 5.25/13.57| 2-7 | 5.70} 4-29/18.56) 2-7 | 5.56 1230/21. 80) 3-7 Dates 8—29}19.92) 3-7 5. 37/11-30/19. 50 4-7 | 6.06 * Indicates time when alcohol cultures were transferred to malted milk. Thus, individuals of the second generation after the return to normal conditions showed a marked increase in egg production; in other words, only a partial retention of the influence of the aleohol; and the averages for the generation whose ancestors were subjected to alcohol approach those individuals continually reared in malted milk. Isolations of the third generation made under the same condi- tions as have just been described at the beginning of the twenty- seventh week, show an average egg production of 21.80 for de- scendants of individuals subjected to } per cent alcohol, and 19.92 for those subjected to 4 per cent, as compared with an average LIFE-HISTORY OF ROTIFER 293 of 19.50 for checks reared continuously in malted milk. At this time, after three generations spent in malted milk, all the effects of alcohol upon egg production has been lost. To summarize: In Proales decipiens individuals of a line subjected to the fumes of } and } per cent ethyl alcohol con- tinuously for nineteen weeks show a decided reduction in egg production while under the influence of the alcohol, but their progeny, returned to normal conditions, regain the normal egg-producing power after the third generation. SUMMARY This paper is an account of the normal life-cycle of Proales decipiens, with experimental studies of the production of males, of the effects of selection during parthenogenetic reproduction, and of the effects of alcohol on inherited characteristics. 1. Statistics are given as to the length of life, the rate of reproduction, the number and kind of eggs deposited, with study of the variations in these matters. The animal lives about a week, then dies with characteristic symptoms of senility. During its life it produces several eggs per day, the number increasing to a maximum, then decreasing with the onset of old age. 2. Reproduction by parthenogenesis for about 250 generations gave no indication of reduction of vigor in the race in any respect. 3. During this period, no males appeared. Alteration of the environment by changes in the nature and concentration of the food, by changes in the temperature at which cultures were reared, and changes in the chemical constitution of the medium were not accompanied by the appearance of the male form. So far as known, the species may be quite without males. 4. An attempt to increase the egg deposit and average length of life through artificial selection carried on for three months, in fifteen generations, was without avail, placing this organism in the list with other parthenogenetic forms in which selection is ineffective. 5. Treatment with ethyl alcohol in a concentration of + and $ per cent for twenty weeks reduced the number of eggs produced from an average of 15 to 24 in normal malted milk to an average 254. BESSIE NOYES of 10 to 14 in } per cent alcohol, and 3 to 5 in 3 per cent alcohol, although the length of life was little influenced. 6. The reduction in egg deposit brought about by alcohol was not retained beyond the third generation of descendants restored to normal conditions of culture. LITERATURE CITED Acar, W. E. 1914 Experiments on inheritance in parthenogenesis. Phil. Trans. Roy. Soc. London B, vol. 205, pp. 421-489. Esrarsrook, A. H. 1910 Effect of certain chemicals on growth in Paramecium. Jour. Exp. Zo6l., vol. 8, pp. 489-534. Ewine, H. E. 1914a Pure line inheritance and parthenogenesis. Biol. Bull., vol. 27, pp. 164-168. 1914b Notes on regression in a pure line of plant lice. Biol. Bull., vol. 27, pp. 164-168. 1916 Eighty-seven generations in a parthenogenetic pure line of Aphis avenae Fab. Biol. Bull., vol. 31, pp. 53-112. Heener, R. W. 1920 The relations between number, chromatin mass, cyto- plasmic mass, and shell characteristics in four species of the genus Arcella. Jour. Exp. Zo6l., vol. 30, pp. 1-95. Jennines, H. 8. 1916 Heredity, variation and the results of selection in the uniparental reproduction of Difflugia corona. Genetics, vol. 1, pp. 407-534. JoHANNSEN, W. 1903 Ueber Erblichkeit in Populationen und in reinen Linien. Ketiy, J. P. 1913 Heredity in a parthenogenetic insect (Aphis). Amer. Nat., vol. 47, pp. 229-234. MippteTon, A. R. 1915 Heritable variations and the results of selection in the fission rate of Stylonychia pustulata. Jour. Exp. Zodl., vol. 19, pp. 451-503. Peart, RaAyMonD 19174 The experimental modification of germ cells. I. Gen- eral plan of experiments with ethyl alcohol and certain related sub- stances. Jour. Exp. Zodél., vol. 22, pp. 125-165. 1917 b The-.experimental modification of germ cells. II. The effect upon the domestic fowl of the daily inhalation of ethyl alcohol and certain related substances. Jour. Exp. Zo6l., vol. 22, pp. 165-187. 1917 c The experimental modification of germ cells. III. The effect of parental alcoholism, and certain other drug intoxications, upon progeny. Jour. Exp. Zodl., vol. 22, pp. 241-311. Pearson, Kart 1910 Darwinism, biometry and some recent biology. I. . Biometrika, vol. 7, pp. 368-385. Root, F. M. 1917 Inheritance in the asexual reproduction of Centropyxis aculeata. Genetics, vol. 3, pp. 174-206. RovussEtet, C. F. 1897 On the male of Rhinops vitrea. Jour. Roy. Micr. Soc., 1897, pp. 4-8. 1897 On the male of Proales werneckii. Jour. Quekett Micr. Club, ser. 2, vol. 2, no. 41, pp. 415-418. LIFE-HISTORY OF ROTIFER 200 Rovssetet, C. F. 1903 Liste der bis jetzt bekannt gewordenen miannlichen Radertiere. Forschungsbericht aus der Biolog. Station zu Plon. Bd. 10, S. 172-176. Suutu, A. F. 1912 Studies in the life cycle of Hydatina senta. Jour. Exp. Zo6l., vol. 11-12, pp. 283-319. SrockarD, CHarLes R. 1918 Further studies on the modification of germ- cells in mammals; the effect of alcohol on treated guinea-pigs and their descendants. Jour. Exp. Zodl., vol. 26, pp. 119-226. Wuaitney, D. D. 1912 The effects of alcohol not inherited in Hydatina senta. Amer. Nat., vol. 46, pp. 41-56. 1917 The influence of food in controlling sex in Hydatina senta. Jour. Exp. Zoél., vol. 17, pp. 545-558. cle Age ate = AUTHOR’S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, MARCH 20 THE REACTIONS OF AMBLYSTOMA TIGRINUM TO OLFACTORY STIMULI J. S. NICHOLAS Osborn Zoélogical Laboratory, Yale University, New Haven, Connecticut ONE FIGURE INTRODUCTION The existence of the sense of smell in fishes has been clearly shown by Parker, Sheldon, and others. This work has settled the question for this group in accordance with Sherrington’s definition, which places the sense of smell in the exteroceptive system and that of taste in the interoceptive system. Comparatively little work of an exact nature has been done upon the olfactory sense of amphibia. Reese (’12) tested Diemyc- tilus with food and meat juices, but as he failed to control the sense of sight, his results are not fully conclusive. - Copeland (13) repeated the experiments of Reese, using more exact methods. He controlled the visual sense by stimu- lating the olfactory epithelium with diffusion from a motionless source, and he concludes that there is a definite olfactory sense in this animal. Risser (714), working upon larvae and adults of the frog and the toad, showed that tadpoles of the toad possess an olfac- tory sense, while frog tadpoles do not. In the adult toad, he was unable to obtain satisfactory data in regard to the exist- ence of an olfactory sense except in regard to specific odors. He did not, however, obtain any response to food substances with which the animals normally come in contact. Burr (716) obtained definite olfactory responses in larval Amblystoma punctatum. An investigation was begun in the spring of 1919 in an endeavor to study the correlation of the senses of sight and smell after 1 Anat. Rec., vol. 20, 1921, p. 189. 257 258 J. S. NICHOLAS extirpation of the sense organs in the larvae. Unfortunately, all of the animals so treated failed to survive. The present work was started in the fall of 1919 in order to get control results for the data which it is hoped can later be obtained upon adult forms which have been operated upon in early stages of development. I take pleasure in thanking Dr. Henry Laurens, under whose direction the work has been carried on, for many helpful sugges- tions and much useful criticism. The general observations upon the behavior of the operated larvae here described were made by Doctor Laurens, who performed the necessary operations in the spring of 1917. MATERIAL AND METHODS The same general method of approach was used as in Burr’s (16) work, 1.e., removing the embryonic anlage of the sense organs. Since the method of operation has been described in earlier papers (Laurens, 714, and Burr, 716), no description will be necessary here. The operated larvae were kept in battery jars stocked with the small crustacea, Daphnia and Cypris, and were allowed to grow to a length of about 25 mm. before any of the observations were made. The adult animals, which were used for the main portion of the work, were three individuals selected from a number that had been sent to Doctor Laurens from Albuquerque, New Mexico, in the spring of 1919. The same individuals were used under all the different conditions described in the experimental section, in order that the general control of the reactions might be more accurate and a more reliable average of reaction obtained. The methods applied to these forms were mainly those which have been used by Parker in his work on Fundulus and Ameiurus, with an application of some of Risser’s methods of experimenta- tion. In all the experiments the test aquarium was cleansed after each operation and filled with fresh water before continuing the tests. 2 The larvae on which the general observations were made were reared from eggs sent to Doctor Laurens by Prof. C. P. Sigerfoos, in the spring of 1917. Dr. W. C. Allee also was kind enough to send the author the eggs of A. tigrinum, but unfortunately, owing to the vicissitudes of the journey and the shock of sub- sequent operation, these animals did not survive. REACTIONS OF A. TIGRINUM 259 NORMAL BEHAVIOR The larvae are in almost incessant motion under laboratory conditions, swimming about the aquaria apparently in a con- stant search for food. Only after they have become gorged with food will they sink to the bottom, and even then they are seldom motionless, but continue to move about by combined walking and swimming movements, stopping to examine any object that attracts their attention. The adults used in this work had lived under laboratory con- ditions for three months previous to the time of the experimental work. The movements of the adult animals are ordinarily not particularly rapid, but the application of stimuli causes a re- markable acceleration of reaction. If placed in an aquarium provided with a float and left undisturbed, the animals will generally be found perfectly motionless upon the float, or else partly submerged with the head protruding above the surface. When placed in an aquarium filled with water to a depth of 6 or 7 inches, but without a float, the animals are restless for a time after their introduction, due probably to the handling which they have received. They soon come to rest, however, with only an occasional sluggish movement. The animals were admirable for the experimental conditions imposed. They can be handled singly and show no trace of the educative process which is said to occur in some forms, viz., Ameiurus, after having been sub- jected to trials of this nature. An attempt was made to obtain evidence of any difference in reaction which might occur in this way by taking the speed of the reaction as an indicator. The last control experiments, with the conditions as nearly as possible like the original experiment, showed an average time slightly longer than the original. During the process of acclimatization, the animals were fed solely upon small pieces of earthworm. The degree of visual sensibility is high. In many cases the animals will secure the earthworm while it is sinking to the bottom of the aquarium. If the food has come to rest, the animals often take great interest in the observer, and for this reason the walls of the aquarium 260 J. S. NICHOLAS must be screened for accurate observation in experimental work. The screening has the added advantage of reducing the amount of light in the aquarium and rendering the animals more quiet than they are under a light of higher intensity. The animals have a slight tendency to be more restless than usual under conditions of hunger. RESPIRATORY MECHANISM Bruner (?14a) has described the mechanism of the olfactory sense organ in relation to the respiratory mechanism in a com- parative study of a number of Amphibia. He states that in Amblystoma larvae the olfactory process is similar to that of frog tadpoles. ‘‘Respiratory water is taken through the nos- trils and is the only medium of smell.’”’ He finds that the passage from the mouth to the nasal cavity is occluded by a fold of mucous membrane, the choanal valve, which prevents either air or water from passing out in this direction. These choanal valves per- sist until metamorphosis. He distinguishes in general two dis- tinct types of respiratory mechanism: in one the respiratory medium is allowed to pass from the nasal cavity to the buccal cavity, the passage of return being occluded by the choanal valves; in the other the respiratory mechanism is completely under muscular control and the respiratory medium can’ pass freely in and out of the nasal cavity. According to the type of respiratory mechanism present, he classifies the forms as ‘‘monos- matic, single smellers, in which the olfactory organ is used to test only the external medium; and diosmatic forms, double smellers, including Siren, Cryptobranchus, Amphiuma, larvae of lungless salamanders and the adult stage of higher amphibians.” In the diosmatic forms, the olfactory organ is used in testing the content of the buccal cavity as well as that of the external medium. Bruner (14b) mentions that Jacobson’s organ is present in Amblystoma adults and that, from the existence of this organ, this form should be classed as a diosmatic form. As a preliminary to the experiments on the olfactory sense, it was deemed expedient to make a study of the water currents REACTIONS OF A. TIGRINUM 261 passing through the respiratory mechanism and their direction and disposal. Blinded adult Amblystoma were used on account of the ease with which they can be handled as well as the fact that they show less reaction to the light stimulus occurring under the conditions necessary for observation. The animals were placed in a flat glass dish under the binocular microscope and the water currents observed by means of a car- mine suspension. Three different courses of these currents have been observed. 1) A current taken in through the nasal open- ings and expelled through the mouth. This condition is the customary one persisting during the state of rest of the animal. 2) A current entering through the nasal passage and the mouth and expelled through the same channels. This is the condition occurring under stimulation, both the current and the excurrent phases taking place with a quick spurt. 3) A current of water expelled through the nasal passages (this being the discharge of the content of the buccal cavity), a special case occurring only when the animals comes to the surface to obtain ‘gulps’ of air. Air is taken in either through the nose or the mouth. The latter is used for the quick ‘gulps’ of air which the animal secures as it appears momentarily upon the surface, while the nasal apertures are used when the animal is in a state of rest, lying in the water with the snout protruded above the surface. The current designated above as 3 occurs infrequently as the aftermath of the second method. As a stimulus becomes at- tenuated by diffusion, the animal seems to take in water more quickly through the mouth that it can be discharged. At such a time the currents seem to be simultaneously going into the mouth and out through the nasal passages.’ 3 Vincent and Cameron find that respiratory movements cease in the frog as soon as the nares of the animal are under water. This condition has been observed by them over several weeks. A preliminary set of experiments, in which normal, anaesthetized and decerebrate frogs were used, shows that the inhi- bition of the respiratory movements is temporary and that water currents can be detected by means of the method given above. The efficiency of the valves covering the nares decreases with the time of immersion. None of the animals used in this work survived twenty-four hours of immersion. 262 J. S. NICHOLAS EXPERIMENTS ON LARVAE Four series of larvae were kept under observation, namely, 1) eyeless, those animals in which the optic vesicles had been removed; 2) noseless, those which had been deprived of the nasal placodes; 3) eyeless and noseless, those from which both the optic vesicle and the nasal placodes had been removed; 4): a series of normal animals which served as controls and were of considerable use in a comparative study. Under practically the same conditions, it could be easily seen that the normal larvae soon outdistanced the operated in growth. - The noseless animals were slightly smaller than the normals of the same age. Next in size were the eyeless forms, and finally the eyeless and noseless, which were considerably smaller than the eyeless. In a general way, the rate of growth shows the relative im- portance of the senses of sight and smell. The sense of sight is of more use in the obtaining of food at this stage than is that of smell. If the conditions of the amount of food present are varied, however, there is a change in the relative importance of the senses, for when food is scarce the sense of smell seems to domi- nate. This is shown in the following observation. The noseless forms seem to have no ability for distinguishing whether a particle is a food substance unless that particle has motion. The eyeless forms have a distinct advantage in this connection for they will distinguish pieces of organic material. If pieces of dead earthworm are placed in an aquarium, the noseless animals will push them around as they do other bits of debris, but will seldom snap at them. The eyeless, in direct contrast to this, will devour the earthworm immediately after coming in contact with it. A marked difference was noted in the responses to rapidly diffusing food substances obtained in the different groups of the larvae. When freshly cut earthworms are placed in the aquaria, the normal animals begin to swim rapidly about without making any attempt to locate the stimulus. They will often snap as they pass through the area occupied by the diffusing substances, REACTIONS OF A. TIGRINUM 263 but keep swimming on and do not stop at the point of the diffu- sion which would supposedly be, in this case, the center of at- traction. This stimulus is not sufficiently definite and the animal swims about until attracted (by sight) to the motion of the piece of worm, which is then seized. The same indications are given with beef juice or with strips of beef instead of earthworm. The animals are stimulated by the diffusion, but seldom engulf the food if it remains motionless. When the strip of beef is moved by means of an attached thread, the animal is quickly attracted and speedily engulfs it. The eyeless forms, under the same conditions, also show signs of stimulation. Their ability to locate food is less than that of the normal larvae. They are easily attracted to any slightly moving object, and since they lack the sense of sight it must be concluded that there is another factor entering into the reac- tion, that of sensitivity to vibration stimuli. Under conditions of starvation, an increase in sensitivity to the diffusion of passive or motionless objects must result, for the eyeless larvae will locate and devour the fecal material within the aquaria. The noseless forms are not stimulated by diffusing substances. They swim passively about the aquaria after the introduction of food materials of different sorts and unless the food exhibits motion it is entirely overlooked. Pieces of beef are completely ignored, producing no modification in the behavior of the animal unless they are agitated, when they are quickly engulfed. These experiments, made on larvae of Amblystoma tigrinum, thus confirm the results of Burr (16), who used A. punctatum. The work has been further extended by the study of individuals deprived of both eyes and nose. The eyeless and noseless forms show the same behavior as the noseless. In general, these larvae remain quiet in the aquaria and move about only at intervals. They are absolutely de- pendent upon mechanical stimuli in finding food. This is clearly shown by the way in which they ignore organic material such as killed earthworm or bits of beef. Under conditions of starva- tion, they exhibit a certain foraging reaction, pushing the objects about within the aquarium in order to discover whether inherent — 264 J. S. NICHOLAS motion is present in these objects. If such motion occurs, the animal is stimulated mechanically. At times, however, these . animals will snap at any object with which they may come in contact and engulf it with a subsequent elimination of the sub- stances unfit for food. In addition to the observations described above, several ex- periments were made using balls of filter-paper which had either been previously treated with juices of beef or earthworm or had been left untreated. A few experiments with sacks containing test substances were also made. As the result of these tests it may be said that there is a strong indication that the eyeless larvae distinguish their food substances by the sense of smell and mechanical stimulation. The normal animals are mainly de- pendent upon sight for the final location of food. The noseless larvae are dependent upon visible motion. The noseless and eye- less larvae have only mechanical stimuli to guide them. EXPERIMENTS UPON THE ADULTS 1. The reactions of normal animals in the light Reactions to motionless objects. 'The method used in this case was the same as that used by Parker and Sheldon in their experiments with Ameiurus, Mustelus, and Fundulus. An individual was placed in an aquarium, the sides of which had been screened to obviate any distractions that might occur from the presence of the observer. The aquarium was a glass- enclosed one, 42 cm. in length and 26 cm. wide. This was filled to a depth of 10 em. with water. After allowing a period of ten minutes for the animal to become adapted to the test aqua- rium, small cheesecloth bags, equal in size and suspended at the same height in the water, were introduced. These bags allowed a rapid diffusion of any substances that might be contained within them. The bags were hung from a rod at the top of the aquarium by means of a cotton thread. The time taken for the introduc- tion of the test materials was at that period at which the animal had come to rest at some position near the central portion of the aquarium. The bags were then introduced with as little dis- turbance of the medium as possible. Because of the variation of REACTIONS OF A. TIGRINUM 265 the conditions under which experimentation took place, the -stimulus caused by the introduction of the test sacks is negligible. One of the bags was filled with minced earthworm and was weighted with sand, while the other bag contained sand alone. Ten trials were made in each series, the three animals being tested in turn with at least a thirty-minute interval between trials. Under normal conditions, the adult Amblystoma moves slowly about the aquarium seeking for a projection upon which it can rest, preferably upon the surface of the water. Under condi- tions of stimulation the animal becomes restless, hastening its ordinary movements and proceeding with a greater degree of rapidity about the aquarium. It is this increase in activity that is spoken of in the following series as the ‘beginning reaction.’ All reactions in this series were considered to be consummated when the testing material was seized by the animal. The conditions under which each animal was tested were con- stant in regard to the motion of the test substances. The other factors, however, such as the distance of the test substance from the animal, the distance of the test substances from each other, and the position of the test substances with reference to the animal were considered. The variations in the conditions preclude any factor that might arise in connection with the above. The response to the stimulus was, in every case, a positive one. Differentiation between the test substances was always in favor of that sack containing the food material. In order to show a few of the experiments and the variations imposed as well as the procedure, two of the individual cases are given below. Trial 4. Animal A. Placed in aquarium at 11.05. Animal is restless. Test sacks introduced at 11.20. The animal moves restlessly about sides of aquarium, moving to center of the aquarium where the sacks are suspended within 2 inches of each other, noses earthworm bag, circles sand bag, goes to earthworm bag, noses, snaps, and seizes. Reaction is completed in two minutes. Trial’7. Animal A. Placed in aquarium at 3.10. Test sacks introduced 3.15. The animal swims about, passes sand bag, 266 J. S. NICHOLAS approaches earthworm bag, and seizes. The animal was de- tached from bag as quickly as possible and sand bag substituted - for it. Animal pays no attention to this bag after nosing it. Meanwhile earthworm bag had been placed at far end of aqua- rium. Animal turns quickly after nosing sand bag, swims to far end of aquarium, noses bag and seizes. The first reaction is complete in three minutes, the second, seven minutes after the introduction of the sacks. The results of this series are tabulated in table 1, series 1 (a), at the end of this section. Reactions to moving objects. The same kinds of testing sacks were used, the conditions of the trials being duplicated in so far as possible with but a single variation, that of a moving object. A rod, from which the testing sacks were suspended, was held by the observer. The time element is partially lost in this case, and because of the varying distances of stimulation due to the motion of the test sacks, it is deemed an unimportant factor n this connection. It is included in the description of the individual trials, in order to show the difference in the complete action time for the trials under consideration. The reaction is considered positive when the test sack is seized or an attempt is made to seize it. ; The record of one individual case will be given here as typical of the series. The results are tabulated in table 1, series 1 (b). Trial 1. Animal C. Placed in aquarium at 10.50. Sand bag, introduced at 11.00, kept in motion in front of the animal, but held at short intervals at different parts of the body. Animal turns at the first approach of the bag and noses it, after which it paid no attention to it, although it was kept moving about the animal for five minutes. Earthworm bag was substituted for sand bag after sand bag has been drawn to end of aquarium away from animal. Harthworm bag played before animal. Animal follows bag about aquarium, noses it when it comes to rest and then seizes. Sand bag substituted and animal allowed to nose it, does not follow it, but moves away, coming to rest at opposite side of aquarium. REACTIONS OF A. TIGRINUM 267 All the experiments were positive in this series, the animal showing a distinct discrimination between the test substances. 2. Terrarium experiments Reactions to motionless objects. A few experiments were made with the animals in a terrarium. The same substances were used as in the preceding trials. The animals are almost insensible to the presence of test substances, and in but one trial in ten was there any indication of a detection of the difference between the substances contained in the bags. This was the only reaction given, all other trials giving no reaction to either bag. The procedure followed was simply to place the bags near each other at the bottom of the terrarium and await the reaction. The one case recorded as positive showed a reaction about fifteen minutes after the introduction of the materials, the animal nosing the earthworm bag after passing the sand bag, and snap- ping the earthworm bag. A comparison between the sensitivity of the adult animal in the air and in the water can easily be obtained in the ordinary process of feeding. Cut earthworm can be held against the nasal apertures of the animal and elicit no response. If, however, the animal is so situated that its nasal openings are immersed in water, a very small diffusion from the forceps in which earthworms have been held is sufficient to give an extremely active response. Reactions to moving objects. If the test sacks are moved about in a terrarium in the same fashion as that described above for the animals in the water, a decided reaction is obtained. In many cases the animal seizes the bag with such force that it can be lifted from the bottom. This is particularly true with refer- ence to the earthworm bag, for the animal will quickly relinquish the sand bag. It is probable that the basis of discrimination in this case is one of taste. It is likely that this is the same reaction as that obtained in the noseless larvae, when the sense of taste or perhaps of touch, must enter into the reaction of the rejection of particles unfit for food. 268 J. S. NICHOLAS 3. The reactions of normal animals in the dark room. (In aquarium) The animals were dark adapted for two days before the experi- ments were undertaken and were given no food for three days. After dark adaptation, the animals are noticeably more sensitive to handling and the time required to come to rest is lengthened appreciably. A dim red light of mmimum intensity, Just suffi- cient for the observer, after a dark adaptation of forty minutes, to see the animal in profile, was used for the observations. The intensity was so weak that the figures upon a watch dial could be read with difficulty just in front of the screened light source. This degree of light is probably below the visual threshold of Amblystoma (Arey, 19), as evidenced in the reactions of the animals during the period of observation preceding the actual experiments. Ordinarily, with one side of the aquarium more brightly illuminated than the other, the animal, particularly if hungry, will tend to move about the sides of the aquarium with a preference for the brighter side, being often attracted to its own reflection in the glass wall, as well as by slight visible move- ments at the side of the aquarium. Under the dark condition, the animal moves slowly up and down until it is stopped by the ~ aquarium wall, then turning, it may proceed down the center of the aquarium, showing no particular preference for the wall. Reactions to motionless objects. The test bags were used as in the preceding section and the description of one individual case will show the procedure with the unvarying positive result (table 1, 3a). Trial 1. Animal B. Tested by the offering of the two test sacks. 'The earth worm bag and the sand bag were placed at opposite ends of the aquarium. The animal shows the character- istic motor reaction soon after the introduction of the material, but seems unable to locate it. Finally locates the object. Posi- tions of bags reversed and operation repeated. Animal swims restlessly about for some time before approaching object, noses it, and seizes it. Asan additional test, a glass vial containing freshly cut earthworms was placed in the aquarium. The animal immediately becomes restless and, after swimming about for REACTIONS OF A. TIGRINUM 269 short while, makes for the glass tube, approaching it from closed end which was purposely placed toward center of aquarium while open end of tube was situated toward side of aquarium and touching it. After nosing tube, animal follows up side of tube to get to open end, pushing tube across aquarium in an effort to obtain diffusing contents. An empty vial produced no reac- tion. Throughout this series of experiments, the reaction to the vial filled with earthworms was much faster than that given to the test sacks. The amount of diffusion is greater in the vial experi- ments and the diffusion takes place from one side of a definite focus in contrast to the test sacks which diffuse on all sides. It is likely that the strength or concentration of the diffusing sub- stances plays a large part in the difference of the speed of the two reactions. Reactions to moving objects. The animals were subjected to tests with moving objects as in the preceding section. One typical case will suffice to show the manipulation used and the general result. Trial 1. Animal B. Sand bag twitched gently in front of animal. Animal follows bag as it is drawn across the aquarium. Animal noses bag when it comes to rest. No further reaction. Trial repeated with same result. Earthworm bag twitched be- fore animal. Response immediate, animal follows, noses when at rest and seizes. The reaction time of normal animals in darkness is much slower than the reaction time in the light, but none the less definite. The sensitivity of the animal to vibration stimuli seems to be increased, as shown by the fact that the animal will follow the sand bag although moved very gently so that the amount of motion given to the surrounding medium is quite small. 4. Experiments upon nose-stopped animals Many methods for occluding the sense of smell have been tried by experimenters. The closure of the external nares by stitching has been used by Parker and his coworkers. The cotton-wool plug used by Sheldon and later by Parker has been 270 J. S. NICHOLAS satisfactory on the forms used by them. In some cases the olfactory tracts have been cut. The latter method is the most effective, but has the one disadvantage of not permitting a re- turn to the normal for control experiments. The cotton-wool plug has been found to have certain irritating effects upon the nasal epithelium and for this reason as well as for ease of handling, a collodion mask was used for occluding the nasal passages in this work. The method is quite simple. The animal is removed from . the aquarium and wrapped in a small hand towel with only the head projecting. The anterior portion of the head is dried as - much as is possible and the collodion applied over the surface prepared in this way and across the nares, making a mask the shape of an inverted T, the base of the T extending well up be- tween the eyes, while the cross line covers the external nares and extends well around the maxilla. The mixture dries rapidly and during drying causes some irritation. The animal will get rid of the cap if its forefeet are free at this stage. The cap will sometimes be removed by the force of the water from the mouth or by an excess of mucous secretion on the front of the head, but after three or four applications the animal becomes accus- tomed to the mask and shows little or no sign of irritation. At this stage the nose cap is more or less firmly adherent to the head of the animal. The animal is replaced in the water of the aquarium and generally becomes quiet in ten or fifteen minutes. If the cap has remained fast during this period, it will probably remain fixed for at least twelve hours, during which time experi- mentation may proceed. The respiratory process of the animal is changed by this occlusion of the nares. The animal comes to the surface and ‘gulps’ air by means of the mouth, then sinks and exhales slowly, the bubbles of air escaping at the mouth. The act of respira- tion affords a constant check upon the effectiveness of the nose mask, for if ineffective the water current under pressure from the buceal cavity will soon remove it. This mask is easily and quickly removed from the animals, after which they respond to control tests in exactly the same fashion as in the normal REACTIONS OF A. TIGRINUM DAG experiments. For the purposes in hand, the method was admir- able, although it did require a great deal of patience and con- siderable observation to make sure that the nose caps were tight. It was with the idea of making a combination hoodwink and nose cap that this method was applied, but the irritation resulting from the application of the mixture of lampblack and collodion to the eyes of the animal was too great to get normal reactions. Reactions to motionless objects. After securing a nose mask which was satisfactory, the animals were placed in screened aquaria in order that the observer might be concealed from view. The reaction time was much increased. Twelve minutes, three times as long as the average reaction time for the normal, was allowed before a test was pronounced negative. The reactions were recorded as negative when the animal made no effort to obtain either bag and also showed none of the characteristic motor activity of a stimulated animal, as positive when the animal attempted to nose either bag. In the reactions so re- corded, this reaction occurred within five to seven minutes after the introduction of test materials into the aquaria. After the attempt to nose, the animal paid no further attention to either of the bags. The reactions obtained show distinctly that animals with the sense of smell occluded cannot discriminate between test sub- stances, for there was in no case any preference shown, the animals nosing the sand bag, on the average, just as frequently as the earthworm bag. The reaction is, to a large extent, one of con- tact and is accidental. (Table 1, series 4 (a).) Reactions to moving objects. In contrast to the above reactions, the reactions of the nose-stopped animals to moving objects is quite rapid. The test sacks were moved about the aquarium at various distances from the animal. All the reactions gave an average reaction time of two and two-tenths minutes, the animals showing no discriminaton between the test substances, in marked contrast to the selective reaction obtained from ani- mals with the possession of the normal respiratory mechanism. (Table 1, series 4 (b).) 262 J. S. NICHOLAS ° 5. Experiments with blinded animals The animals were anaesthetized with chloretone and the eyes removed with a pair of fine scissors. After several hours the animals seemed normal and were fed. They were then not fed for a period of five days preceding the commencement of the following experiments. Animal A, which was not normal, due to an inflammation of the external nares, was not used in the course of these experiments. The tests were made with the usual test sacks. Results are tabulated in table 1, series 5 (a) and (b). The average reaction time in this series is almost the same as that for the normal. All the tests showed that the animals discriminated between the test substances, reacting positively to the earthworm sack. . In the reactions to moving objects, ten trials were made, all giving the same end result. The animals were quickly attracted by any movement in the water. They would not, however, follow the sand bag, although stimulation was shown by their motor activity. They did follow the earthworm bag and would, if possible, snap and seize it. 6. Experiments with nose-stopped and blinded animals The external nares of the blinded animals were covered with collodion and the animals were subjected to the same trials as those given above. The results are the same for both the moy- ing and the motionless objects. There was no attraction to either object in the motionless series. In the series with the moving object, there was a slight motor activity present but no definite objective. The results are tabulated in table 1, series 6 (a) and (b). 7. Partitioned aquarium experiments An aquarium was divided into two compartments by the inser- tion of a glass partition which ran diagonally across it, having a slit three-eighths of an inch high between the base of the parti- tion and the concrete bottom of the aquarium. This slit afforded the means by which diffusing substances could pass to the test REACTIONS OF A. TIGRINUM BS side. The arrangement proved very useful in the differentiation of sight stimulus from that of smell and was used on all of the - four series of animals. Two kinds of substances were used, dif- fusible and non-diffusible but visible. For the first, an infusion was made of minced earthworm, generally allowed to stand over night and filtered in the morning. The resulting extract con- tained no visible solid substance, but simply the juices and body fluids of the earthworms. This proved to be a very potent sort . TABLE 1 Reactions to test sacks NUM- g CONDITION OF TEST SERIES etc SUBSTANCE REACTION TIME minutes 1 10 |Aquarium (a)Motionless | Positive 3.8 10 (b)Moying Positive Nae zs 10 |Terrarium | (a)Motionless | Positive 15 ai = 10 (b)Moving Positive 3 10 |Dark room | (a)Motionless | Positive Over 30 j 10 |In aquarium] (b)Moving Positive : ore 9 locesteoped ..... 4 10 |Aquarium (a)Motionless | Attracted 1 10 (b) Moving equally 3 10 |Aquarium (a)Motionless | Positive 4.2 1 CLAUSE cme ie é 10 |Aquarium (b) Moving Positive Nose-stopped and 6 10 {Aquarium | (a)Motionless | Negative blmded® ...22% 3.22 10 |Aquarium | (b)Moving Negative of stimulus. It was applied by means of a pipette along the wall of the aquarium, and diffusing through the water passed under the partition. The observer was screened from view. The reactions were remarkably quick and vigorous, in many cases the animal moved to the central portion of the partition and snapped at the diffusing substances which it perceived but certainly did not see. The non-diffusing substances used were bits of rubber tubing and uncut earthworms. According to the work of Olmsted (718), 274. J. S. NICHOLAS uncut earthworms will stimulate Ameiurus by slime secreted by the skin. This is not the case in A. tigrinum, a fact which may indicate that the probable sensitivity of combined air and water ~ breathers is not so great as single smellers. Red rubber tubing of small diameter was used to simulate earthworms. Uncut earthworms were allowed to move about in the half of the aqua- rium not containing the Amblystoma. The rubber tubing was more effective in eliciting a response if set in motion by means of TABLE 2 Diffusing substances TRIALS SERIES SUBSTANCE REACTION 10 (1) Normal Earthworm infusion Positive 10 (2) Nose-stopped Earthworm infusion Negative 10 (3) Blinded Earthworm infusion Positive 10 (4) Blinded and nose-stopped | Earthworm infusion Negative Controlsdistilledswaters ee ves Mere so Seton a oiicives Giese een Negative TABLE 3 Non-diffusing substances NUMBER OF SERIES STIMULUS REACTION TRIALS 10 (1) Normal Uncut earthworm and tubing |Positive, 10 (2) Nose-stopped Uncut earthworm and tubing |Positive 10 (3) Blinded Uncut earthworm and tubing |Negative 10 (4) Nose-stopped and blinded | Uncut earthworm and tubing |Negative a cotton thread. Both objects acted solely as visual stimuli, as can be seen from the accompanying table (table 3). S§. Experiments with odor streams Risser has described an apparatus with which he tested the ol- factory reactions of the toad. The apparatus consisted of a test- ing chamber into which odors were brought from an odor chamber by means of an air current. A modification of this method was devised. The source of the air current was a compressed-air system, the air being run REACTIONS OF A. TIGRINUM 25 through a bath of water in a 10-liter container. This container acted as a reservoir and served to keep the pressure constant (fig. 1). The apparatus was first thoroughly tested for the effects of air currents under various pressures. It was found that an air current evidently stronger than that which Risser used had no effect upon the reactions of the experimental animal. ‘This current could be distinctly felt on the lips or the surface of the tongue of the experimenter. Text-figure 1 Diagram of the arrangement of the odor-stream apparatus From the flask (A) the air was run through a manometer (J7), and from there to the odor chamber (O). A fluid, the odor of which was to be tested, was placed in a vial in the odor chamber and the air bubbled through the liquid. ash sb Se hs -_ rag, \"! tales a Ais wr 9 eh 3 ’ P . Beare Gal Dy ibs am ¢ ¥ *h Ahh hadley ley hel Seagal 4iyvU1 eu bhe lie > , bo wid Fes "eel ty f-SUAYS hehe tek A la} ; ; ‘voor ae ey y's jira oI vali ek wv tess 3 oe Lh nye: U7 ay te Bory Golrric. UT ad eve ATT aR RR CS te a sida ks ae) aty ' is a a edt a rat fhe yh { tf 7 pate” MO Sani SOUT Ie RL SR Miiaatsitis here isi: rence. (MELE) |) Load LE a singly, mtpL Ned e i hig eis kal iNT Meh emda ote Dios, Mh Bigttie o 2411 hi Mit 124 ee ets aly at BE. gabe) 14 \0,,. 91st ie aan Ar. is y 4 A hoe ol) ge F Ee aan, ie ine (Comey a BG By MERIT INS ih speed CU) ay ‘ { Ar deh wie octrwn eat cate |) aap + Moas ce bf ' rus eer FA een ipa ify 1) UR ALOT ie PE CAT ASS) | et fe, 3: sad ae lay f + ine ' res) : Ds ’ SPO 89) 1558 ae : ¥ A ere | i hy nas rh w A ye fay. : hy tt if ] i c yas. a hic a i Srincey } sat; he es: Ce # pon hi pen igh ite virgo Ae Fis a ee ae a it Poa 7 4 ele a Resumen por el autor, Henri C. van der Heyde. Sobre la respiracién del Dytiscus marginalis L. El autor estudia el mecanismo de la respiracién de Dytiscus marginalis mediante un nuevo aparato. Cuando el animal respira en un espacio cerrado por una gota de petréleo pueden observarse movimientos en dicha gota. El primer movimiento observable parece ser una expiracién, cuyo hecho esta en abierta contradiccién con las afirmaciones previas de otros autores. El autor propone una teorfa para explicar este hecho tratando de verificarla de diversos modos. Mediantc un nuevo aparato estudia el aire después de haber permitido respirar al animal en él durante algun tiempo. El hecho sefialado por Ege y Kreuger de que el O, acumulado en la cidmara aérea es consumido durante la permanencia del animal debajo del agua has sido comprobado por el autor. La importancia biolégica de la difusién del oxigeno en la camara aérea es objeto de discusién bajo el punto de vista fisico-quimico y el autor la considera de un modo matematico. Por ultimo, la importancia relativa de los gases del agua y del aire para el comportamiento del animal son tam- bién objeto de estudio. Parece que atin cuando el animal se ha emancipado hasta cierto punto del medio ambiente origi- nario depende mucho mds del aire que del contenido gaseoso del agua. Translation by José I’, Nonidez Cornell Medical College, New York AUTHOR’S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, MARCH 20 ON THE RESPIRATION OF DYTISCUS MARGINALIS L. H. C. VAN DER HEYDE Physiological Laboratory of the Free University in Amsterdam, Holland THREE FIGURES For a comparative physiologist fresh water is the best medium from which to gather his material. It is, as Professor Jordan! pointed out, very poor with regard to its more highly organized primitive forms. Nearly all representatives of those groups which are called ‘high’ in the scale of phylogenesis of the mor- phologists and systematists, especially the insects, prove by their structure that doubtlessly they originally inhabited the land. As the structure of an organism is determined on one side by its phylogenesis, but on the other hand by the claims of ‘Milieu’ and ‘Umwelt,’ we may expect that this change of sur- roundings has caused certain morphological and physiological peculiarities which we are accustomed to call phenomena of ‘adaptation’ or better perhaps of ‘being adapted.’ Without trying to give an explanation of these phenomena, it is the task of the comparative physiologist to state them and to study them with all applicable methods. ‘Then it will be of special interest to see how far a certain species has emancipated itself from its original environment and how the same problem has been solved in quite a different way in related groups of animals, when we compare, for instance, Dytiscus and Hydrophilus with regard to their respiration in the groups of the Adephaga and the Poly- phaga. The respiration of Dytiscus has already been studied by numer- ous investigators, though mostly from the morphological stand- point. In this paper I shall mention some experiments which seem to throw a new light on the whole problem. 1 Prof. Dr. H. J. Jordan. Het leven der dieren in het zoete water. Utrecht. Oosthoek, 1918. 309 336 H. C. VAN DER HEYDE The mechanism of their respiration was already known to Swammerdam; he also knew that breathing in Dytiscus is entirely different from that in Hydrophilus. The old naturalist Frisch? had not seen it, and in observing Dytiscus only was led to gener- alize. Nitzsch,? a more careful observer, corrected this and emphasized the difference between the two species. It is not my intention to give a complete history of our knowl- edge of this problem. Such review should be absolutely com- plete or not be given at all. Moreover, the interested reader can find a complete though not wholly impartial summary of the older investigations (till about 1912) by Babak in Winterstein’s Handbuch der vergleichenden Physiologie, I. Bd., 2 Halfte, p. 452 seq., while most of the more recent studies have been mentioned in Ege’s paper.‘ The problem to be solved here has many sides. In the first place, the animal must be able to stay for a certain time in the water. In consequence of its type of organization, however, it is not entirely independent of the atmospheric air. Intermit- tently it must come to the surface to breathe, and now the problem has to be solved how to penetrate the surface layer. This prob- lem has been solved in an entirely different way by Dytiscus and Hydrophilus. Dytiscus breathes with the edge of its abdomen. By com- plicated movements of this abdomen which have been studied in great detail by du Bois-Reymond? and Plateau® and in which the muscles of the ‘Genitalkapsel’ play an important r6le—these 2J. L. Frisch. Beschreibung von allerley Insekten in Teutschland. 2 Teile. Berlin, 1721. 3Ch. L. Nitzsch. Ueber das Athmen der Hydrophiliden. Arch. f. Phys. Reil und Autenrieth., Bd. 10, 8. 440, 1811. Rich. Ege. On the respiratory function of the airstores carried by some aquatic insects (Corixidae, Dytiscidae and Notonecta). Zeitschr. f. Allg. Physiol., Bd. 17, S. 81, 1918. 5 R. du Bois-Reymond. Ueber die Atmung von Dytiscus marginalis. Ver- handl. der deutschen Physiol. Ges. Arch. f. Phys., 1898, 8. 378. 6 F. Plateau. Recherches expérimentales sur les mouvements respiratoires des insectes. Bull. Acad. Roy. Belg., T. 3, p. 727, 1882. RESPIRATION OF DYTISCUS MARGINALIS L. Jan muscles have for this reason been called accessory respiratory muscles—a cleft is formed between the edge of the abdomen and the two elytra. In that way the large space between the back of the animal and these elytra communicates freely with the atmospheric air. The edge of the abdomen is fatty, so that the water is prevented from coming into the dorsal space. In the opening itself we see a number of stiff hairs which probably have the purpose of keeping out dust particles and on the other hand prevent the air from escaping after the animal has dived into the water. Corresponding to this abdominal breathing, the whole bunch of stigmata has been moved backward. Without the aid of a lens two very large stigmata can be seen at the edge of the abdomen while the animal is breathing; moreover, all the other stigmata have moved in the same direction. Numerous anatomical details are given by Willy Alt.? In connection with this brief description of the breathing process I wish to state that I am still in doubt whether the opening of the abdominal cleft occurs actively or simply in con- sequence of capillary forces as, for instance, the ‘hair-crests’ of Notonecta do. I base this doubt on the following observation: In order to determine the consumption of oxygen of the animals in the water, I put two beetles into a beaker and covered the water with a layer of paraffin oil. After some time the beetles moved the edge of their abdomen towards the limiting layer ap- parently with the purpose of breathing. It appeared that they were not able to keep the cleft closed, though they had not yet, reached the air. In that way the abdominal space was filled in less than no time with the oil, and the animals were completely motionless after a quarter of an hour. In order to keep them alive, I brought them back into the fresh air, but in less than two hours they died. When we take into consideration that the animals can live for much longer than an hour in the water with- out reaching the surface, we cannot explain this result by lack of oxygen. 7 Willy Alt. Ueber das respirationssystem von Dytiscus marginalis L. Zeit- schr. f. wissensch. Zool., Bd. 99, S. 357, 1912. Auch Zool. Anz., Bd. 34,8. 793, 1909. 338 H. C. VAN DER HEYDE The method of respiration of Hydrophilus is entirely different; it is, however, not my purpose to enter into details with regard to,this species in the present paper. Above all, it seemed desirable to me to study the breathing mechanism on the intact animal. Curves such as have been traced by Babak* may give rise to the objection that they do Fig. 1. Apparatus for the registration of the respiratory movements of Dytiscus. ' not represent the normal breathing mechanism of the beetle. This author made an opening in the elytra, fixed the animal with pins, and fastened a hooklet in its back. In that way he could register the movements of the back on a kymograph. The great difficulty however, which must be surmounted is the fact that the whole respiratory apparatus is completely covered by the large elytra. On the suggestion of Professor Buytendyk, I finally used the apparatus of figure 1. 8 Prof. Dr. Edw. Babak. Mitw. stud. med. J. Hepner. Untersuchungen iiber die Atemzentrentitigkeit. I. Ueber die Physiologie der Atemzentrentitigkeit von Dytiscus mit Bemerkungen iiber die Ventilation des Tracheensystems. Pfliiger’s Arch., Bd. 147, S. 349, 1912. RESPIRATION OF DYTISCUS MARGINALIS L. 339 In the vessel A I put the beetle in a certain quantity of water. The vessel B has the purpose of compensating occasional dif- ferences in temperature and vapor pressure. Both vessels con- tain about the same amount of water. In the capillary tube, a, a drop of petroleum moves up and down. Petroleum ‘is to be preferred to water in consequence of its smaller friction coefficient. The utmost care should be taken to keep the capillary tube absolutely dry; the sensitiveness of the whole apparatus is lost as soon as any water penetrates into this tube. In b and c we have a three-way cock, the purpose of which can be easily under- stood from the figure. I must still mention, in the first place, the glass stopper, d, with which we can bring the droplet to any desired spot when the apparatus is ready for use; in the second place, the tube e through which the water can leave when the apparatus is filled from above with some gas, e.g., Hz or COs. In ordinary circumstances they are closed with a glass stopper. Care must be taken to keep the tube a absolutely horizontal; finally, the apparatus was put in water to keep the temperature as constant as possible. Let us now describe systematically the movements which the droplet makes in the different periods of the breathing process. A. As soon as the animal reaches the surface and opens its abdominal cleft, the droplet first does not move. Whatever may take place during this period, it is clear that in this way an occasion for an exchange of gases between the dorsal space and the atmospheric air is given. B. A short time after this the droplet moves first slowly, then more quickly, sometimes intermittently and with pulsations in the direction of the distal end of the tube. Obviously, the air is expelled in some way either from the dorsal space or from the tracheae. The first phase of the breathing process is thus an expiration, not an inspiration, as Babak (l.c., p. 350) and almost all other investigators postulate. Details will be given later on. C. Almost always the animal dives immediately after this. Sometimes I could observe a recession of the droplet before the animal dived; the distance covered, however, in these cases was never more than about one-tenth of the distance traversed during the expiration. 340 H. C. VAN DER HEYDE After the animal has dived we can observe the following move- ments of the droplet: Animal opens cleft. Fig. 2 Schematic representation of the movements of the droplet D. In the first place, a regular retrogression of the droplet at a slower rate, but over a greater distance than was covered during the expiration. When the animal stays under water for a long time, the droplet scarcely shows any more movements. RESPIRATION OF DYTISCUS MARGINALIS L. 341 E. In the second place one can regularly observe very distinct hydrostatic movements. When the animal moves up and down in the bottle the air in the reservoir is alternately expanded and compressed. The extreme sensitiveness of the apparatus is shown by the fact that even these minute changes in volume can be read in the capillary tube. This gives, in the second place, a good test as to whether or not the apparatus is working all right in a certain series of experiments. Sometimes these movements do not show because the cleft is completely closed. F. Finally, the definite proof can in this way be given that the animal breathes under water. This fact has been denied by > several authors. Immediately after diving the animal is usually very motile. After some time, however, it comes to rest and may sit quietly at the bottom of the vessel for a while. Then one can observe several times a very distinct moving up and down of the droplet: in the meantime the gas-bubble in the cleft of the abdomen makes synchronous movements. The latter move- ments have already been observed by some authors, while others did not seem to realize their importance. A diagrammatic representation of the whole process is tried in figure 2. The movements of the droplet are supposed to be projected on a vertically moving plane. In that way the ordinate gives the time, while from the abscissa one can read the location of the droplet at every moment. We must now try to give an adequate explanation of these facts, and in doing so we must keep in mind the following considerations: 1. It has been shown by Ege‘ that the oxygen of the reservoir is consumed during the animal’s stay in the water. I could prove the exactness of this statement by the following analyses, which have been made with Krogh’s apparatus for micro-gas analysis:° 9 Aug. Krogh. On micro-analysis of gases. Scand. Arch. f. Physiol., Bd. 20, 1908. 342 H. C. VAN DER HEYDE TABLE 1 LENGTH OF! TEMPERA- EXPERIMENT BUBBLE TURE RESULT 1. Dytiscus, immediately after diving I press | 15.35 15.3° | 0.65% COs with my fingers a little air-bubble out of the | 15.25 15.5°..| 14.01%" Os air-store; this bubble is immediately brought | 13.10 15.7° | 85.34% Ne into the apparatus and analyzed 2. The same, somewhat later 13.90 16.2° | 0.72% COz 13.80 16.5° | 12.81% Os 12.02. 16.6° | 86.47% Ne 3. The animal is prevented from coming up for | 15.49 16.3° | 0.64% CO» some time by being shut in an inverted beaker. | 15.39 16.5° | 3.05% O2 After this analysis as before 14.92 16.7° |96.31% Ne Analogous results have been obtained by Elsa Kreuger? in a series of very careful experiments. This author studied, more- over, the rate of oxygen consumption, and could show that after a very rapid decrease in oxygen content in the beginning a kind of equilibrium is reached after a while between the tracheal air and the air of the reservoir, so that after that the curve proceeds much less steeply. 2. The’animal really expires when it opens its breathing cleft at the surface. I could prove this fact which had already been made very probable by the movements of the droplet in the apparatus of figure 1 in using the following little apparatus, not previously described (fig. 3). It simply consists of a beaker covered by a flat cork. A space, A, has been made in the center of the cork" in which a bubble of air can be brought through the pipette, B. The animal is allowed to breathe into this bubble after an equilibrium is practically established between the 10 Wlsa Kreuger. Ueber die Bedeutung des Elythralraumes bei Dytiscus. Lund Universites Arsskrift. N. F., Bd. 10, No. 13, 1915. Kongl. Fysiografiska Sallskapets Handlingar. N. F., Bd. 25, No. 13. 11 The air is kept in a hollow metal platelet, which is fixed to the cork by means of screws while the spaces above it were filled up with plasticine. 12 For this reason the cork must not fit too tightly in the beaker so as to allow the water to escape at the rim. RESPIRATION OF DYTISCUS MARGINALIS L. 343 bubble and the water," and after a sample has been taken of the air with one of the pipettes, C, D, and FE. These pipettes are built exactly like the ones described by Krogh. Care should be taken not to make them too long. After breathing, another portion is taken, and quickly the analysis is made, first of the third one, because the air is not in equilibrium with the water, afterwards of the control portion. Fig. 3. Apparatus for the study of the expired air of Dytiscus. By moving a water-filled funnel connected with B upwards, the air of the bubble A is pressed into the space F, whereas the superfluous water escapes at the rim. The animal is allowed to breathe in the space F. With the pipettes C, D, and HE (D and E have been indicated with a broken line, as they do not lie in the plane of the picture) we can at every movement get a sample of the air. With this little apparatus I obtained the following results with regard to the composition of the air before and after breathing (the analysis has again been made with Krogh’s micro-gas analy- sis apparatus): -BIt is very easy to prevent the animal from breathing for a while by shaking the beaker at the very moment when it intends to open its cleft for this purpose. 344 H. C. VAN DER HEYDE TABLE 2 BEFORE BREATHING AFTER BREATHING sth Gite Ateoonipaemae Length of | T zen em Ta- nN. empera- Babble. For Result Bubble bore Result 1 14.33 23<0> 0.21% COs 14.07 PALS Oe 1.07% COz 14.30 DRA ZOE ZAG Os 13.92 2 iis 13.07% Os 11.40 23002 79.55% Ne 12.08 BP? 85.86% Ne DZ; 14.90 DASRS 0.14% CO. 1 Aa? 23 ae 1.68% COsz 14.88 24.9° PABA IOs 16.99 23.8° 15.29% Os U2 25.02 78.65% N2 14.35 DA oe 83.08% Ne These figures show once more: 1) that the animal expires as soon as it comes to the surface; 2) that the air-store at the end of the period during which the animal remains in the water con- tains much nitrogen, little oxygen, and less carbon dioxide than one would expect. 3. In insects, contrary to what we find in vertebrates, expira- tion takes place actively as, according to Ege, Brocher™ strongly emphasizes. 4, Dytiscus is able to obtain some oxygen through diffusion from the water. What this means and in how far this oxygen intake has a biological importance will be discussed later on. Here I simply describe the following experiment. One beetle was put into a beaker. In order to avoid as much as possible the diffusion of oxygen from the air into the water, I covered the surface with a floating cork with a little hole in the center for breathing (paraffin oil cannot be used, as stated pre- viously). Moreover, the beaker was covered by a glass plate. Before and after the experiment the oxygen content of the water was determined by the method of Winkler.“ Experiment A. Animal in the water during 24 hours. Before the experiment: 11.27 cc. thiosulphate. After the experiment: 10.47 ce. Difference: 1.80 ce. _ 14His paper, Recherches sur la respiration des insectes aquatiques adultes. II. Les Dytiscides. Ann. de Biol. Lacustre, T 4, 1909/11, was not available for me. 15 T did not yet know the remarkable improvement of this method by Edwin B. Powers, published in the Bull. o. the Ill. State Lab. 0. Nat. Hist., vol. 11, May, 1918. RESPIRATION OF DYTISCUS MARGINALIS L. 345 Experiment B. Animal in the water during 1$ hours. Before the experiment: 10.18 cc. thiosulphate. Afterwards: 9.82 ec. Difference: 0.36 cc. thiosulphate. Unfortunately, I cannot find the volume of the water used in these experiments in my records. Nevertheless, their results prove that some oxygen has disappeared from the water, and this is the only point I wished to state. 5. A glance at the figures given in table 1 for the composition of the air in the animal’s store immediately after diving shows us that this air contains an abnormally high percentage of carbon dioxide. The figures of Ege and of Kreuger agree in this respect completely with my own. We may therefore safely conclude that this air is a mixture of expired and atmospheric air, a ‘Mischungsluft.’ My first impression of the fact that the first phase of the breathing process is not an ‘inspiratorische Schluckbewegung,’ but an expiration, was that it would complicate our present knowledge of the problem to such an extent as to make it a chaos of contradicting facts. Certainly, it proves that a great deal of the literature based on this hypothesis is worthless. It is clear that the only way in which the animal can cause a movement of the droplet in the direction of the distal end of the tube is by reducing the pressure somewhere else, and it is very probable that this reduced pressure will be in its tracheal system and tissues. As the animal’s body fluids are incompressible, no other possibility can be realized. Let us now suppose that in diving the animal takes with it a certain quantity of air, A. The oxygen in the air-store is used up as I showed in 1. The animal gives off carbon dioxide instead of this oxygen, but, as the analyses of three authors show, this CO, diffuses out very soon into the water. Consequently the volume of the air-store must diminish. Moreover, after some time, as both CO, and O, are very low, the partial pressure of nitrogen will be higher than in the water, which is in equilibrium with the air, and nitrogen will diffuse out. All these factors tend to diminish the total volume of the air-store and make it less 346 H. C. VAN DER HEYDE than A. And now when the animal comes to the surface it expires! After that it dives again! It is obvious that in this system of reasoning there must be a mistake. In the beginning I tried to find its solution in the process under water and tried to solve the following question: Is the diffusion of oxygen into the bubble sufficient to cover this loss? This means: ‘Has this oxygen diffusion a biological importance? This has been denied by Ege in his paper cited above. My first hypothesis was that perhaps this author had made a mistake in his complicated calculations or in the determination of the different data used in these calculations. I** therefore took up the problem in my own way, but my results are the same as those of Ege. a. Bohr has given a formula (used also by Ege) which enables us to find out the quantity of gas which diffuses into or out of a bubble in a certain time. The modification which Ege used is ie = in which p: — pz is the difference in pressure 10s) = 0 between bubble and water (Ap), M the change in volume (Av), y the diffusion coefficient of the gas and S the surface through which diffusion takes place. What I wanted to see was whether O2 would diffuse in faster than N, diffuses out or not. Now, when we suppose that one diffuses as fast as the other, we have: = = — & z Tl Op At Ne Substituting Bohr’s formula, we have: BS. Vo. 4Do, =! Biyn,. ADw: 760 7a000. Substituting y,, = 0.029 and yy, = 0.009 (Ege), we get: Nog. == 0:31 Apy. . In “the water py, 1s, 601° cme See, pressure O,. is 159 mm. Hg. When we take, for instance, the analytical result of our table 1, analysis 3, as standard for the composition of the air after the animal has moved for some time in the water, Apo, appears to be — 136 mm. Hg. To balance 16 T am very much obliged to my friend the physical chemist, Prof. Dr. Allen E. Stearn, who helped me in working this problem and had the kindness to look over these pages. RESPIRATION OF DYTISCUS MARGINALIS L. 347 this Apo,, Apy, must have the value of 421 mm. Hg. Actually this value is only 130.8 mm. Hg. This proves that oxygen will diffuse in quicker than N, diffuses out, even when we do not take into consideration the higher value of CO, in the bubble. 6. In this way it has been proved that oxygen really diffuses faster into the bubble than nitrogen diffuses out. It seems even possible that this inflow of oxygen may compensate to a certain extent for the decrease in volume of the bubble. Another question, however, arises: Will this inflow of oxygen enable the animal to stay under water? In other words: Will the inflow of oxygen compensate the consumption of oxygen of the animal and have in that way a vital biological importance? In that case the only reason for the animal to move upwards would be the lack of nitrogen, though this seems to be a paradox. This question has been denied by Ege, and I completely agree with him, though I believe that it can be demonstrated in a much more simple way. According to Ege’s figures, one Dytiscus consumes in one minute 8 mm.’ of oxygen. The quantity which diffuses in can be calculated by means of Bohr’s formula: M = use ea =a. Taking p = 136 mm., and s = 10 mm.? (as Ege did), 0.029 x 10 x 136 760 of oxygen diffusing in. The discrepancy between these two figures is evident. vy. In this way we see that the term ‘biological importance’ which Ege introduced is ambiguous and should be avoided. As far as the oxygen economy is concerned, the oxygen diffusion has no ‘biological importance’; that it plays a role in preventing the volume of the bubble from decreasing too quickly has been shown in the given calculations. 5. The possibility which occurred to me in the beginning that by the inflow of oxygen the volume of the bubble might increase so that the animal would be obliged to move upwards simply to get rid of its superfluous air—in that way the frequent escaping = 0.52 mm.’ 348 H. C. VAN DER HEYDE of air bubbles from the dorsal space would be explained—is dis- proved by the fact illustrated in table 1 (decrease of oxygen till about 3 per cent) and by our above-given calculations. e. Concluding these theoretical considerations, we may state: 1) that from the statistical standpoint a state of sliding equilib- rium between bubble and water will be reached after a while in which O, diffuses in and compensates the outflow of N» (and CO,?); 2) taking into consideration the animal’s consumption of oxygen, we can see that this inflow has no real ‘biological impor- tance’ in the sense in which Ege used this expression. This attempt at an explanation revealed some interesting facts, but does not give the solution of our problem. It appears that the problem cannot be solved by applying physical chemistry to the processes which happen during the animal’s stay in the water. So we are logically obliged to find its solution in the process of breathing itself and in doing so we must keep in mind: 1) that the animal expels the droplet of petroleum—this proves that a vacuum must be made somewhere as shown above; 2) that the air in the dorsal space in the moment of diving is a mixture of atmospheric and tracheal air (see 5). I believe that these facts enable us to give only one explanation of the whole mechanism as follows (though it took me a long time to find it—it is something like the egg of Columbus and extremely simple after one has realized it): As soon as the animal has opened its abdominal cleft, it moves its back slowly downwards. Consequently, the air in the stig- mata is compressed just a little bit and flows out, partly through the dorsal, partly through the terminal stigmata. In that way no movement of the droplet can be expected because no vacuum nor noticeable compression is made. In this connection it 1s remarkable that in several experiments I could observe that the droplet continued its backward movement for a few seconds after the cleft had made communication between the dorsal space and the atmosphere. This may perhaps be an indication in favor of my hypothesis. The only effect of the postulated move- ment is that the tracheal air is expelled from the tracheae into the dorsal space and perhaps into the free air (in the latter case RESPIRATION OF DYTISCUS MARGINALIS L. 349 a compensatory inflow of air from the atmosphere into the dorsal space is to be expected). Most probably the period called A in our description corresponds to this movement. Now the back of the animal is raised again. ‘Through the stigmata the air sucks into the newly formed vacuum, but it is not able to pene- trate soon enough into the tracheae to compensate the vacuum. Consequently, we must expect the droplet to move into the direction of the distal end of the tube, as in fact was observed. Part of the air in the dorsal space is expelled, and as soon as the back has returned into its position of equilibrium the animal dives (this position will most probably depend on the hydro- static function of the air-store which has been emphasized by Brocher (p. 344) and Wesenberg-Lund?!’). From that moment the movement of the droplet must go in the opposite direction, which is in fact observed. Never could I notice any exception to this postulate. Now here is another feature in favor of my conception of the whole process. When the backward movement of the droplet was due only to the diffusion of the expired CO: into the water, we would expect it to be very slow—in the case the oxygen inflow compensates the outflow of N. and that of CO, completely the droplet would even stay where it was:- This is, however, not the case: the droplet goes back rather speedily though not as speedily as during the expiration (observation D and fig. 2). This must be due to the gradual filling up of the vacuum by the air which sucks through the stigmata! After a while this vacuum must be filled up, and now we may expect the droplet to move much more slowly, perhaps even to come to a standstill. This was actually observed (observation D). In this way no difficulty can be encountered any longer in the explanation of the movements, so far as I can see. Some more experiments have been made to study the degree of emancipation of Dytiscus from its original medium—the air. I had the animal respire in different atmospheres while, more- over, the gas content of the water was varied. In this way I hoped to get an impression of the relative importance of these 17 Wesenberg-Lund. Biologische Studien iiber Dytiscus. Internat. Revue d. Hydrobiologie und Hydrographie. Biol. Suppl., 5 Sér. 1912, p. 89. 350 H. C. VAN DER HEYDE two factors. In each experiment the time spent at the surface was determined by means of a stop-watch and tabulated. The experiments lasted half an hour; in the table I also indicated the ten-minute periods. The experiments were: TABLE 3 Experiment 1. Dytiscus @. Water normal. (33 .0mins3) 355,20 mines 25 )otalt4s: Experiment 2. Dytiscus o&. Water boiled. §.2,32.1004.9.10.1,4/; 10. min. > '7.5.3.3.4.7.109.6.0.10; 20> min. 7 die 160.10.63. Total 270 + 162 + 264 = 696. Experiment 3. Dytiscus &. Water normal. CO: atmosphere. In the first 10 min. the animal stayed about 100 sec. under water, after that it was about continuously at the surface. Experiment 4. Dytiscus o&'. Water normal. Hz, atmosphere. 20.2.1.2.60.5.34.40.55.50.3.25; 10 min.; 50.4.1.15.150.230.5.10.120.5.10.120; 20 min. ; 70.30.145.31.65.80.150. Total 297 + 585 + 570 = 1452. ~ : The animal gave the impression that it was not able to dive. Perhaps the hydrogen gives a marked decrease in specific gravity so that it is made very difficult for it to dive. Experiment 5. Dytiscus ~. Water boiled, allowed to cool and then saturated with carbon dioxide. Air normal. 2.2°6.2,0.1,0.16.1.2.1.1.1.4:7; 5 min.) 5:1,29. 11-211 4s 0) min 2 1.60; 15 ‘min. ; ‘55.1.3; 20 min.; 2.28.32; 25° min.; 14.3:5.4.2.4.21 4. Total 111 + 197 + 119 = 427. These experiments show clearly that not only the composition of the atmosphere, but also the gas content of the water has a noticeable influence on the animal’s behavior, as far as breathing is concerned. When we compare, for instance, experiments 1 and 2, we see a remarkable increase in the number of times in which the animal goes to the surface and in the time spent there. This is only due to the fact that in one case the gases have been expelled from the water so that the air in the air-store must dif- fuse out. Very remarkable in this connection is the fact that less time is spent at the surface when the water after having been boiled and allowed to cool is saturated with CO, (exp. 5). In that way CO, cannot diffuse out and decrease in volume is at a much slower rate. In the second place, these experiments show once more that the importance of this gas diffusion has not been realized sufficiently RESPIRATION OF DYTISCUS MARGINALIS L. 351 by most students of this problem. Once more we can see that it really has a ‘biological importance,’ though not in the sense in which Ege has used this expression. Much more important than the air content of the water is the composition of the atmosphere. A glance at experiments 3 and 4 shows us immediately how in case of substituting some other gas for the normal air the animal almost continuously hangs at the surface. CQO, has a stronger influence than H., perhaps be- cause it stimulates in some way the centers for breathing. In this way we are able to demonstrate that Dytiscus, though it changed its medium, is still dependent of its original milieu and that changes in this initial environment affect it much more than changes in the water. This fact is also nicely illustrated by the observation of Plateau,!8 that some ‘land’ beetles could endure immersion in the water without contact with the air longer than the water-beetle Dytiscus. Oryctes nasicornis, for instance, could on the average stay ninety-six hours under water, Dytiscus only sixty-five hours. Nevertheless, we can mention certain facts, phenomena which without doubt prove a certain adaptation to the new medium. Among these could be mentioned anatomical features, as the structure of the legs, the backward movement of the stigmata, the establishment of the dorsal air-chamber, and the structure of the cleft, but also physiological peculiarities, such as the whole breathing mechanism, as it has been analyzed in the present paper. The facts that the beetle is in some way in- fluenced by the gas content of the water, that the oxygen inflow into the air-store is necessary in order to prevent it from decreas- ing in volume too rapidly, show that the animal has established some relations, some ‘Wechselbeziehungen’ with the water. These phenomena we might call phenomena of adaptation. 18 F. Plateau. Recherches physicochémiques sur les articulés aquatiques. 2me partie. Bull. Acad. Roy. Beg., 41me Ann. 2me Sér., T. 34, p. 263. a0 H. C. VAN DER HEYDE SUMMARY With a new apparatus the mechanism of breathing of Dytiscus marginalis L. is studied. The movements of a droplet of petro- leum are observed when the animal is allowed to breathe in a space which is closed by this droplet. It appears that the first observable movement is an expiration, which fact is in contra- diction with previous statements of other authors. A theory is given to explain this fact and in several ways the verification of this theory is tried. With a new apparatus the air is studied after the animal is allowed to breathe in it for some time. The fact reported by Ege and Kreuger that the O, in the air-store is consumed during the animal’s stay under water was proved again. The biological importance of the oxygen diffusion into the air-store is discussed from the physico-chemical standpoint and treated mathematically. Finally, the relative importance of the gases in the water and in the air for the animal’s behavior is studied. It appears that though the animal has become emancipated to a certain extent from its original environment, it is still more dependent on the air than on the gas content of the water. ; Thanks are due to Prof. Dr. Withrow Morse who tried to make my English readable even for the English reader—and I trust he succeeded—and corrected the many mistakes. Morcantown, W. Va., U. S. A., January 6, 1921 AUTHOR’S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, MARCH 27 THE TRANSPLANTATION OF SKIN IN FROG TADPOLES, WITH SPECIAL REFERENCE TO THE ADJUSTMENT OF GRAFTS OVER EYES, AND TO THE LOCAL SPECI- FICITY OF INTEGUMENT WILLIAM H. COLE Zoélogical Laboratory, Harvard University TWO TEXT FIGURES AND FOUR PLATES (TWENTY-TWO FIGURES) CONTENTS iG OGIC OTM ee ent 5 ee re wne cai aie ta. o:Byc sors, 2 bi bopaspratdey Poe eis eile adece 354 eer esteeer TUPI TIES GILGORIE cory yee oe a hue cLanctane seca cpa. 2ieie aus e d agea cgreys's bee raleleye 357 (Slam erl GLS Sere ie a0 RS tA A ae ie a Me Aa 360 Lt LET OUST 120 PCr Ea Aan ge gear og RAL De el te a RS fT 360 2D acncripiionrol tie healing processss. |. jccs, ya sove woe « boca avers byes anew a> toners 363 Seorohiteratiomeotuohessransplantse a: ge scien s< 5 qe le we aoe ssn chee 365 Meriva krapustment OF STATS). ceccc cot ccc acs. ccscocreeacds se Shane cca 367 1. Grafts over eyes kept in the light (series LN) or in darkness (series ENS ee peg tester bs ick 9 oi Aaya! Bie oo, ty, Ba en Mie wicat ces clas apanstin sca) bk toed cays Fle 367 Game elas kalinnn. pte ces ann Sinha ee teh coc mem incurs cave tine: Bic Resmi sree 367 (Ba, LBAEY Cle eb erie ys phe sti alpen i am als gen «Mage Fai Ni ti eee a ae 377 2. Grafts not over eyes (series LB) or over operated eyes (series LO).... 378 _ 3. Grafts over artificial eyes, series LBE and DBE..............6....4.. 383 MacAWspeciicibygOlamMpe owen bc. .cl yhieciy ot sedis-o link ciciasom aeysse ets «obese iSyrieue 386 mm TTA A BIEN COULEIN CIM apc ta aya ters. c. cies cta evel cass eeeavecege SALE Mere, avd tgs ASR anaaetanels see CSAs 386 EM LERCAT UPL OUTROS cr hs Rea Sherry SAR PRL BEE! BA ATE Eh Aer, ME, cate 387 PPeLMORIGiOLE Aa An iat aceee her aes si tth SSeS cAiis waldo estaceels sees gala ae 390 Acquisition of melanophores by white grafts on a black region............. 395 CALLEN gRPAEE SESE IES este OMEN STi ch hci Sys S/ehege os cs wasn a She peas Sean So ne OS 395 PPMELOMTOTO ELAS OAT LSM Pama ecites scree aie Corclecs Oita Geers cceiticle one 396 Pigmentation of the conjunctiva caused by injury...................0 cece 401 ee otal pease GL GOR CISION Bs ic;2:.tgsalres Sra5 5 1 eee ‘ Ant. only | Post. only | Both ends | Amorphiec 86 | h. tail | 270° | sym | 100% Ex. 10 | tail N | sym | 100% Ex. $s 18 | tail R' | sym | 100% Ex. re 27 | tail N | sym | 100% Ex. 31 | tail R | sym 50% Ex. EE: 32,| tail R |sym{| 33% Ex. s 43 | tail N_ | sym i “s 44°| tail N | sym | 100% Ex. 45 | tail R | sym sl. abs. e 48 | tail N | sym 5 50 | tail R |sym| 50% Ex. re ae) tail N |sym]| 50% Ex. ue 52 | tail N | sym | 75% Ex. ee Be etal B, | sya!) (2599 ux. 54 | tail N | sym | - 57 | tail N | sym] 75% Ex. % 58 | tail R | sym | 100% Ex. 59 | tail N | sym | 100% Ex. 60 | tail N | sym 75% Ex. a 61 | tail N | sym ne 62 | tail N | sym sl. abs. 63 | tail R | sym 64 | tail N | sym | 100% Ex. 65 | tail 270° | sym 4 66, | h. tail Re Josymar tO a G7a|) tal N | sym i 69 | tail R | sym < 70 | h. tail | 270° | sym | 100% Ex. 71 | hz. tail R | sym | 100% Ex. 72.) b. tail N | sym 50% Ex. 7 ta R | sym | 100% Ex. e 75 | tail N | sym _ Col tail N | sym sl. abs “2 78 | tail N | sym 4 79 | tail N | sym =f Py 80 | tail N | sym sl. abs. 81 | tail N | sym nt 82 | tail 270° | sym a 83 | tail 90° | sym | 100% Ex. 2 84 | tail 270° | sym | 100% Ex. rr 85 | tail N | sym | 100% Ex. SKIN TRANSPLANTATION IN FROG TADPOLES 369 TABLE 1—Concluded : SYM. PROLIFERATION REM oece | OREIN-| | op AMOUNT OF BER DABION | orig ABSORPTION: laser ag artist] Ree pES TE? OES ae eS EP : Ant. only | Post. only | Both ends | Amorphic 8/2 h. tail 90° | asym] sl. abs. i 88 | h. tail | 270° | asym| 50% Ex. 89 | tail 270° | asym] 100% Ex. 90 | tail 270° | asym 8 91 | back N | asym 92 | tail 270° | asym] sl. abs. a 93 | tail 90° | asym . 94 | tail 90° | asym] sl. abs. 96 | tail 90° | asym 7: 97 | tail 270° | asym! 33% Ex. ¥ 98 | tail 270° | sym re 99 | back N sym 100 | back N | sym 101 | back R | sym 102 | back N | sym 103 | back R | sym 104 | back R | sym 105 | tail 270° | sym. | 100% Ex. . 106 | tail 270° | sym | 100% Ex. $F 107 | tail 270° | sym | 25% Ex. " 108 | tail 270° | sym sl. abs. ce 109 | back N | sym 110 | back R | sym Totals 64 | 55 tail 9 back Bes ric = Noon el i 1 The abbreviations used in this table are as follows: Sym., symmetrically placed over the eye Asym., asymmetrically placed over the eye % Ex., percentage of exposure of eye after absorption Sl. abs., slight absorption not reaching eye Ant. only, proliferation at originally anterior end of graft Post. only, proliferation at originally posterior end of graft N., graft placed in original orientation R:, graft placed in reversed orientation 90°, graft placed with originally anterior end ventral 270°, graft placed with originally anterior end dorsal h., homoiotransplant; all others were autotransplants 370 WILLIAM H. COLE of the fitting of the transplant to the wound. If the latter is exactly covered, the healing process, as already stated, is very short, complete union being established by the end of twelve hours. If, on the other hand, the transplant is too small or too large for the wound, the healing process is greatly prolonged and firm union may not be established until the end of thirty-six or forty-eight hours. A graft which is too small has a stretched appearance during this period, as though the epidermal cells, moving out over the wound, exerted a centrifugal pull on the whole mass. This favors firm and complete union of the trans- plant with the host integument along all its edges, provided the distances between the four sides of the graft and the host integu- ment are nearly equal. If one side is much farther away from the host integument than the opposite one, the former will be prevented from uniting because of being pulled away from the incision (fig. 1). Such a free edge will hasten the absorption process, which will be described in the next paragraph. A trans- plant of excessive size exhibits a turning under of its projecting edges. If all sides project, an increase in the convexity of the whole graft is produced during the healing period; if only one, the increase is limited to that side. If the overlap is not too great (less than about 2 mm.), the curved-in edges and the edges of the host integument unite. But since such grafts commonly are puckered at one place or another, union is rarely complete. In any event, a transplant too large for the wound is more convex than one of the right size or one somewhat too small and is less likely to be completely united. The first period, then, varies, its length being from twelve to forty-eight hours, at the end of which time the graft is attached to its host along the whole, or nearly all, of its periphery. During this time there is no external sign of vascular congestion, although sections show accumula- tions of blood cells in the dermis. (2) Second or adjustment period. Since all living tissues are continually being worn away or used up, the effete elements be- ing replaced through regeneration, it is to be expected that trans- planted tissue will undergo similar changes. If the replacement of tissue does not keep pace with wastage, then the transplant SKIN TRANSPLANTATION IN FROG TADPOLES 371 will disappear. But if wastage does not exceed replacement, the transplant will continue to exist. In many of the tail-skin grafts placed over the eye, there was a local disappearance of tissue; that is to say, in a certain area of the graft, the location varying in individual cases, wastage predominated over the for- mation of new cells. The result was the disappearance of tissue in that area. The word ‘absorption,’ as used in this paper, is defined as the disappearance of tissue in a small part of the graft, without any intimation as to the method of such disappearance. The word ‘adjustment,’ because of its broader meaning, has been selected to designate the period in which absorption takes place. During the adjustment period, a majority of the grafts, thirty- seven out of fifty-five, showed absorption in varying amounts—a visible proof that some kind of adjustment in the grafts was oc- curring. Although the other eighteen grafts did not show this sign, it is believed that they also passed through an adjusting process. The reasons for the non-appearance of absorption in those cases, as will be described later, are probably correlated with a firmer state of union between graft and host, or thicker tissue of the graft. The beginning of the absorption is seen first in those transplants which have one edge unattached. The free edge shrinks back from the incision, causing the contour of the edge to become con- vex to the center of the graft (figs. 2, 6, and 18). In symmetri- cally placed grafts, i.e., where the eye is beneath the center of the graft, the apex of the absorbed area is directed toward the eye- ball, no matter which edge was free in the beginning. Absorp- tion continues until the eye is partly or fully exposed to view (figs. 3, 4). The process is then checked, no further absorption taking place. The time at which evidence of absorption appears depends upon the amount of free edge in the beginning. ‘Trans- plants LN 18 and 75, for example, whose ventral edges were unattached, began to be absorbed during the third day after the operation, and on the tenth day the eyes were entirely exposed. When the length of the free edge is small or when only a corner of the graft is free, the absorption process is delayed. LN 60, with its anterior ventral corner free, illustrates such a condition. 372 WILLIAM H, COLE Absorption did not appear in that graft until the end of the fifth day after the operation. On the sixteenth day, about one-half of the eye was uncovered. Absorption then ceased. The grafts which had established complete union all along their edges showed absorption later than any of the others. In LN 44 ten days elapsed before the appearance of absorption. On the twentieth day after the operation the entire eye was exposed (fig.12). As will be seen from the figure, which is a photograph of the living animal, the graft is roughly crescent shaped. In that condition it re- mained until death on the one hundred and twenty-seventh day after the operation. The only change that occurred was a slight proliferation of tissue around the edges of the graft. Of the thirty-seven grafts which showed absorption, twenty-six had established complete union. In these twenty-six cases absorp- tion began on the average twelve days after the operation. In the other eleven grafts, each one of which had some free place along its edges, the average number of days preceding absorption was five. Complete union, therefore, delays the absorption proc- ess about one week. The delay is usually compensated, however, in the grafts with complete union by a greater rate of absorption when it does begin. Thus in LN 51, with complete union, the first signs of absorption appeared on the fourteenth day. At the end of the eighteenth day about one-half of the eye was exposed, and absorption then stopped (fig. 18). In this case, absorption continued only four days—a period shorter by nearly a week than that seen in other grafts where union was incomplete. There were a few grafts in which absorption began very late in the adjustment period and produced a very small U-shaped area not reaching the eye. It is supposed that in these cases the third period, one of proliferation, began before the absorption ~ had accomplished what it would have, if proliferation had not begun. Complete union may also cause a second type of adjustment, which was shown by six transplants, LN 64, 74, 83, 86, 105, and 106. In these grafts a circular area near the center gradually became thinner by absorption, and at last was perforated, thus exposing the eye. In none of them were there any signs of in- SKIN TRANSPLANTATION IN FROG TADPOLES aia fection or other abnormal conditions. ‘The behavior in all of them was similar, absorption appearing in the second week, and continuing, on an average, for seven days. When the process stopped, the opening was about the size of the eyeball (fig. 13). As a rule, then, absorption begins at any free place along an edge of the graft. When there is no free place, it starts at the weakest point on the line of union. When all edges are firmly and com- pletely united, then the center of the graft, which is at the point of greatest convexity and farthest away from the host, is the starting-point of absorption. The place where absorption begins is thus determined by the mechanical state of union between the transplant and the host. ~ It will be noticed from the foregoing description of the absorp- tion process, first, that the amount was by no means constant, varying from the maximum, which exposed the entire eye, to that which produced only a small U-shaped area at a free place without reaching the eye; and, secondly, that absorption usually began before the close of the second week and came to an end during the third week after the operation. ‘There were only two exceptions to this rule. LN 62 established complete union and thereafter up to the fiftieth day showed no changes. At that time it became loosened near the posterior ventral corner, and absorption proceeded anteriorly along the ventral edge. After a slight withdrawing of that edge, the process stopped and no further changes took place (fig. 6). In the other case, LN 88, with complete union, the middle of the anterior edge broke loose on the thirty-second day. Absorption continued slowly until the fortieth day, when one-half of the eye was exposed. ‘These two grafts were exceptions in another respect, since they never showed any proliferation of tissue. There comes a time, there- fore, in the history of the transplant, after which an adjustment by absorption does not occur, and this time, according to aver- ages, is about twenty-one days. It marks the close of the second period. (3) Third or proliferation period. During the adjustment period, the eighteen grafts which were not absorbed remained without visible changes. Union had been established along all edges. At about the beginning of the fourth week all transplants, 374 WILLIAM H. COLE excepting LN 62 and 88, with or without absorption history, began to proliferate. The formation of new tissue by the graft was never observed to occur during the adjustment period. It is assumed that a profound change takes place in the tissues of the graft which makes them incapable of further adjustment and initiates growth. Previous to such a change the connection of the blood vessels of the graft to those of the host has been com- pleted. It is likely, then, that with the normal blood supply restored, the transplant is able to form new tissue. This new activity, recognized by an increase in size of the graft and directly opposed to absorption seen in the second period, predominates during the third period. No further adjustment ever takes place. It is suggested by this fact that the absorption process in a certain part of the graft may be aided by a poorer blood supply in that region than in other regions. The originally anterior and posterior ends of the giaft show new tissue before the sides, and the amount produced at the ends is much greater than that at the sides. In many grafts, however, the proliferation was distributed irregularly over the surface, constituting amorphic regeneration. The limits of the old tissue are easily distinguishable from the new by the larger number of melanophores in the former—a condition which persists for months. When growth begins it proceeds rapidly for a time, and then almost suddenly ceases. The following abbreviated notes from the records of LN 79 illustrate the history of a typical case of proliferation: Oct. 4, 1920. Auto-tail-graft over right eye in original orientation. Oct. 5. Complete union along all edges. During the next two weeks no noticeable changes occurred. Oct. 20. Outgrowth along posterior edge of the graft. Oct. 22. Very slight proliferation along the two sides. Noticeable increase in the amount of new tissue at posterior end. Oct. 24. Posterior outgrowth is distinctly triangular in shape. No increase in lateral growth. rat 30. Posterior outgrowth appears like a normally regenerated tail tip. Noy. 3. No further increase in proliferation. Nov. 17. No changes. Jan. 9,1921. Nochanges; animal killed and graft fixed for sectioning. SKIN TRANSPLANTATION IN FROG TADPOLES one Proliferation in this case began on the sixteenth day after the operation, continued for ten days, and then ceased. Up to death, over two months after the last recorded change, no further growth took place. In other cases growth continued slowly for a month before stopping. The transplants which had shown no absorption did not differ from the others in regard to proliferation. Both kinds of grafts passed through the growth period, and at its close reached the final stage of equilibrium. This behavior was constant in all tail-skin grafts, many of which were kept under observation for as long as five months after the last noticeable change had oc- eurred. Morgulis (’09, p. 639) summarized the regeneration of the marine worm, Podarke, as follows: There is a lapse of some time, which varies with different individuals, and under different conditions, before new tissue is proliferated; this is followed suddenly by a period of rapid formation of new segments to be in turn followed soon by a period of slower regeneration. Finally the process is brought to a standstill. The histories of the proliferation of the tail-skin grafts on frog tadpoles and the regeneration of Podarke are thus seen to be similar. In the DN series (table 2) fourteen transplants of tail skin were made, and the animals kept in darkness. Four of them showed absorption, a lower percentage than that found in the LN series. In two cases the eye was one-half exposed; in the others the absorbed area just reached the eyeball. The adjustment went on in a manner similar to that seen in the light grafts. There- fore, it cannot be said that the absence of light was the direct cause of the smaller percentage of absorption cases. The dif- ference is correlated with the earlier appearance of proliferation. The average number of days at the end of which growth was first observed in the DN series was twelve. In the LN series the average was twenty-one. The adjustment period of the DN grafts was nine days shorter than the same period in the LN grafts. Further, the rate of growth was higher in the DN series, and was indicated by the greater irregularity of the new tissue formed in darkness. It is a well-known fact that in plants the 376 WILLIAM H. COLE absence of light results in a higher rate of growth. It is not sur- prising, then, to find that the same thing is true of animal tissue. Amorphic regeneration is typical of DN grafts. Only three of them showed tail-like regeneration in addition to the amorphic type. In the light series twenty-four out of fifty-five showed the amorphic and nineteen produced miniature tail tips. It seems that the absence of light hastens the adjustment period, allowing TABLE 2 The DN series NUM- orren-| 92™- AMOUNT OF a ie BER Jouniee TATION th. ABSORPTION Sil ft ae Selo. = Ant. only | Post. on'y | Both ends | Amorphic 2 tail R | sym . iM 3 tail N. | sym ss 4 tail N .| sym : 5 tail N |-sym | 50% Ex. ce 6 tail N -| sym ~ 7 tail i. |) siya “4 8 tail N | sym om 9 tail R | sym < 10 tail RB | sym $s 11 tail N | sym sl. abs. 12 tail R | sym | 50% Ex. % 13 tail N | sym 7 14 tail N | sym i 15 tail 270° | sym sl. abs. iW - 16 back R | sym 17 back N | sym Totals 16 14 tail 4 1 1 a 14 2 back proliferation to get an earlier start, and also increases the rate of proliferation (fig. 14). These two conditions tend to prevent adjustment in those grafts which are slow to show it, and to bring adjustment to a close, in those grafts which have begun to be absorbed, earlier than if they were in the light. . With these ex- ceptions, the histories of the LN and DN series were similar. SKIN TRANSPLANTATION IN FROG TADPOLES ae b. Back skin. The second group of transplantations over the eyes consisted of grafts of integument taken from the back region of the same animal or from another animal. A square of skin from the middle of the back was cut out and grafted over the eye in the same way as with tail-skin grafts. The histories of such grafts differ very markedly from the histories of tail-skin grafts, since the former never show absorption or proliferation. The healing period is essentially like that described for tail-skin grafts, but the subsequent behavior varies according to the source of the tissue. Autotransplants remain unchanged and enter the period of equilibrium after about four days. The line of union is marked by a denser mass of melanophores than elsewhere for several weeks afterward. But in time this condition disappears and the graft becomes indistinguishable from the surrounding skin (fig.17). Vision is thus permanently inhibited, since back skin is thoroughly opaque. Neither increase in size nor any other indication of growth is shown by such grafts, some of which were under ob- servation for five months. Homoiotransplants of back skin over an eye are not usually successful. When placed flat on other parts of the body, they unite readily, but the healing period is prolonged. This is probably due to the action of the homoio- toxin upon the protoplasm of the graft, which has been shown by Loeb to delay healing of grafts. When the graft is placed over an eye the tissue is forced into a curved position by the convexity of the eye. Because of the delayed healing and the curvature of the new position, homoiotransplants of back skin are not able to establish good union. Usually only one edge succeeds in becom- ing attached, so that nearly all of the graft soon dies. But the homoiotransplants which did unite completely had _ histories similar to those of the autotransplants; they showed neither absorption nor proliferation. The striking difference in behavior between tail-skin and back- skin grafts may be due to the physical difference in structure. Tail skin is mostly epidermis, with a very thin dermal layer, while back skin has a well-developed dermis with many glands and a dense fibrous layer. Therefore, back skin is thicker, more com- pact and much less plastic than tail skin—facts that may well 378 WILLIAM H. COLE explain its indifference to any stimuli tending to change its size and shape. 2. Grafts not over eyes (series LB) or over operated eyes (series LO) The outcome of the back-skin transplants contributed only negative evidence to the question of a relation between the visual function and the absorption process, because the difference in behavior between the tail-skin and back-skin grafts seemed to be due to the difference in structure. It was then planned to make grafts of tail skin on the back region of tadpoles, not over the eye, to determine whether absorption would ever occur in such cases. These grafts constituted a part of the LB series (table 4). The operation was like that in the other series. The tail skin was fitted to the incision and grafted in place (fig. 15). The orientation was varied and some of the animals were kept in darkness. None of the twenty-two grafts so made were absorbed. They all established complete union along all edges and at the end of two weeks they began to proliferate new tissue. From the anterior end only, growth occurred in three cases; from the posterior end only, in one case, and from both ends equally, in eight cases. Sixteen cases showed amorphic regeneration. Five weeks after the operation, on an average, these LB grafts had reached the period of equilibrium. Thereafter the tissue remained unchanged. ‘There is no doubt that the tail-skin grafts placed on the body region not over an eye differ entirely in their behavior from those placed over an eye. The former never show absorption; a majority of the latter do. When the operation was varied by using back skin, the results were the same as far as the absence of absorption is concerned. Autotransplants and homoiotransplants of tail or back skin on the back region are never absorbed, and only the tail-skin grafts proliferate. ‘These results indicated definitely that the eye in some way is responsible for the absorption of the tail-skin grafts. Whether the relation between the organ and the adjustment of the graft was due to the function or the structure of the eye was still undetermined. It seemed as though the severance of the optic nerve before the graft was placed over the eye would afford evidence on this ques- SKIN TRANSPLANTATION IN FROG TADPOLES 379 tion. Tail-skin grafts were, therefore, placed over eyes whose optic nerves had been cut, and also over empty sockets after the removal of the eyes (the LO series, table 3). In doing the former operation a small piece of the optic nerve, about 1 mm. long was removed, so as to prevent regeneration. The results of these grafts may be briefly stated. No absorp- tion was seen in any of the LO grafts. The explanation is as follows. After the severance of the optic nerve, the eye rapidly TABLE 3 The LO series PROLIFERATION NUMBER] SOURCE OPERATION pe ae ee ae es a on ree Se Ant. only | Post. only | Both ends | Amorphiec 1 tail op. n. cut none ei 2, tail op. n. cut none a 3 tail op. n. cut none < ie 4 tail op. n. cut none gs 5 tail op. n. cut none e he 6 tail eye removed | none ri 7 tail eye removed | none ts se 8 tail eye removed | none 4 9 tail eye removed | none x 10 tail eye removed | none o 11 back | eye removed | none 12 back op. n. cut none 13 back op. n. cut none Totals 13 10 tail |7op. n. cut] none 3 back 3 0 2 8 6 eye removed degenerates. At the end of several weeks, postmortem examina- tion showed only the remains of the whitish opaque lens. The operation and the degeneration of the eye which follows greatly lessen and often obliterate the curvature of the conjunctiva, so that grafts placed over such eyes are not under the same condi- tions as grafts over normal eyes. Those over empty sockets are either flat or slightly concave on their outer surface. The absence of curvature or the destruction of vision in the LO series may have been the reason why absorption did not take place. All 380 WILLIAM H. COLE TABLE 4 The LB series PROLIFERATION NUMBER| SOURCE |SKIN BENEATH ee aie > ee Ant. only | Post. only | Both ends | Amorphic 1 tail “ none es 2 tail ce none es 3 tail or none &6 #6 4 tail “ none 6 tail ad none ad 7 tail ee none s 8 tail se none cs He 9 tail es none s ili tail ef none 3 12 tail ss none ie 13 tail ef none cs 6 14 tail ee none - 115 tail ee none s 16 tail & none es 17 tail af none % 18 tail s none « 19 tail none : 4: 20 tail es none “ Dill tail s none ‘ 22 tail ee none ce o 23 tail oe none ‘ 24 tail é none f Zo h. back ss none 26 h. back a6 none 27 h. back ee none 28 h. back re none 29 h. back ee none 30 h. back se none 31 h. belly 5 none 32 h. belly as none 34 h. belly cs none 35 h. belly cs none 36 h. belly cs none 37 belly $ none 38 belly a none 39 belly none 40 belly as none 41 belly none 42 belly pr none 43 belly os none SKIN TRANSPLANTATION IN FROG TADPOLES 381 TABLE 4—Concluded PROLIFERATION NUMBER| SOURCE |SKIN BENEATH sca rau eh AS SE en Ant. only | Post. only | Both ends | Amorphic 44 belly a none 45 belly e none 46 tail Skin and} none < glass be- = neath “ 47 tail <6 none 48 tail.. i gl. out 49 tail ss gl. out ** 50 tail e none et 51 tail ¢ none 52 tail s§ gl. out 53 tail sS none ce 54 tail ee none e 55 tail id gl. out o 56 back se none 57 back ss none 58 back « none 59 back id none 60 back s none 61 back sé none 62 back ! 5 " fy b # > rs 4 ‘i a ald TT aon? 7 y ry ate PK is my, WT) g aD i a) Sai an P btay? Os Bad 8 ; WBS Be Sioud . 32. re = lawaey sal asPORaLg TUES Pt: - — “4 aS y 411 PLATE 1 EXPLANATION OF FIGURES 1to4 Drawings of graft LN 18 at successive periods in the adjustment proc- ess. s, scar in host integument where incision was made. The upper edge of each figure represents the anterior end of the graft. X 7. 1 Graft LN 18 as it appeared on the second day after the operation. 2 Same graft one week after operation. 3 Same graft two weeks after operation. 4 Same graft six weeks after operation, showing the maximum amount of proliferated tissue that was formed. 5 Graft LO 5 four months after the operation. Proliferation from both ante- rior and posterior ends is shown, as well as amorphic regeneration over the surface of the graft. The anterior outgrowth is cylindrical and twisted back on itself, while the posterior one is tail-like. No absorption occurred. X 7. 6 Graft LN 62 four and one-half months after the operation. The ventral edge is unattached and has shrunken back toward the eye*-a condition which persisted. X 6. 412 SKIN TRANSPLANTATION IN FROG TADPOLES PLATE 1 WILLIAM H. COLE PLATE 2 EXPLANATION OF FIGURES 7 A transverse section of graft LN 79, illustrating amorphic regeneration. Only grafted tissue is shown. The notochord (n) is in two parts, the larger being the old notochord and the smaller being the regenerated notochord. The nerve cord is seen just beneath the notochord. Due to the irregularity of the proliferated tissue, there are several spaces lined by epidermis, all of which open to the outside, however. X 34. 8 Transverse section of graft LB 10 three days after the operation. Beneath the graft is seen the integument of the host (hi) with many blood cells in the der- mis. Blood cells have also penetrated the notochord (nm) of the graft. Subse- quent behavior of such autotransplants results in the lining of the space between graft and host by epidermis, with an opening to the outside somewhere. X 513. 9 Transverse section of graft LO 2 five weeks after the operation. It will be noticed that the dorsal side of the graft (lower edge in figure) has become con- tinuous with the skin of the host, due to establishment of union distally. The other side has united proximally. X 38. 10 Transverse section of graft LO 2 near the posterior edge, showing the point of union on the dorsal side. The epidermis on the under side of the graft and the epidermis of the host, covered by the graft, have become continuous, thus lining the cavity. X 45. 11 Tangential section, nearly longitudinal, of graft LN 78, showing boundary between old grafted tissue and the new proliferated tissue. mn, notochord of old tissue; rn, regenerated notochord; nc, regenerated nerve cord. X 973. 414 PLATE 2 SKIN TRANSPLANTATION IN FROG TADPOLES WILLIAM H. COLE - pRSste ee 5 ia s —_ = aS pada Ate wee ep ieett De audi 415 THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 35, NO. 4 PLATE 3 EXPLANATION OF FIGURES 12 Photograph of graft LN 44, taken twenty-six days after the operation. The only change up to death, on the 127th day, was a slight increase of prolifer- ated tissue at the ends. X 3. 13 Photograph of graft LN 64, taken eleven days after the operation. The darker region surrounding the central opening is due to an abundance of melano- phores there. X 3. 14 Photograph of graft DN 2, taken fourteen days after the operation. Proliferation has begun and is easily seen at the corners of the graft. The irreg- ularity of the proliferated tissue is noticeable. X 3. 15 Photograph of graft LB 9, taken four months after the operation. Slight amorphic regeneration is shown. No absorption occurred and the characteristic features of tail skin were preserved. X 3. 16 Photograph of graft LN 78, taken three months after the operation. Pro- liferation from both ends, absence of absorption, and preservation of tail-skin characteristics are the features of this graft. Refer to figurell. X83. 17 Photograph of graft LN 104 taken three months after the operation. The back-skin graft over the eye has become almost indistinguishable from the sur- rounding skin, but the tail-skin graft has proliferated and preserved its charac- teristics. X 3. 416 SKIN TRANSPLANTATION IN FROG TADPOLES PLATE 3 WILLIAM H. COLE 417 PLATE 4 EXPLANATION OF FIGURES 18 Photograph of LN 51 taken sixteen days after the operation. Absorption began along the ventral edge. Two days after this photograph was made, about one-half of the eye was exposed ina dorsal view. No further absorption occurred. xX 3. 19 Photomicrograph of dermal melanophores of an animal which had been kept in ice-water (0°C.) for four days. An intricate network of melanophore processes entirely obliterates the limits of individual cells. X 2938. 20 Photomicrograph of dermal melanophores of an animal which had been kept in ice-water (0°C.) for two hours. The cells, which at the beginning of the experiment were completely contracted (fig. 21), are shown to be about one- third expanded. X 298. 21 Photomicrograph of dermal melanophores of an animal which had been kept in warm water (35°C.) for two days. The cells are maximally contracted. X 293. 22 Photomicrograph of epidermal melanophores of graft LB 34, showing the parallel arrangement of the melanophores and masses of melanin granules. X 320. 418 PLATE 4 SKIN TRANSPLANTATION IN FROG TADPOLES WILLIAM H. COLE 2 | 419 Resumen por el autor, Leon 8. Stone. Experimentos sobre el desarrollo de los ganglios craneales y el sistema de 6rganos sensoriales de la linea lateral en Amblystoma punctatum. Los experimentos llevados a cabo por el autor han consistido en la extirpacién de las placodas y la cresta neural. La cresta neural se origina en el tubo neural, en el margen dorsal de la fusion de los pliegues neurales; en el desarrollo normal, todas estas células, con la excepcidn de una pequefia parte de ellas, se transforman en el ‘‘mesectodermo”’ emigrante que se extiende ventralmente sobre el mesodermo de los arcos viscerales, en- volviéndolos, y finalmente se sittia en sus superficies medias, en las cuales forma los cartilagos del esqueleto visceral. La extirpacioén de las células de la cresta produce ausencia o defectos en la mandibula, cuadrado, porciones anteriores de las trabéculas, y en todos los cartilagos branquiales, con la ex- cepcidn del segundo basibranquial. Este ultimo se origina a expensas del mesodermo situado a lo largo de la pared anterior de la camara pericirdica. La ausencia de la cresta neural en la region branquial est’ acompafnada siempre de pequenas bran- quias externas, que contienen una porcién subnormal de tejido conectivo. La extirpacién de las placodas epibranquiales de los nervios VII, IX y X produce ausencia de los ganglios y ner- vios viscerales especiales. El] sistema cutdneo general deriva en gran parte, si no enteramente, de las placodas, mientras que el sistema visceral general deriva aparentemente de la cresta neural. Cuando se extirpan regiones del ectodermo anteriores y posteriores a la placoda auditiva en el momento de cerrarse los pliegues neurales, la linea del cuerpo, los primordios occipital y supraorbital y sus ganglios de la linea lateral correspondientes faltan. Todos los grupos de los 6érganos de la linea lateral pare- cen originarse a expensas de primordios separados. Translation by José F. Nonidez Cornell Medical College, New York AUTHOR’S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, MARCH 27 EXPERIMENTS ON THE DEVELOPMENT OF THE CRA- NIAL GANGLIA AND THE LATERAL LINE SENSE ORGANS IN AMBLYSTOMA PUNCTATUM}! L. 8. STONE Osborn Zoélogical Laboratory, Yale University NINETY FIGURES CONTENTS UT Econ aa tiga Mee 2a ho ee ee Le CAL oe OR PRA Sm 421 Material, methods, and normal development..................00000000eeeee 427 Dre pee sa ome ee Ee ek AN kB ohh cc hg ENS ag sisvaie mass, «dapat? 457 LE ORI TI SE AAIEN CL SPI COWES ta: Kates. colety co.) 2ya!siae-aih ve cede SHE Gade Ade we 44 ga ahe, es 459 ipivemoval of Ophthalmic placode. )..8/5) 6.0066. Js es oe. 459 Aaavemovaliol eagssrian placode: 565.0) L)eianceal eo ga. Meee See 462 3. Removal of the preauditory placode and the supra-orbital pri- TEL CLUSUREELE Raz vx AMEN S's bo ult Rede eee eS o Ay ae ie kee fas OS oe eee 464 4. Removal of infra-orbital, hyomandibular, ventral hyomandibular, ardimaandibilar primordiacs (eco Wa Me. SAL A edn wk 467 5. Removal of the epibranchial placode of VII and surrounding SE OLE 6 RE eS Re SEROUS oN hi. a 91 eee 467 6. Removal of epibranchial placodes of IX and X and postauditory Paterd-lne primordia th) 80.) 8e0, OUT SEO, N's Bs 471 Peehemosalietoneural) crestiyiacp).22 hess. cAteswrd oh biky. Gebidet plas Oo. week 474 i Coutripution to. mesodenmal tissue. <....222.2.0 «2406 66 oyys sides. « oo 474 2. Contributions to ganglionic components..................0eeeeeees 483 a. Centripution to Vandy VEL. oe ee ae Pa icant 483 be, Contmibntion) to Mxeant irs S res ics) sips bob eri kien (Sines oe ae 485 “oS EES TAL, Sy ae ee 2 ee a SR en ae 487 SY UNTILITTS GIES eee Band oy ata eR oe ae ah oe I OR 493 PePCER EME TCILEME ert A Te nek eR Ome et TTR Eros it oae he tees 495 INTRODUCTION The experiments set forth in this paper were undertaken for the purpose of extending our knowledge concerning the part which placodes and neural crest play in the formation of cranial ganglia and nerves, and also to determine, if possible, the extent of their 1 Read before the American Association of Anatomists, March 25, 1921. 421 422 L. S. STONE contribution to the formation of the mesoderm and the exact origin and fate of the ‘mesectoderm’ tissue. Experiments have never before been applied to the solution of these problems, all of our knowledge having been obtained from studies made upon successive developmental stages of normal embryos. In the present study of the early stages of development it has been found that cells are given off for the formation of certain ganglia by epibranchial, lateral-line, gasserian, and ophthalmic placodes. From the early appearance of the neural crest in the dorsal portion of the neural canal at the time of the closure of the neural folds the crest cells have been followed as they migrate over the mesoderm of the visceral arches, increase in number and finally arrange themselves upon the median and ventral sides of this mesoderm, where they form the cartilages of the visceral skeleton except the second basibranchial. The distinctness of the early crest-cell groups and placodes makes it possible to re- move these structures by employing the operative methods now so commonly used by investigators upon amphibian embryos. It is also possible to remove areas of ectoderm from early embryos in which the placodes have not yet become prominent enough to be identified, and thus to ascertain how early these placodes in the ectoderm are laid down as definite entities. Much confusion arose in the earlier descriptions of placodes by the failure to distinguish clearly between ectodermal thicken- ings in relation to the gill clefts and those related to the forma- tion of the lateral-line system. Not until von Kuppfer (91) clearly pointed out the distinction between dorsolateral and ven- trolateral, or epibranchial placodes, was there hope of obtaining light on the formation of the components of cranial ganglia. De- scriptions of cells breaking off en masse from the epibranchial placodes and being added to portions of the cerebral ganglia have repeatedly appeared in the anatomical studies of all types of vertebrates, especially among the fishes and amphibians (van Wijhe, ’82; Beard, ’85; Froriep, ’85; Platt, ’96, and Landacre, 10 and ’12). Landacre has assigned a definite function to these epibranchial placodes. He has shown in Ameiurus (710) that the special visceral portion of the [IX ganglion appears to come from CRANIAL GANGLIA OF AMBLYSTOMA 423 the epibranchial placode of the first true gill. Since Herrick (07)? had found only gustatory fibers arising from the visceral ganglion of IX, Landacre naturally concludes that the epibran- chial placode of IX gives rise to the ganglionic cells from which the gustatory fibers arise. He further expresses the opinion that this is the function of the epibranchial placodes in all types of vertebrates (Landacre, 712), since these placodes occur only on those nerves of VII, IX, and X which contain special visceral fibers, and since the relative size and growth of the epibranchial placodes in Lepidosteus and Ameiurus seem to be in relation to the area and time of appearance of taste buds supplied by fibers from special visceral ganglia. Coghill (716) finds conditions in Amblystoma which seem to indicate that the visceral ganglia of VII, IX, and X all receive masses of cells from the epibranchial placodes. There is also evidence that there are other placodes besides those concerned in the formation of lateral-line ganglia, L.e., general cutaneous placodes. Platt (’96) notes in Necturus that an ectodermal thickening above the eye appears to be concerned in some manner with the formation of the ophthalmicus pro- fundus V nerve. A similar fact has been recorded by Coghill (’16) in Amblystoma and by Landacre (’12) in Lepidosteus, where in early stages of the formation of the ganglion of the ophthalmicus profundus V there is a distinct anchorage to the ectoderm over the eye. The former made the interesting observation that it was during this period, in which the anchorage is intact, that root fibers make their connection with the brain. However, neither of the latter investigators satisfied himself as to the significance of this early contact. Coghill (16) also found in the earliest of his stages described, the ‘non-motile,’ an area of adhesion between the distal end of the gasserian ganglion and the skin. It is ven- tral to the primordium of the preauditory lateral-line organs and is not so extensive nor so intimate as the adhesion of the oph- thalmic ganglion and skin. 2A note included in Landacre’s paper (’07) by C. J. Herrick on the distribution of the [IX nerve of Ameiurusmelas. The conclusions from this reference are found in Landacre’s paper (12), page 3. 424 L. S. STONE In connection with the general cutaneous component of X, Coghill (16) finds in the ‘non-motile’ stage an intimate relation between an ectodermal thickening and a loose aggregation of cells upon the lateral aspect of the lateral-line ganglion of X. This cluster of cells, which he identifies as representing the jugular ganglion, is slightly more condensed in its most rostral portion where an incipient root is forming, which does not, however, at this stage reach the brain. Lateral-line primordia. Platt (96), Landacre (10), Coghill (’16), Goette (’14), and others have shown that a different system of placodes is concerned in the formation of the lateral-line gan- glia of VII, IX, and X. Not all of the investigators agree as to the exclusive derivation of these ganglia from placodes, for some have expressed the belief that neural crest also enters into their formation. The primordia of the lateral-line system have been shown very clearly by Platt (96), Landacre (’10), and Landacre and Conger (13) to have origins independent of the auditory vesicle. How- ever, anterior and posterior prolongations of the auditory thick- ening have been observed in many forms, and to these Landacre (10) applies the term of ‘preauditory and postauditory’ placodes. He does not have a clear idea as to the relation of these placodes to the sensory lines, for in the case of the preauditory placode (Landacre and Conger, ’13) it seemed to disappear by a process of degeneration before the appearance of the lateral-line primordia, and in the case of the postauditory placode (710) it loses the char- acteristic cell arrangement and does not give rise to the lateral- line organs. Moreover, in Lepidosteus Landacre and Conger (18) fail to find any postauditory placode. Although separate origins have been given to each of the groups of the lateral-line sense organs by various investigators, none have described such an early pattern of the lateral-line system as the one observed in Necturus. According to Platt (96, p. 491), the plan of this system is early laid down in three longitu- dinal lines on each side of the embryo, connected by intersegmen- — tal cross-lines with special differentiations at points of intersec- tion, and out of this pattern certain portions are retained in the CRANIAL GANGLIA OF AMBLYSTOMA 425 final system. However, when one compares this description with skin amounts of early stages of Amblystoma, another interpreta- tion seems to be more plausible. No such pattern can be found in Amblystoma embryos, but whenever the contour of underlying structures makes its appearance on the surface certain transitory thickenings appear due to mechanical molding, as one finds be- tween somites, gill swellings, around the early optic vesicles and the early limb bud. It seems possible, therefore, that most of the early pattern described in Necturus has no significance in the formation of the lateral-line system. Neural crest. It is generally accepted that portions of the cranial ganglia are derived from the neural crest; in fact, early investigators of cranial nerve problems made the crest cells their sole source of origin. Landacre (’10) would derive all general cutaneous ganglia and the general visceral system exclusively from neural crest. In view of the fact that there have appeared descriptions of extensive wanderings of neural crest, it seems strange that so little emphasis has been put upon such an outstanding feature. The literature contains comparatively few descriptions of any complete investigations of the growth of the neural crest outside of its supposed connection with the formation of cranial ganglia. In most of the descriptions of the early stages of developing cranial ganglia the assumption that the crest cells were concerned only with the formation of ganglia and nerves has been so general that many investigators have been led to overlook the fact that any further wandering of these elements occurs. However, various careers have been assigned to them by a few investigators. Marshall (’78), working on the chick, and van Wijhe (’82), on the selachians, were among the first to observe that the anterior part of the neural crest, which could not be identified with the formation of ganglia, gradually disappeared. They could not determine its final fate. Kastschenko (’88) went a step farther in suggesting that the cells which were lost from the neural crest did not degenerate, but added themselves to mesenchyme. Addi- tional descriptions of the neural crest then followed, by Gorono- witsch (’93) in birds and Platt (’96, ’97) in Necturus. Both 426 L. §. STONE agreed that a greater part of the neural crest, augmented at the margin of the gill clefts by a proliferation from the lateral ecto- derm, becomes a wandering mass of ectodermal cells to which Platt gave the name ‘mesectoderm’ and that this wandering mass of ectodermal cells is the origin of certain mesenchymal tissue. It was Platt, however, who suggested that the branchial cartilages, surrounding connective tissue, and the anterior portions of the trabeculae were derived from the ‘mesectoderm.’ In re- spect to the branchial cartilages, this agrees with von Kuppfer (95) on Petromyzon, although he claims that the wandering ecto- dermal cells are derived from deeper layers of ectoderm in situ. Furthermore, Dohrn (’02), working on Torpedo, and Brauer (04), on the Gymnophiona, agree in the main with Platt’s description of the mode of formation of branchial cartilages, but they ascribe the sole origin of the cartilages to the neural crest. Recently Goette (’14) has discussed the formation and develop- ment of the ‘ectomesoderm,’ as he calls it, in Siredon (Amblys- toma tigrinum) which also illustrates what he finds in Petromy- zon and Torpedo. He confines the anlagen of the ‘ectomesoderm’ to the ectoderm in the regions of V, VII, IX, and X, which he claims forms visceral skeleton, the outer gill muscles, and the surrounding connective tissue. To the neural crest, epibranchial placodes, and lateral-line placodes belongs the function of the formation of the cranial ganglia and nerves. From this summary of the literature it is evident that there is much disagreement concerning the formation of cranial ganglia and the origin and further distribution of the wandering masses of cells of ectodermal origin. In the light of these facts, an investigation of this problem was suggested by Prof. R. G. Harrison on an animal, such as Amblys- toma punctatum, which would lend itself to an experimental analysis of all the factors involved. It gives me pleasure to express here my appreciation to Doctor Harrison for his kind criticisms and suggestions during its progress. Since the completion of the present investigation, there has appeared a paper by Landacre (’21) on the fate of the neural crest in Plethodon glutinosus, a urodele. In so far as the visceral CRANIAL GANGLIA OF AMBLYSTOMA ADT cartilages are concerned, the morphological findings in this paper correspond very closely to the observations recorded in the pres- ent study of Amblystoma. In Plethodon the manner of the migration and disposition of the neural crest upon the branchial arches is similar to that described by Platt (97). In addition to contributing to the ganglionic portions of V, VII, IX, and X, the neural crest is described as forming the greater portion of the mesenchyme in the ventral part of the head and in the branchial region, the anterior portions of the trabeculae, Meckel’s cartilage, the palatoquadrate bar, and all the branchial cartilages except the second basibranchial. MATERIAL, METHODS, AND NORMAL DEVELOPMENT All the experiments have been made upon embryos of Amblys- toma punctatum in stages ranging from 21 to 27 as shown in figures 1 to 4. The technique employed in the operation is similar to that already described by Harrison (’18). The special types of operation employed in this investigation will be subse- quently described under separate sections. In order that the experiments may be more clearly interpreted, a description of the stages in the normal development of the neural crest and of certain placodes will first be given. This description was obtained chiefly from a series of dissections sup- ported in the more minute details by serial sections. In order to accomplish the dissections, embryos which had been preserved in the ordinary corrosive sublimate acetic mixture were placed in a 5 per cent aqueous solution of nitric acid, where they were kept from twelve to twenty-four hours for the purpose of softening the brittle ectoderm and at the same time rendering it pliable. They were then placed under the binocular microscope, and by means of a small pair of operating scissors and suitable needles a con- tinuous incision was made up the middorsal line and down the midventral line. The two halves of ectoderm were carefully removed, stained lightly with haematoxylin, cleared in the usual manner, and mounted in damar on glass slides. After such treatment the positions of ectodermal thickenings can be observed accurately. 428 L. S. STONE The dissected embryos were in some cases preserved in glycerin after they had been stained with haematoxylin. However, the neural crest may be more satisfactorily observed unstained in water or alcohol, because its grayish-brown color sufficiently differentiates it from the surrounding tissue up to a certain stage as will be described later. Although the picture of the crest cells and placodes is approxi- mately constant for each stage, occasionally normal embryos whose external appearances indicate similar age disclose upon dis- section noticeable differences in the early rate of growth of the crest cells or placodes. This difference may even be confined to the same individual where, for example, the advance of the trunk lateral-line primordia of the one side lags conspicuously behind the primordia of the other side for a distance of one or two somites. Stage 21 Neural crest. A layer of neural crest, light brown in color, over the dorsal and dorsolateral portions of the medullary tube extends from a point above the eye to a point above the posterior border of the second somite (fig. 1) where it is continuous with the crest cells in the spinal region. Ventrally there are two distinct proliferations separated by a constriction lying approximately midway between the two ends. The more anterior proliferation becomes the band which later wraps around the mesoderm of the hyoid arch. Where the dorsal margin of the mesoderm of the branchial region meets the lateral surface of the medullary tube there is a trough-like depression, filled in by a deep longitudinal ridge of ectoderm extending from above the primitive hyomandibular cleft to near the first somite, where it becomes continuous with less marked ridges over the somites. This is represented by dotted lines in the figure. Figure 13 is a frontal section passing through the longitudinal ectodermal ridge and the anterior and posterior extremities of the layer of neural crest. The considerable thickness of the ectodermal ridge can be seen on either side in the region of the CRANIAL GANGLIA OF AMBLYSTOMA 429 hind-brain. A cap of crest cells, whose base extends as a wedge between the fused neural folds, lies in the region of the mid-brain. These cells are represented by well-formed nuclei surrounded by a pigmented cytoplasm in which lie fine, closely packed yolk eranules which are similar in size to those of the neural canal. The mass of cells lies close to the ectoderm and extends laterally a short distance over the walls of the mid-brain. Over the region of the hind-brain a similar small cap can be seen lying in the same relation to the ectoderm and brain, except that its base does not appear in this plane of section to be wedged between the fused folds. Placodes. Figure 1 shows an area over the dorsal border of the eye, which represents the position of a very slight thickening of the ectoderm where cells are given off later to the ophthalmic ganglion. It may be said here that throughout the different stages studied, this ectodermal thickening never acquires a thickness comparable to that of other placodes. Many times in the dis- sections the area was thinner than the surrounding ectoderm because of the fact that its cells adhered to the loose ganglionic mass of cells forming the ophthalmic ganglion. This thickening may be called the ophthalmic placode. Below the anterior extremity of the longitudinal ridge les a small ectodermal thickening in the dorsal extremity of the hyo- mandibular cleft, which can be detected at this stage. This small ectodermal thickening takes on the contour of the cleft at this point, giving it approximately a crescent shape. In the position of the anterior portion of this thickening there later arises the primordium of the infra-orbital group of lateral-line sense organs, while in the position of the posterior portion there later arises the hyomandibular group. Stage 23 Neural crest. This stage shows a decided ventral growth of the neural crest over the lateral walls of the brain (fig. 2). Above and posterior to the dorsal border of the eye an extensive pro- liferation has taken piace, carrying the neural crest downward toward the primitive mandibular arch. Figure 14 shows above 430 L. S. STONE ABBREVIATIONS alv., alveolaris VII nerve an., angular group of sense organs au., ear buc., ramus buccalis VII br., brain bv., blood vessel c. c., neural-crest cells ch., notochord chy., ceratohyal dl., dorsal body line of sense organs €., eye ect., ectoderm epi. VII., epibranchial vlacode of VII epi. 1X., epibranchial placode of IX epi. X., epibranchial placode of X ex. g., external gill gas. g., gasserian ganglion gas. pl., gasserian placode jg., jugularis VII nerve hh., hyohyal hm., hyomandibular group of sense organs hmd. th., ectodermal thickening in dorsal hyomandibular cleft hy. a., mesoderm of hyoid arch hy. ce., hyoid crest-cell group hy. m., hyoid muscles io., infra-orbital group of sense organs l.e.t., longitudinal ectodermal thicken- ing mb., mandible md., mandibular group of sense organs md. a., mesoderm of the mandibular arch md. cc., mandibular group of crest cells md. m., mandibular muscle md. V., truncus mandibularis V ment., mentalis VII nerve m.1., midbody line of sense organs ms., mesoderm na., nose 0., occipital group of sense organs op. v., optic vesicle oph. g., ophthalmic ganglion oph. pl., ophthalmic placode oph. p. V., ophthalmicus profundus V nerve ph., pharynx pl., palatinus VII nerve q., quadrate s., somite so., Supra-orbital group of sense organs sup. VII, ramus ophthalmicus super- ficialis VII th. m., thoracicohyoideus muscle imp. m., temporalis muscle tr., trabecula v. hm., ventral hyomandibular group of sense organs vl., ventral body line of sense organs 1-2 bb., first and second basibranchial 1 bb. cc., first basibranchial crest- cell group 1-4 br. a., mesoderm of first to fourth brancbial arches 1-4 br. cc., first to fourth branchial crest-cell groups 1-2 cbr., first and second ceratobran- chial 1-4 ebr., first to fourth epibranchial 1 s., first somite ; 1 int. r., first intersomitic ridge VII g., facial ganglion VII lL. g., facial lateral-line ganglion VII pl., facial lateral-line placode IX lg., glossopharyngeus lateral-line ganglion IX pl., glossopharyngeus lateral-line placode IX vis. g., glossopharyngeus visceral ganglion X lg., vagus lateral-line ganglion X pl., vagus lateral-line placode X vis. g., vagus visceral ganglion X vis. t., vagus visceral trunk at Me “p epilk vim thd fim 8 Figs. 1 to 8 Camera-lucida drawings of stages 21, 23, 25, 26-27, 30, 33, 39, and 35+ of embryos of Amblystoma punctatum, showing development of neural crest, ectodermal placodes, and primordia of lateral-line sense organs. N eural crest is stippled and placodes and primordia are solid black. X 10. For ab- breviations see page 430. 431 THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 35, NO. 4 mul dl o auhm so 12 Figs. 9 and 10 Camera drawings of stages 36 and 37-38, showing further de- velopment of lateral-line primordia. X 10. Figs. 11 and 12 Camera drawings of lateral and ventral views, respectively, of stage 46+, showing final distribution of lateral-line groups of sense organs. The gills have been removed. X10. For abbreviations see page 430. 432 CRANIAL GANGLIA OF AMBLYSTOMA 433 the longitudinal ectodermal ridge small portions of the neural crest descending upon the lateral walls of the hind-brain con- nected dorsally by a cap of cells still wedged tightly between the ectoderm and brain. Posterior to the mandibular arch are the two prolongations of the neural crest which appeared in stage 21. Approaching the 15 Fig. 13 Frontal section passing through longitudinal ectodermal ridge and anterior and posterior extremities of layer of neural crest in stage 21, showing origin of crest cells from dorsal and median portions of neuralfolds. X 37. Fig. 14 Cross-section passing through longitudinal ectodermal ridge and ectodermal thickening in dorsal portion of hyomandibular cleft in stage 23, showing neural crest over dorsal portion of neural tube and portions of neural crest descending to lateral walls of neural tube. X 37. Fig. 15 Cross-section through upper half of optic vesicles in stage 25, showing crest cells descending along sides of neural tube and over mesoderm of mandibular arches and over anterior portion of right optic vesicle. X 37. first somite is an additional proliferation which later covers the mesoderm of the second, third, and fourth branchial arches. In this stage the neural crest has extended posteriorly (fig. 2) to a position above the anterior half of the fourth somite. Placodes. ‘The ophthalmic placode lies in relatively the same position as in stage 21. Its indefinite borders outline a slightly 434 L. S. STONE thickened portion of the ectoderm which is in close contact with the crest cells at its posterior border. The longitudinal ectoder- mal ridge (fig. 2) is still prominent, but extends slightly more an- teriorly and less posteriorly than in the previous stage. The thickening in the dorsal portion of the hyomandibular cleft is slightly larger, but still remains crescentic in shape (fig. 14). The approximate positions of two placodes, anterior and posterior to the probable position of the auditory placode, are indicated by dotted lines. That which lies just anterior to the auditory placode gives rise to part of the lateral-line ganglion of VII. That which lies posterior to the auditory placode gives rise to part of the lateral-line ganglion of X. The positions of other placodes cannot be exactly located in sections of this stage by the appearance of any ectodermal thickenings. Stage 25 Neural crest. A further very rapid ventral growth of the neural crest has taken place, the most pronounced proliferation involving those cells which are in the region over the dorsal border of the eye (fig. 3). The upper portion of the mandibular arch as far as the middle of the posterior border of the eye is covered by the neural crest. A growth of the neural crest passes over the dorsal border of the eye. 37. the first basibranchial is attached to mesoderm which continues ventrally toward the pharynx. Out of this mesoderm along the anterior part of the pericardial chamber is formed the second basibranchial. Its distal half lies ventral to the level of the other cartilages. The mesoderm continuous with its distal extremity is the anlage of the thoracicohyoideus muscles, while the meso- derm continuous with its proximal or attached extremity is the anlage of the geniohyoid muscles. It has not developed cartilage CRANIAL GANGLIA OF AMBLYSTOMA 455 cells at this stage and its large characteristic mesodermal yolk granules (fig. 37), which are retained for a long time, make a sharp contrast with the other branchial cartilages, which now contain practically no yolk granules. Stage 46+ Lateral-line growps of sense organs. The subsequent growth of the lateral-line primordia is merely a matter of the development 37 Fig. 37 Frontal section at the level of first basibranchium, showing at stage 42 formation of second basibranchium from mesoderm near anterior wall of peri- cardial chamber. X 250. of sense organs out of the ectodermal thickenings as they are laid down in stage 37-38 (fig. 10). The general distribution of the lateral-line sense organs has been described in a number of amphibians. Descriptive figures are shown by Kingsbury (’95) of the distribution of different groups of sense organs connected with the lateral-line system in eight groups of amphibians. However, two figures (figs. 11 and 12) have been included in this paper to show the groups of sense organs developed from the primordia of stage 37-38 as they are finally laid down in a 133-mm. larva of Amblystoma punctatum. 456 L. S. STONE The groups of sense organs may be described in the following manner: 1) the supra-orbital, a grouv usually arranged in a double line passing dorsal to the eye then ventral to a region median to the external nares; 2) the infra-orbital, a group sometimes in a single, sometimes in a double row, passing ventral to the eye to a region on the upper jaw ventral to the external nares; 3) the occip- ital, an approximately triangular group of variously arranged organs with a short line of four or five sense organs extending caudally as far as the region above the third external gill; 4) the hyomandibular or postorbital, an irregularly arranged group, usually a double line associated with the infra-orbital passing caudally and ventrally from the latter to the ventral side of the head; 5) the mandibular, a single line of sense organs extending along the side of the lower jaw from the symphysis back to the ventral extremity of the hyomandibular group and giving off, at the angle of the jaw, a branch of two or three sense organs, sometimes called the angular group, extending dorsally to the infra-orbital line; 6) the ventral-hyomandibular group, a double line of sense organs on either side, beginning at the symphysis of the jaw passing median to the mandibular group to meet the latter at its posterior extremity, where it gives off a single line of seven or eight sense organs which passes directly mesially, but does not join with the corresponding branch of the opposite side; 7) the midbody line, a single line of sense organs extending from.a region above the proximal end of the limb behind the external gills to the region above the level of the anus, where it bends up- ward to pass along the upper border of the somites to the end of the tail; 8) the dorsal body line, a single line of sense organs extending from behind the dorsal end of the third gill along the upper border of the somites, where it joins the midbody line in the region above the anus; 9) the ventral body line, a single line of sense organs curving around the ventral border of the fore limb and extending to the ventral border of the hind limb. In this morphological study, in which critical stages have been described, it is obvious that at the time of the closure of the neural folds (stage 21) no definite placodes can be accurately located, unless it be a slight indefinite thickening dorsal to the CRANIAL GANGLIA OF AMBLYSTOMA 457 eye. The ectodermal thickening over the dorsal extremity of the hyomandibular cleft as well as the longitudinal ectodermal thickening seems to have no significance except in so far as it is the result of a change in contour of the underlying tissue. As early as stage 25 (fig. 3), however, the placodes of lateral-line ganglia of VII and X can be identified, while all the epibranchial placodes can be located at stage 26-27 (fig. 4). The primordia of all the groups of the lateral-line sense organs are established at stage 35+ (fig. 8), and out of this pattern the sense organs are laid down in the skin. EXPERIMENTAL All the operations were confined to the right side of the em- bryos and consist of two groups: a) extirpating the placodes in the ectoderm, followed by grafting into the wound indifferent ectoderm taken from another animal of the same age and, 6) re- moving the neural crest either by scraping it loose or by removing it along with the upper half of the neural canal from the anterior border of the third somite to the anterior extremity of the brain above the eye. Where indifferent ectoderm was transplanted into denuded areas the grafts were removed from animals which had previously been stained in the jelly capsule with Nile-blue sulphate (Det- wiler, ’17).. This made possible daily observations upon the early growth and position of the implanted tissue. The mortality of operated animals was lowered by keeping them during the three or four days following the operation in a glass chamber, around which cool tap-water ranging from 11° to 15°C. was constantly running. This retarded wound healing at first, but no disturbances were observed to affect the results of the experiments... In stage 25 the auditory placode becomes marked externally as a small round pigmented spot a little below an elevation in the region of the hind-brain (fig. 3). This may be used almost invariably as a landmark in orienting operations upon the placo- dal regions in this and early embryos. Since there is no external evidence of the pigmented spot previous to this stage, the eleva- 458 L. S. STONE 40A CRANIAL GANGLIA OF AMBLYSTOMA 459 tion in the region of the hind-brain lends itself as a landmark in orienting similar operations upon embryos from the closure of the neural folds to stage 25 (figs. 1 and 2). A. Extirpation of placodes 1. Removal of ophthalmic placode. A triangular piece of ecto- derm, including skin dorsal to the optic vesicle, was removed in the manner shown in figure 38 and indifferent ectoderm was grafted upon the denuded area. Such areas of ectoderm were taken from embryos varying in stages from 23 to 26. A typical picture of the extension of the implanted blue ecto- derm may be obtained from figure 39. Gradually the blue area extends ventrally until it includes the ectoderm over the eye and the regions anterior and posterior to it, always showing the more extensive migration towards the ventral half of the mandibular arch and towards the first external gill. As a result of this, the blue ectoderm several days after the operation comprises an area far greater in extent than the original extirpated area. After seven or eight days the Nile-blue sulphate gradually dis- appears and the transplanted area can often be located by its scarcity of pigment—a characteristic of the ventral ectoderm of older embryos. The same condition of the extension of blue implanted ectoderm may be seen in many other cases (figs. 48, 49,55, 63). This may be largely due to a migration of the super- Fig. 38 Showing triangular piece of ectoderm, indicated by dotted line and middorsal line which was excised in order to remove the ophthalmic placode at stage 28. X 10. Fig.39 Camera-lucida drawing of the same individual a few days after opera- tion, showing extent of migration of implanted ectoderm stained in Nile-blue sulphate. X 10. Fig.40aandb Showing, respectively, lateral views of a reconstruction in the trigeminal region of the normal and operated sides. Only a small gasserian ganglion appears on the operated side. X 50. Figs. 41 and 42 Frontal sections of the specimen reconstructed in figures 40a and b comparing the normal and operated sides at levels of the root of V. xX 37. Fig. 43 Frontal section of specimen killed nineteen days after operation, showing on right side only the gasserian portion of V after the ophthalmic placode has been excised. X 387. 460 L. S. STONE ficial cells of the ectoderm, for the deeper-lying sense organs in the intruded area are undisturbed. However, this is a subject for further investigation. A specimen killed eight days after operation shows a complete absence of the ophthalmic ganglion and the ophthalmicus pro- fundus V nerve (figs. 40 a and b). The size of the ophthalmic ganglion of the normal side of this individual is very large (fig. 41), and a comparison between it and the region of the root of the V on the operated side (fig. 42) renders a striking picture. The crest cells concerned in the formation of the visceral skeleton along the region of the mandibular arch are approximately nor- malin amount. This condition indicates that if any of the crest cells had been disturbed there has been a regeneration, and ample opportunity has been offered to contribute to the forma- tion of the ganglion if such were their function. This individual also shows a diminution in the size of the gasserian ganglion as illustrated by figure 40 a. Another individual killed nineteen days after a similar operation at stage 23 shows no ophthalmicus profundus V nerve and the small ganglionic mass which does appear is apparently concerned only with the gasserian portion (fig. 43). An individual which had been operated upon at stage 26 shows nine days later a very small ganglion represented by possibly no more than ten cells (fig. 44). The only representative of the ophthalmicus profundus V nerve appears in but one section and extends anteriorly as a slender nerve fiber. During stage 26 cells are being given off from the ophthalmic placode, and it is quite possible that a few of the placodal cells remained in the wound and later gave rise to the remnant of the ganglion which appears in this section. ‘The gasserian ganglion is also smaller in this individual than is the normal ganglion of the left side. An individual which was operated upon at stage 23 shows a small ophthalmic ganglion in an abnormal position (figs. 45 a and b). This ganglionic mass lies median to the eye and some- what dorsal to the optic nerve. It is connected to the brain by a slender group of fibers which pass back through the gasserian ganglion into the brain. At the lower anterior portion of this CRANIAL GANGLIA OF AMBLYSTOMA 461 ganglionic mass there arises a slender nerve which sends a few cutaneous fibers to the skin over the anterodorsal portion of the eye. It then follows over the dorsal border of the nasal organ, in the region of which it gives off cutaneous fibers. In the distribu- tion of its fibers this nerve issuing from the anteriorly placed ganglion simulates the ophthalmicus profundus V nerve. A Fig. 44 Frontal section, showing relative difference in size of the ophthalmic ganglia on normal and operated sides. Part of the ophthalmic placode was pos- sibly not removed. X 37. Fig.45aandb. Lateral views of a reconstruction, showing, respectively, nor- mal and operated sides in the region of V and VII. The posterior portion of VII ganglion is not shown. The operated side (fig. 45b) shows a small ophthalmic ganglion displaced anteriorly and lying above the optic nerve. It is connected posteriorly with the V and VII complex by a long slender root a few fibers from its dorsoposterior border. A small ophthalmicus profundus V nerve extends from its anteroventral border. X 50. 462 L. S. STONE study of the records of the growth of the transplanted blue ectoderm of this individual shows that not all of the ectoderm concerned in the formation of the ophthalmic ganglion was elim- inated from the ectoderm in the region anterodorsal to the optic vesicle. A control operation was made on several specimens in which the ectoderm was excised as in the usual operative procedure and replaced and allowed to heal into its normal position. When these specimens were sectioned they showed perfectly normal ophthalmic ganglia, indicating that if there had been any dis- turbance of the crest cells in the trigeminal region during the operation it did not, in itself, affect the formation of the ganglion. A large number of individuals in which the excised area con- taining the preauditory placode and supra-orbital primordium was removed along with some of the ectoderm above the optic vesicle, show a similar displacement of the ophthalmic ganglion. Among these cases where the embryos were preserved within a few days after the operation ganglionic masses of cells lie close to the ectoderm. ‘Their posterior ends become attenuated and no connection to the brain is discernible. The older individuals which were sectioned show cutaneous fibers issuing from the ganglionic mass comparable in their distribution to the fibers from the ophthalmicus profundus V nerve as in the case already cited (fig. 45 b). It seems quite evident from the above results obtained that the formation of the ophthalmic ganglion is largely if not entirely dependent upon the placode in the ectoderm above the optic vesicle. 2. Removal of gasserian placode. A rather extensive rectangu- lar piece of ectoderm was removed, including all the ectoderm around the dorsoposterior quadrant of the eye. An incision was made beginning at a point on the hyoid arch at a level with about the middle of the eye and passing dorsally some distance in front of the auditory placode to a level about the middle of the latter. The incision was then extended anteriorly parallel with thi. dorsal line to a point above the middle of the eye. From this point it was carried ventrally to a little above the middle of the CRANIAL GANGLIA OF AMBLYSTOMA 463 eye, from which it was extended posteriorly to the incision on the hyoid arch. The ectoderm thus outlined was very carefully removed and the wound was covered by a graft of blue indif- ferent ectoderm. A specimen killed six days after operation at a time when the ophthalmic and gasserian ganglia have just fused at the point of the entrance of the trigeminal root into the brain shows on the operated side a somewhat smaller gasserian ganglion. The fron- tal sections show that the ganglion in its anteroposterior diam- eter varies little in the region near its root from that on the normal side, but as the ganglion is followed ventrally it soon be- comes attenuated in the region near the upper posterior border of the eye, while on the normal side the mandibular nerve does not appear until the ganglion reaches the level with the middle of the posterior border of the eye. The diminution in the size of the gasserian ganglion is therefore represented mostly by a shorter dorsoventral axis. A large portion of the ophthalmic ganglion is displaced and lies near the anterodorsal portion of the eye. A group of fibers connect it to ganglionic cells which lie on the anterodorsal border of the gasserian ganglion. The crest cells over the mandibular arch are apparently normal in amount. Two other specimens killed the same number of days after operation show the same results. In these cases, however, there appears to be a slightly smaller number of crest cells over the mesoderm of the mandibular arch. ‘A control operation was also made on several specimens, in which the ectoderm was excised in the usual manner and then replaced and allowed to heal in its normal position. Sections of these specimens show perfectly normal gasserian and pro- fundus ganglia. It has already been shown (figs. 40 b and 438) that when the ophthalmic placode was removed, including considerable ecto- derm from the posterodorsal region of the eye, the gasserian _ .on was smaller than on the normal side. Among the cases, described in another section of this paper, in which placodes of VII were removed there appear two cases in which there is apparently no gasserian ganglion present (fig. 56 b). THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 35, NO. 4 464 L. S. STONE It seems quite evident from these results that the ectodermal adhesion of the early gasserian ganglion is in the nature of a small placode which contributes cells to the formation of that ganglion. 8. Removal of preauditory placode and the supra-orbital pri- mordium. In this type of operation the rectangular area of Fig. 46 Camera drawing showing area of ectoderm excised in removing pre- auditory placode and supra-orbital primordium. Outlined by means of dotted lines and the middorsal line. X 10. Fig.47 Camera drawing, showing absence of the supra-orbital group of sense organs after removal of an area of ectoderm as shown in figure 46. X 10. Fig. 48 Camera drawing, showing only three or four supra-orbital lateral line sense organs when not all of the early preauditory placode is removed during stage 21. The lightly pigmented area of transplant of indifferent ectoderm at time animal was killed is shown by dotted line. > 10. ectoderm excised is outlined in figure 46. Indifferent ectoderm from the ventral side of another animal of the same age which had been previously stained with Nile-blue sulphate was grafted into the wound. This operation was confined to stages under stage 25. CRANIAL GANGLIA OF AMBLYSTOMA 465 Due to the inability to locate exactly the position of the audi- tory placode, especially in stage 21, it was often included wholly or in part in the excised ectoderm. Among such cases where the auditory vesicle was not entirely absent, it appeared as a small rudimentary vesicle near the skin. An examination of specimens which were treated in the manner just described shows in eight cases no supra-orbital group of lateral-line sense organs (fig. 47), and consequently no ophthal- micus superficialis VII nerve. The operations in seven of these cases were performed at stage 23 and in one case about stage 21. The auditory vesicle appears as a rudiment in six cases, while in the other cases it is entirely absent. The larger portion of the lateral-line ganglion of VII which supplies nerve fibers to the supra- and infra-orbital groups of sense organs is represented in these cases by a very small ganglion. The infra-orbital line of sense organs with its corresponding nerve was never absent in this type of operation and it received its fibers from the small remaining portion of the above-mentioned lateral-line ganglion. In a few cases where a small portion of the placode was, pre- sumably, not entirely extirpated, the VII lateral-line ganglion is smaller than on the normal side and a small line of sense organs is the only representative of the supra-orbital line (fig. 48). In some cases a sensory line was represented by small, poorly de- veloped organs and the ophthalmicus superficialis VII is so slender that it can be followed with great difficulty and then only when most favorably stained. Such cases show diminution in the size of the VII lateral-line component which supplies the supra-orbital and infra-orbital sensory lines. Throughout all these cases in which the supra-orbital sensory line is represented in part or in its entirety the ear is present and apparently normal in size. The hyomandibular, ventral, and mandibular groups of lateral- line sense organs were never disturbed when such an area of ectoderm was removed. The VII lateral-line component of the ganglion supplying nerves to these groups was always normal. It is quite apparent that no other portions of the VII ganglion are lacking. 466 L. S. STONE Fig. 49 Camera drawing, showing absence of hyomandibular group of sense organs after removal of an area of ectoderm shown in figure 50. X 10. Fig.50 Camera drawing, showing by dotted line an area of ectoderm removed at stage 26 in which was included the primordium of the hyomandibular group of sense organs as well as epibranchial placode of VII and supra-orbital primordium. x 10. Fig.51 Camera drawing, showing by dotted line an area of excised ectoderm in which was included the primordium of the ventral hyomandibular group of sense organs and part of the primordium of mandibular group. X 10. Fig. 52 Camera drawing, showing by dotted line position of implanted blue ectoderm a few days after operation in same individual shown in figure 51. The balancer is absent on operated side. X 10. Fig. 53 Camera drawing twenty-one days later of the ventral side of embryo shown in figure 51. On the operated side there is only a small portion of the man- dibular group of sense organs, while the ventral hyomandibular group is entirely absent. X 10. CRANIAL GANGLIA OF AMBLYSTOMA 467 4. Removal of the infra-orbital, hyomandibular, ventral hyomandib- ular and mandibular primordia. A number of trials were made upon embryos under stage 26 to remove separately, if possible, the primordia of the infra-orbital and hyomandibular groups of lateral-line sense organs on the side of the head by excising small segments of ectoderm behind the upper posterior border of the optic vesicle. These attempts proved to be fruitless, for in all cases the ingrafted tissue failed to suppress their development and complete regeneration of the two primordia took place. However, among the cases where an area of ectoderm including placodes of the VII ganglion was removed there appears one case, operated upon at stage 26, in which the hyomandibular group of sense organs is entirely absent, although the infra-orbital group is intact as well as the groups of sense organs on the lower side of the jaw (fig. 49). The area removed (fig. 50) also includes the epibranchial, the preauditory, and supra-orbital placodes. This seems to indicate that the infra-orbital and hyomandibular groups have separate primordia. A few operations were made in an attempt to remove the pri- mordia of the sense organs on the under side of the jaw. The most successful case appears in an individual operated upon at stage 25. The area removed is shown in figure 51, one day after operation, and it includes a small amount of ectoderm anterior and posterior to the lower portion of the hyomandibular cleft. Several days after operation (fig. 52) the area of the transplanted ° indifferent ectoderm may be seen at a lower level than its original position and the migration of its posterior portion extends into the anteroventral border of the first external gill. The balancer is lacking on this side. Twenty-one days after operation this individual reveals no ventral hyomandibular group of sense organs on the right side of the ventral portion of the lower jaw (fig. 53). However, there is a remnant of the mandibular group in the form of a few small sense organs along the anterolateral border of the lower jaw. This may possibly be an indication of the existence of a separate primordium for the mandibular group of sense organs. 5. Removal of the epibranchial placode of VII and surrounding ectoderm. ‘The area of the extirpation in this type of operation 468 L. S. STONE wee Fig. 54 Camera-lucida drawing, showing by dotted line extirpated area of ectoderm one day after operation. The operation was done at stage 25 and in- cluded auditory and preauditory placodes as well as epibranchial placode of VII. 37. 478 CRANIAL GANGLIA OF AMBLYSTOMA 479 region of the mesoderm of the hyoid and the first and second branchial arches a distinct lack of crest cells on the operated side, while on the normal side they are abundant and still surround the mesoderm of the arches. Case 3. A similar specimen was killed nine days after opera- tion, when it had reached about stage 37. The external gills at this stage have become prominent finger-like processes, but have not “developed branches. A frontal section of this specimen (fig. 70), although obliquely cut, shows to even more advantage the comparison of the two sides, for the level of the operated side is lower where there would be normally an abundance of procartilage-forming cells. There are but three crest cells in the region of the third branchial arch, while the others lack cells of this kind altogether. This is the condition in sections above this level. The crest cells in the hyoid region are also noticeably less in amount than on the left side. At a lower level (fig. 71) the difference is more striking, for the hyoid, the first, and part of the second branchial regions have very few cartilage-forming cells compared with the abundance on the normal side. Loose connective tissue in the external gills on the operated side is also subnormal in amount. Case 4. This specimen was killed fourteen days after the operation, when it had reached about stage 39. The gills on the operated side are smaller than on the normal side and their con- nective-tissue cells are less numerous. Compared with the nor- mal side, there are only a few procartilage cells in the third and fourth branchial arches (Sg. 72). When followed farther ven- trally (fig. 73), the few loose cells on the right side increase but slightly in number and always remain much more loose in ar- rangement as they approach in the midline the normal compact procartilage. The procartilage in the hyoid region is also much less in amount. Case 5. This specimen was killéd seventeen days after opera- tion and also shows a decided diminution in the size of the gills on the operated side. A frontal section shows well-formed branchial cartilages on the left, while on the right a few cells form a small cartilage only in the third gill arch. When followed 480 L. S. STONE farther ventrally, very small rudiments of the second, third, and fourth branchial cartilages appear, which, after they approach each other at their ventral extremities, become completely lost and do not approach in the midline the cartilages of the normal side. The hyoid cartilage is completely absent. The first and second ceratobranchials and the ceratohyoid are absent on the operated side (fig. 74). The first and second basibranchials are present. ° A specimen in which all the crest cells were included in the operation along with the upper half of the neural tube was killed eleven days after operation, and it shows many deficiencies in the visceral skeleton. Very small first, second, and third epibran- chial cartilages are found on the operated side (fig. 75); they soon disappear as the sections are followed ventrally. The quad- rate in this and other sections is smaller than on the normal side. The hyohyal, ceratohyal, and all the ceratobranchial carti- lages are entirely absent. The mandible is smaller on the right side. The first basibranchial cartilage is present but small and lies toward the right of the midline, while the anlage of the second basibranchial is apparently normal. No change can be seen in the anterior portions of the trabeculae. A similar attempt was made in another specimen, but due to incomplete removal of crest cells in the branchial region regenera- tion had taken place. However, seventeen days after operation it showed externally a diminution in size of the right side of the lower jaw. 23. Fig.81 Frontal section, showing in another specimen at the level of the optic nerves relative sizes of the quadrates when the crest cells have been removed at stage 24. XX 37. 482 L. S. STONE Another individual was operated on at stage 24 and the crest cells along with the upper half of the neural canal were removed from the hyoid and mandibular regions. There was evidently a regeneration of the crest cells, but the specimen showed certain deficiencies in the visceral skeleton. At a level with the optic nerves a section (fig. 81) shows a small quadrate cartilage on the operated side. In levels above and below this section the dif- ference in the size of the two cartilages is even more striking. The anterior portions of the trabeculae show a difference in size. The anterior portion of the trabecula on the normal side extends somewhat beyond the level of the optic nerve, while on the operated side it extends only to the posterior border of the optic nerve. The ceratohyoid cartilage is poorly developed and joins with the first basibranchial posterior to its normal position. The mandibular cartilage in this individual also shows a defi- ciency in size. It was found in the large number of operations made in re- moving the neural crest that the most favorable stage was around stage 26. However, this applies only to the branchial and hyoid regions. Here their removal is a very simple matter and no damage to the mesoderm need be expected, for the crest cells lie loosely upon the mesoderm and do not extend very far ventrally. In the mandibular region the elimination of crest cells is more difficult, for several factors are involved. From stage 23 to 26, when the regeneration of the neural crest is most persistent, a very rapid ventral growth takes place far in advance of the ven- tral proliferation of crest-cell groups in the hyoid and branchial regions. ‘The crest cells lie tightly against the mesoderm of the mandibular arch, and in order to remove these cells considerable damage is caused to the mesoderm. The wound is very exten- sive ventrally, and the pulling of the cut edge of the ectoderm near the optic vesicle, accompanied by the large cavity made in the anterior portion of the brain, always resulted in the death of such individuals. It is hoped that by doing a large number of operations in this region a few specimens may survive which will show a complete removal of the crest cells over the trigeminal region. CRANIAL GANGLIA OF AMBLYSTOMA 483 An examination of the specimens done at the earlier stage shows a deficiency of the mandibular and quadrate cartilages, but not a complete absence. One case shows a deficiency in the anterior portion of the trabecula. In the branchial and hyoid regions it is quite evident that the branchial cartilages, the cerato- hyal, the hyohyal, and the first basibranchial are derived from the wandering neural crest. 2. Contributions to ganglionic components. a. Contribution to V and VII. As already shown in the cases described, there is no condition in which the crest cells in the region of the trigeminus were entirely removed, although in the other regions they have been almost entirely eliminated. A specimen operated upon at stage 25 was killed two days after operation and shows on the operated side a portion of the ophthalmic placode which is giving off cells, even though few crest cells may be found in this region. It lies close to the brain (fig. 82) while farther ventrally placodal cells from the posterior portion of the VII lateral-line placode are projecting medially to form the lateral-line ganglion (fig. 83). At its anterior portion it is contiguous with a small group of cells which belong to the gasserian placode. The epibranchial placode of VII at this stage . is prominent, but has split off no cells on either side. Although the crest cells have regenerated to a small degree in the trigeminal regions of two other specimens killed within a few days after operation, placodal cells may be seen contributing in large numbers to an ophthalmic ganglion. The condition of the ganglion is practically the same as on the normal side. One of the specimens shows a large well-formed ophthalmic ganglion (fig. 84), but the number of crest cells on the mesoderm of the mandibular arch farther ventrally is far less than on the normal side. Although there are but few crest cells on the mesoderm of the hyoid arch of this specimen, the VII lateral-line ganglion is present and normal (fig. 84). This level does not show the comparative size of the two ganglia. The visceral portion of the VII seems to be a little smaller than on the normal side. Another typical case killed eleven days after operation shows deficiencies in the mandibular group of crest cells and an entire 484 L. S. STONE absence of the ceratohyal cartilage. The ophthalmic ganglion is large and normal in size, although it is situated somewhat dorsal to its normal position (fig. 85). The portion of the lateral-line ganglion of VII which supplies the supra-orbital and infra-orbital group of sense organs also appears in this level and is normal in size as well as the rest of the lateral-line ganglion. gas. --Mpl 82 83 Fig. 82 Frontal section, showing the ophthalmic placode two days after the crest cells have been removed. X 37. Fig. 83 Showing in the same specimen the posterior portion of VII lateral- line placode giving off placodal cells. At its anterior extremity is shown the contact of the gasserian ganglion with the ectoderm. X 37. Fig.84 Showing a normal VII lateral-line ganglion when crest cells have been removed from the hyoid arch. X 37. Fig. 85 Frontal section, showing a large ophthalmic ganglion after crest cells had been removed at an early stage. X 387. Fig. 86 Frontal section further ventrally in the same individual, showing the palatinus VII at the point of leaving the ganglion. X 37. CRANIAL GANGLIA OF AMBLYSTOMA 485 The gasserian ganglion is slightly smaller than on the normal side and lies somewhat separated from the ophthalmic ganglion. The visceral portion of VII is a little smaller and no definite alveolar nerve can be found. The special visceral or palatinus VII is present (fig. 86). Fig.87 Frontal section of a specimen killed three days after removal of crest, cells over the branchial region, showing lateral-line placodes of IX and X forming ganglia. Between them in ectoderm lies the epibranchial placode of X. On the normal side lies a group of crest cells near the epibranchial placode of X, possibly forming a portion of visceral ganglion of X. X 37. Fig. 88 Portion of a frontal section in a specimen killed a few days after re- moval of crest cells, showing normal epibranchial placodes of IXand X. X 37. Fig.89 Showing a large vagus lateral-line ganglion and ramus lateralis supe- rior vagiin a specimen killed eleven days after an attempt had been made to remove all crest cells on right side. X 37. Fig.90 Showing in the same specimen the IX visceral ganglion. X 37. b. Contribution to IX and X. A specimen killed three days after operation shows on the operated side a large lateral-line ganglion placode of IX just anterior to the epibranchial placode of X (fig. 87). On the anterior surface of the large vagus lateral- line placode are a number of loose placodal cells given off from the placode. No crest cells are found in the branchial region of the operated side. ‘On the normal side anterior to the vagus 486 L. S. STONE placode is a mass of crest cells near the ectoderm and epibran- chial placode of X. These are crest cells which possibly give rise to a portion of the visceral ganglion, for when followed ven- trally they lie close to the epibranchial placode of X. A specimen killed five days after operation shows on the operated side a large normal lateral-line ganglion. The placode of the lateral-line ganglion of IX seems to be a little less advanced in development than on the normal side. Farther ventrally (fig. 88) loose masses of cells are being given off from the epi- branchial placodes of IX and X and are not of crest-cell origin, for there are no crest cells in the branchial region of the operated side. A specimen killed eight days after operation shows a large vagus lateral-line ganglion, in front of which is a portion of a small visceral ganglion. The ganglia are still made up of loose cells and nerve fibers from the visceral ganglia cannot be determined. Posterior to the auditory vesicle on the operated side is a small lateral-line ganglion which lies above the small visceral ganglion of IX which is connected with the epibranchial placode. Only a very few scattered crest cells appear on the median surface of the mesoderm of the branchial arches. A number of other specimens killed between eight and ten days after operation show small visceral ganglia, which are derived from the epibran- chial placodes, and perfectly normal IX and X lateral-line eanglia. In these cases also there are only a few loose crest cells in the branchial region on the operated side. A typical specimen killed eleven days after operation shows a large vagus lateral-line ganglion with the ramus lateralis superior vagi nerve extending posteriorly to innervate the body line of sense organs (fig. 89). The anterior portion of this ganglion is part of the visceral ganglion of X. From it pass slender fibers to the epithelium of the second external gill and also a number of motor fibers to the branchial trunk of the vagus can be seen in the second branchial arch. It gives off motor fibers to the bran- chial muscles, and when followed ventrally into the branchial arch it is lost near the epithelium on the pharyngeal side. There is no definite second branchial trunk. As the superior lateralis CRANIAL GANGLIA OF AMBLYSTOMA 487 vagi nerve leaves the ganglion, fibers are continued ventrally, as on the normal side, to form the visceral trunk of the vagus. A few motor fibers are given off along its path, and when followed farther ventrally it is finally lost near the wall of the pharynx. No ramus intestinalis from the visceral trunk can be determined. In the dorsal portion behind the ear a number of cutaneous fibers may be seen, along with the lateralis fibers, to innervate the skin. The visceral portion of [IX may be seen at the lower border of the ear coming off from the root of 1X which is at a lower level than on the normal side (fig. 90). The visceral ganglion is small and gives off no trunk to the first branchial arch, but ventral to the ear capsule it sends out a nerve which passes some distance along the median border of the internal ceratohyoid muscle, where it is followed ventrally until its fibers are lost against the wall of the . pharynx. The only visceral fibers that can be identified on the operated side appear to be of the special visceral system. The other fibers which appear in the IX and X ganglionic complex on the operated side are of the lateralis and general cutaneous sys- tems. In the ventral positions of the third and fourth branchial arches only very ‘small rudiments of cartilages appear, which shows that very few crest cells remained which could have con- tributed to the visceral ganglia. It appears from the study of the specimens described that the only contribution of the crest cells to the ganglionic complex of IX and X is to the general visceral component. DISCUSSION It has been shown in the study of early stages of Amblystoma embryos that extensive contributions from the lateral ectoderm take part in the formation of cranial ganglia, and the experi- mental analysis of the problem has shown how small a part the crest cells play in the formation of these ganglia. The facts which the experiments present lead to the conclusion that the general cutaneous system is derived entirely from placodes. In the trigeminal region there are two definite placodes concerned with the formation of the V; that in the case of the ophthalmic division is the larger, while that of the gasserian is the smaller and 488 L. S. STONE of shorter duration and, therefore, difficult to follow. The earliest stage in which Coghill (’16) described the early contact of the ophthalmic ganglion with the ectoderm corresponds to about stage 34, i.e., a stage between those shown in figures 6 and 7. At the point of contact lies the placode which can be last seen at this stage. The actual contribution of placodal cells to the ganglion must be observed in earlier stages than this, at a time when crest cells are still very numerous in this region. This condition has been the factor which has caused investigators to overlook the placodal contribution. Coghill’s observation that during this contact with the ectoderm the ganglion makes its connection with the brain adds further morphological evidence that the ganglion is of placodal origin. Judging from the many similarities in Amblystoma and Necturus, it is obvious that Platt was correct in assuming an ectodermal contribution to the oph- thalmie ganglion, although she has confused the placodal and erest cells in this region and incorrectly interpreted part of these placodal cells as contributing to the ‘mesectoderm.’ The ecto- dermal cells which Goette (’14) describes being given off above the optic vesicles in Siredon (Amblystoma tigrinum) and Torpedo are likewise cells of the ophthalmic placode and not contributions to the wandering ‘ectomesoderm.’ It has been shown that when the ophthalmic placode is entirely removed in early stages of Amblystoma there is a complete absence of the ophthalmic ganglion and the ophthalmicus profundus V nerve (figs. 40 b and 43). Such cases always show but little disturbance of the crest cells in the trigeminal region. In control operations in which the ectoderm was removed in the usual manner and then replaced there appeared normal ophthalmic ganglia and nerves, which shows that the placode is necessary for the formation of the ganglion. When a small portion of the ectoderm over the eye is left a very small displaced ganglion is often found (fig.45b). In this respect it is similar to other partially removed placodes. On the other hand, when the ophthalmic placode is left intact and the crest cells have been disturbed as much as possible by an attempt to remove them, an apparently normal ganglion is pres- ent, and although there has been a regeneration of the crest cells CRANIAL GANGLIA OF AMBLYSTOMA 489 it shows that an extensive disturbance of the crest cells as early as stages 21 and 23 does not inhibit the growth of the ganglion. As Coghill (16) has already shown in Amblystoma, there is an early contact of the gasserian ganglion with the ectoderm. At the point of contact near the anterior border of the preauditory plac- ode (figs. 21 and 26) is a small thickening in the ectoderm which can be followed only through stages 28 to 30. Although this condition has not been described in any other forms, it is quite - possible that it does exist, but has been overlooked on account of its small size and short duration in a region where the crest cells are very abundant. Among the cases where the removal of the preauditory placode included ectoderm near the posterodorsal border of the optic vesicle there occurred two cases (fig. 56 b) in which there was no gasserian ganglion. When smaller areas of ectoderm were removed from the posterodorsal region of the eye there often occurred small gasserian ganglia. This was pos- sibly due to the fact that not all of the placode had been removed. In one case where there was a deficiency in the crest cells on the mandibular arch after the crest cells had been removed a smaller gasserian ganglion was observed. In this case it seems quite possible in the light of the control and other operations that the gasserian placode was injured when the ectoderm was reflected at the time of operation. The remaining portion of the general cutaneous system of the cranial nerves is to be found in the X. In the observations re- ported in this paper no definite distinction could be made in the early stages between the small general cutaneous and the visceral ganglia of X. Coghill (’16) has observed that during its early contact with the ectoderm, the cutaneous ganglion of X has no connection with the brain. However, when a large area of ecto- derm was removed containing the epibranchial placodal regions of IX and X no definite general cutaneous fibers could be found. This leads one to conclude that the general cutaneous portion of the vagus complex is derived from the lateral ectoderm and the early contact of the small general cutaneous ganglion of X, which was described by Coghill, is the indication of a placode in the ectoderm which gives rise to that ganglion. This fact falls in 490 L. S. STONE line with the results obtained in the removal of placodes in the trigeminal region and shows that the general cutaneous system is placodal in nature and not, as Landacre (10) has suggested, of neural-crest origin. From the results obtained in the series of experiments recorded in this paper there can be no doubt that the lateral-line sensory system in Amblystoma is derived entirely from placodes. The ‘study of the preauditory placode shows that a large part of the lateral-line ganglion of VII is formed from this placode and from its anterior end arises the supra-orbital primordium of sense organs. The other lateral-line primordia in the head region are separate in origin as in Necturus and also contribute to the VII lateral-line ganglion. At no time is there any condition such as that described in Lepidosteus by Landacre and Conger (713) in which the preauditory placode begins to disintegrate at the time when the first trace of the lateral-line primordium can be detected. It is quite possible that Landacre and Conger were misled in this interpretation of the preauditory placode, for, according to their description, it apparently arises very early, and although they describe no cells being given off from the placode, it seems prob- able that there may have been an early contribution which was unobserved. In the case of the postauditory lateral-line primordia, the study of experimental as well as of. normal material shows that the three trunk lines of sense organs have separate primordia, and in this respect Amblystoma is similar to Necturus. The experimental results show that the occipital group of sense organs appeared in a few cases where the ectoderm in the anterodorsal portion of the gill swelling (fig. 61) was not entirely removed, al- though the ear was entirely removed. This condition implies the independence of the occipital primordia from the auditory placode and also indicates in embryos close to stage 21 the ability of the ectoderm in the anterodorsal region of the gill swelling to give rise to occipital primordia. The complete removal of the postauditory lateral-line primordia was not only accompanied by the absence of the groups of sense organs, but by an entire absence of the lateral-line ganglia. When only a few sense CRANIAL GANGLIA OF AMBLYSTOMA 491 organs appeared, correspondingly small lateral-line ganglia were present which innervated those sense organs. No evidence can be obtained that crest cells contribute to the formation of lateral- line ganglia. The morphological studies are misleading in this respect, for in many cases the close arrangement and contact of the early crest cells and placodes make an interpretation of the exact contribution of the two kinds of cells difficult to understand. In the study of the normal embryos the epibranchial placodes of VII, IX, and X could be located as early as stage 26-27 (fig. 4) and their contributions to ganglia could be followed up to stage 36. The observations of these placodes agree in many respects with Coghill as to the placodal relation and contribution to the visceral ganglia. The removal of epibranchial placodes was found to be accompanied by a distinct lack of gustatory fibers in VII, IX, and X with no apparent disturbance to the general visceral system. In this respect the experimental results agree with Landacre’s explanation of the function of the epibranchial placodes in Lepidosteus, viz., that they give rise to special vis- ceral ganglia of VII, IX, and X. Goette (14) expresses the belief that the epibranchial and lateral-line placodes form, with the crest cells, a syncytial mass of cells out of which ganglia and nerves are formed which later join themselves up with the brain. Studies of experimental and normal amblystoma embryos show that certain definite portions derived from placodes and crest cells, although they mingle with each other, maintain their identity and are not to be considered a syncytial mass at any time. It has also been shown in Amblystoma that the neural crest originates from the dorsal portions of the contiguous surfaces of the neural folds at the time of the closure (fig. 13). These crest cells were followed by means of their difference in pigment and by the presence of fine yolk granules in their cytoplasm as they migrate ventrally over the mesoderm of the arches, always re- maining separate from the ectoderm. The wandering mass of ectoderm is of crest-cell origin only and does not in Amblystoma receive any contribution from the ectoderm on the lateral sur- faces of the head. Platt’s descriptions of Necturus show clearly 492 L. S. STONE that the positions of the proliferating lateral ectoderm correspond to placodal regions and since the migration of crest cells soon produces a scarcity of these cells in the dorsal region of the neural canal and an abundance of crest cells in the region where placodal cells are given off, a condition is brought about which would lead to a confusion as to the origin of the wandering ‘mesectoderm.’ Aside from the different interpretation in the origin of the ‘mes- ectoderm’ in Amblystoma the manner of the formation of the branchial cartilages with the exception of the second basibran- chial agrees with the description which Platt (97) gives of the branchial cartilages in Necturus. This is fully in accord with Landacre (’21). The branchial cartilages with the exception of the second basibranchial have been conclusively shown to have their origin in the neural crest. At the time when this skeleton begins to take on a cartilaginous appearance the first basibran- chial extends a short distance posteriorly from the attachment of the ceratobranchial cartilages. This condition is somewhat misleading for it gives the appearance that the second _ basi- branchial is a posterior outgrowth from the first basibranchial. However, such is not the case, for a study of embryos about stage 42 conclusively shows that the second basibranchial is formed out of mesoderm. near the anterior wall of the pericardial chamber and that this cartilage retains large mesodermal yolk granules for a long time after the branchial skeleton from the neural crest has lost all of its yolk granules (fig. 37). Amblystoma in this respect agrees with Landaere’s (’21) description of Plethodon. The experimental results show that some of the neural crest is incorporated in the connective tissue of the external gills as well as in the balancer, as Harrison (’21) has shown. How much more of the connective tissue in the branchial region is formed from the crest cells is impossible to determine at this time. In the case of the mandibular and quadrate cartilages the experimental results, as already stated, did not show as conclu- sively as the findings from the study of normal embryos that they are derived from the neural crest, because of the difficulty in eliminating the crest cells from the trigeminal region. Never- theless, the results do show that there was a decided diminution CRANIAL GANGLIA OF AMBLYSTOMA 493 in the size of those cartilages when the crest cells were removed. Although it was difficult to determine from the studies of normal embryos what became of the neural crest migrating over the dorsal and anterior margins of the optic vesicles, one case (fig. 81) in which the crest cells were removed in the trigeminal region seems to show conclusively that they form the anterior portions of the trabeculae. This is in accordance with the findings of Platt (97) and Landacre (’21). In no case do the crest cells enter into the formation of any part of the branchial musculature as described by Goronowitsch (’93). The musculature of the visceral arches is formed entirely from the mesoderm of those arches. No portion of the skull other than the anterior portion of the trabeculae is formed from the neural crest. The only contribution of the neural crest to the formation of cranial ganglia is probably to the general visceral portions of VII, IX, and X. This conclusion is substantiated by the fact that when the crest cells are removed from the branchial and hyoid regions, there is a distinct lack of general visceral fibers, while the gustatory, lateralis, and general cutaneous fibers are normal. SUMMARY 1. Above the optic vesicle in early stages of Amblystoma there is an elongated ophthalmic placode which gives off cells to the formation of the ophthalmic ganglion. When the ectoderm in- cluding this placode is removed as early as stage 23, the ophthal- micus profundus V nerve and ganglion are absent. 2. Near the anteroventral border of the supra-orbital lateral- line primordium is a small gasserian placode of brief duration which can be followed through stages 28 and 30. When a large area of ectoderm is removed from the region posterodorsal to the eye, deficiencies of the gasserian ganglion are produced. 3. Lying close to the anterior border of the auditory placode is a prominent placode elongated in a direction toward the dorsal extremity of the hyomandibular cleft. It can be located as early as stage 25. When ectoderm in this region is removed, even 494 L. S. STONE before the appearance of the placode, the supra-orbital line of sense organs is absent as well as a large part of the VII lateral- line ganglion. The small remaining VII lateral-line ganglion in such cases gives rise to lateralis fibers, which innervate the group of sense organs on the lower jaw, and also slender fibers to the infra-orbital group. 4. The supra-orbital primordium of lateral-line sense organs arises from the anterior extremity of the VII lateral-line ganglion placode. 5. The supra- and infra-orbital and hyomandibular primordia of lateral-line sense organs have separate seats of origin. 6. The ventral hyomandibular and mandibular groups of lateral-line sense organs also appear to have separate seats of origin. | 7. The epibranchial placodes of VII, IX, and X give off cells which become incorporated in the visceral ganglia, and when these placodes are removed from the ectoderm in early stages (23-26) no special visceral ganglia nor gustatory fibers can be found. ; 8. The complete removal of ectoderm in the region of [IX and X which includes all the primordia of the lateral-line system is accompanied by a complete absence of lateral-line ganglia. When only partially removed, small lateral-line ganglia are produced. 9. Large areas of ectoderm removed from the region of IX and X also show an absence of cutaneous fibers as well as visceral sensory fibers. 10. The lateralis and special visceral ganglia are derived en- tirely from placodes, and the general cutaneous, for the most part if not entirely, is also derived from placodes. 11. The neural.crest cells arise from the dorsal portion of the neural tube at the points of the fusion of the neuralfolds. They can be distinguished as early as the closure of the folds, and from this region they can be followed by their difference in pigmentation from the surrounding tissue and by the presence of small yolk granules in their cytoplasm as they descend upon the mesoderm of the visceral arches around which they wrap themselves and CRANIAL GANGLIA OF AMBLYSTOMA 495 later become arranged on the median surfaces of the arches where they form cartilaginous tissue. 12. The wandering mass of ‘mesectoderm’ is of neural-crest origin and in no region is it augmented by a contribution from cells of the lateral ectoderm. 13. The removal of neural crest in the branchial and hyoid regions is accompanied by smaller external gills and marked deficiencies in the branchial and hyoid cartilages. The hyohyals, ceratohyals, ceratobranchials, epibranchials, and first basibranchial are formed from the wandering neural crest, while the second basibranchial is formed from mesoderm near the anterior wall of the pericardial chamber. In the ganglionic regions the neural crest appears toform only the general visceral components. 14. The removal of neural crest in the trigeminal region shows no complete absence, but deficiencies in the size of the quadrate and mandibular cartilages. However, there is no doubt that these cartilages are formed from crest cells. The regeneration of the crest cells always occurred in the operation of these specimens because they were necessarily confined to very early stages. 15. The neural crest in the trigeminal region which migrates over the anterior border of the optic vesicles apparently gives rise to the anterior portion of the trabeculae. LITERATURE CITED BrarpD, J. 1885 The system of branchial sense organs and their associated ganglia in Ichthyopsida. A contribution to the ancestral history of the vertebrates. Quart. Jour. Micr. Sci., vol. 26, New Series. Braver, A. 1904 Beitrige zur Kenntnis der Entwicklung und Anatomie der Gymnophionen. IV. Die Entwicklung der beiden Trigeminusgang- lien. Zoolog. Jahr., Suppl. Bd. 7. CocuiLt, G. E. 1916 Correlated anatomical and physiological studies of the growth of the nervous system of Amphibia. II. The afferent system of the head of Amblystoma. Jour. Comp. Neur., vol. 25. Detwiter, S. R. 1917 On the use of Nile blue sulphate in embryonic tissue transplantation. Anat. Rec., vol. 13. Donrn, A. 1902 Studien zur Urgeschichte des Wirbeltierkérpers. XXII. Weitere Beitrige zur Beurteilung der Occipitalregion und der Gang- lienleiste der Selachier. Mitteil. aus der Zoolog. Station zu Neapel, Bd. 15. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 35, NO. 4 496 L. S. STONE Froriep, A. 1885 Ueber Anlagen von Sinnesorgane am Facialis, ete. Archiv, f. Anat. und Phys., Anat. Abt. Gorttr, A. 1914 Die Entwicklung der Kopfnerven bei Fischen und Amphibien. Archiv. f. mikr. Anat., Bd. 85. GoronowitscH, N. 1893 Untersuchungen iiber die Entwicklung der sogenann- ten Ganglienleisten im Kopfe der Vogelembryonen. Morphol. Jahrb., Bd. 20. Harrison, R. G. 1918 Experiments on the development of the fore limb of Amblystoma, a self-differentiating equipotential system. Jour. Exp. Zool., vol. 25. 1921 The development of the balancer in Amblystoma. Abstract Proc. Am. Assoc. Anat., Anat. Rec., vol. 21. KastscHenko, N. 1888 Zur Entwicklungsgeschichte des Selachierembryos. Anat. Anz., Bd. 3. Kinesspury, B. F. 1896 The lateral line system of sense organs in some Ameri- can amphibians and comparison with dipnoans. Proc. Am. Micr. Soe., vol. 17. Kuprrer, C. von 1891 Die Entwicklung der Kopfnerven der Vertebraten. Verh. d. Anat. Gesellschaft. 1895 Ueber die Entwicklung des Kiemenskelets von Ammocoetes und die organogene Bestimmung des Exoderms. Verh. d. Anat. Gesell- schaft. LanpacrE, F. L. 1907 On the place of origin and method of distribution of taste buds in Ameiurus melas. (A note on the distribution of the IX nerve by C. J. Herrick, p. 55.) Jour. Comp. Neur., vol. 17. 1910 The origin of the cranial ganglia in Ameiurus. Jour. Comp. Neur., vol. 20. 1912 The epibranchial placodes of Lepidosteus osseus and their re- lation to the cerebral ganglia. Jour. Comp. Neur., vol. 21, no. 1. 1921 The fate of the neural crest in the head of the urodeles. Jour. Comp. Neur., vol. 33. ; Lanpacke, F. L., anp Concrr, A.C. 1913 The origin of the lateral line primor- dia in Lepidosteus osseus. Jour. Comp. Neur., vol. 23. MarsnHatt, A. M. 1878 The cranial nerves in the chick. Quart. Jour. Micr. Sci., vol. Be 1Y Piatt, J. B. 1896 Ontogenetic differentiation of the ectoderm in Necturus. Study II. Development of the peripheral nervous system. Quart. Jour. Micr. Sci., vol. 38, New Series. 1897 The development of the cartilaginous skull and of the branchial and hypoglossal musculature in Necturus. Morphol. Jahrb., Bd. 25. Stone, L. S. 1921 Experiments on the development of the cranial ganglion and the lateral line sense organsin Amblystoma. Abstract Proc. Am. Assoc. Anat., Anat. Rec., vol. 21. WisHE, J. W. vAN 1882 Ueber die Mesodermsegments und itiber die Entwick- lung der Nerven des Selachierkopfes. Amsterdam. SUBJECT AND AUTHOR INDEX DJUSTMENT of grafts over eyes, and to the local specificity of integument. The transplanation of skin in frog tadpoles, with special reference to the.............. 353 Admiral butterfly, Pyrameis atalanta Linn. The chemical sensitivity of the tarsi of the Tyr ticle ee ao seiaoek:. Sei SE Abaco eae ee aa 57 Amblystoma. Further observations on pe- ripheral nerve connections. Experiments on the transplantation of limbsin........ 115 Amblystoma punctatum. Experiments on the development of the cranial ganglia and the lateral line sense organsin...... areee - 421 Amblystoma tigrinum to olfactory stimuli. PNECACHIOHS OLS ss oc. ce eye nt ais diale.c re sleuse 257 Amblystoma toa position at right angles to the normal. Theeffectof transplanting a por- tion of the neural tube of................ 163 Amoeba, Vahlkampfia patuxent, with tissue- culture cells. A comparison ofan........ Anesthetics and CO2 output. I. The effect of anesthetics and other substances on the production of carbon dioxide by certain ONG OP DEER cio. ches ai0 a)sicieie atere siesta «1° see Lo | 5 pee aes of Spathidium, with special reference to the capture and ingestion of ‘its prey. Studies on Spathidium spath- tls mie he struchure and. - 223.020. 189 Bovine, JosepH Hari. Anesthetics and CO2 output. I. The effect of anesthetics and other substances on the production of car- bon dioxide by certain Orthoptera...... 323 Bopine, JosepH Hat. The effect of light and decapitation on the rate of COz output of certain Orthopters:... 2.0.20. 2.0e2 esse 47 Butterfly, Pyrameis atalanta Linn. chemical sensitivity of the tarsi of the red fea ery peed Les Ne ee aa eer NS ee 57 ELLS. A comparison of an amoeba, Vahl- kampfia patuxent, with tissue-culture... 1 CO: output of certain Orthoptera. The effect of light and decapitation on the rate of.. 47 CO: output. I. The effect of anesthetics and . other substances on the production of car- bon dioxide by certain Orthoptera. Anes- MEtACS ANG e: sce Feasts ees awis xccinle s 3 323 Core, Wituiam H. The transplantation of skin in frog tadpoles, with special reference to the adjustment of grafts over eyes, and to other local specificity of integument.... 353 Cranial ganglia and the lateral line sense organ in Amblystoma punctatum. Experi- ments on the development of the......... 421 Cytolysins. III. Experiments with sperma- ROLOsATINS © SLUGICS OD". <2, 0oeen socemere oe 207 D ECAPITATION ontherate of CO2 output or ertals Orthoptera. Theeffectof light ; Bile ve pevs a cic Om a siege volte apart ctwaia cere ae 4 Sera S. R. Experiments on the trans- plantation oflimbsin Amblystoma. Fur- ther observations on peripheral nerve CONNECTIONS: =e seam tole se een een 115 Development of the cranial ganglia and the lateral line sense organs in Amblystoma punctatum. Experiments on the........ 421 (Didelphys virginiana). Studies in mamma- lian spermatogenesis. I. The spermato- genesis of the opossum.. Leas Dytiscus marginalis L. Onthe respiration of. 335 | Pag and to the local specificity of integu- ment. The transplantation of skin in frog tadpoles, with special reference to the adjustment of grafts over.............. 353 prtieg S skin. The temperature senses in = ANGLIA and the lateral line sense organs in Amblystoma punctatum. Experi- ments on the development of the cranial. 421 Guyer, M. F. Studies on cytolysins. III. Experiments with spermatotoxins........ 207 OGUE, Mary Jang. A comparison of an amoeba, Vahlkampfia patuxent, with LISSUe-CULLNTeICelIsS.-. a25) ean tes, ca 1 Hydatina senta. Relative nuclear volume and the life-cycleiotee sce octane gecseaeee 283 NTEGUMENT. The transplantation of skin in frog tadpoles, with special reference to the adjustment of grafts over eyes, and to the local specificityiof-.2- «2-02 2nee be 353 IFE-CYCLE of Hydatinasenta. Relative nuclear volume and the.................. 283 Light and decapitation on the rate of CO» out- put of certain Orthoptera. Theeffectof.. 47 Limbsin Amblystoma. Further observations on peripheral nerve connections. Experi- ments on the transplantation of.......... 115 Vee AN spermatogenesis. I. The spermatogenesis of the opossum (Didel- phys virginiana). Studiesin.......... 13 Minnicu, DwicurE. The chemical sensitiv- ity of the tarsi of thered admiral butterfly, Pyrameis atalanta Linn.................. 57 MorGan, ANN Haven. The temperature senses in the frog’s skin................-0- 83 ERVE connections. Experiments on the transplantation of limbsin Amblystoma. Further observations on peripheral..... 115 Neural tube of Amblystoma to a position at right anglestothe normal. The effect of transplanting a portion of the............. 163 Nicuouas, J. S. The reactions of Ambly- stoma tigrinum to olfactory stimuli....... 257 Noyes, Besstr. Experimental studies on the life history of a rotifer reproducing par- thenogenetically (Proales decipiens)....... 225 Nuclear volume and the life-cycle of Hydatina SENG e mM EVe abl Veae cera eras ce mccsiclointereueacie wre 283 LFACTORY stimuli. The reactions of ‘Amblystoma tigrinum to................ Opossum (Didelphys virginiana). Studies in mammaliag spermatogenesis. I. The spermatogenesis of the.................0e 13 497 498 Orthoptera. Anestheticsand CO2 output. I. The effect of anesthetics and other sub- stances on the production of carbon diox- Ley DYyaCenbaiMeees. Aachen strate eee _. B20 Orthoptera. ‘The effect of light and decapi- tation on the rate of COz output of certain. 47 Output of certain Orthoptera. The effect of light and decapitation on the rate of COz.. 47 Output. I. The effect of anesthetics and other substances on the production of car- bon dioxide by certain Orthoptera. Anes- thetics: and! COs: sce ss acetone ations AINTER, Turopnitus S. Studies in mammalian spermatogenesis. I. The spermatogenesis of the opossum (Didel- phys svirpiniwana)) ceisler ee eels eisisese oye 13 Parthenogenetically (Proales decipiens). perimental studies on the life history of a TObMer TEPLOd MEINE Wades ee eee ete ers 6 Peripheral nerve connections. Experiments onthe transplantation of limbs in Ambly- stoma. Further observations on......... (Proales decipiens). Experimental studies on the life history of a rotifer reproducing parthenogenetically. 7... i: ames esses = 2 225 Production of carbon dioxide by certain Or- thoptera. Anesthetics and CO: output. I. The effect of anesthetics and other SUbstaNCEs ONGthe > neha cereale, stele 323 Pyrameis atalanta Linn. The chemical sensi- tivity of the tarsi of the red admiral but- Gently: Pee oe cites iittala ote mate ctet roars etaterels 57 225 115 EACTIONS of Amblystoma tigrinum to olfactory: stimuli, “hele... -.- ss 257 Red admiral butterfly, Pyrameis atalanta Linn. The chemical sensitivity of the Past Of, Heh eerie Meretcte rts tos ewes, asavais amy 57 Reproducing parthenogenetically (Proales decipiens). Experimental studies on the lifevhistory of arotifer,! 5.27.0. s~ sie sodas 225 Respiration of Dytiscus marginalis L. Onthe 335 Rotifer reproducing _parthenogenetically (Proales decipiens). Experimental studies on the life history Ola .ccsssc cere oe sees 225 ENSE organsin Amblystoma punctatum. S Experiments on the development of the cranial ganglia and the lateral line....... 421 Senses in the frog’sskin. Thetemperature... 83 Sensitivity of the tarsi of the red admiral butterfly, Pyrameis atalanta Linn. ‘The CHEMICAL cp centers Saleaiard 4 japsietsraste iS gakenisie 57 Sauti, A. FRANKLIN. Relative nuclear vol- ume and the life-cycle of Hydatina senta.. 283 Skin in frog tadpoles, with special reference to the adjustment of grafts over eyes, and to the local specificity ofintegument. The _ transplantation of..........-.---+---.2+-- 353 Skin. The temperature senses in the frog’s... 83 Spathidium spathula. I. The structure and behavior of Spathidium, with special ref- erence to the capture and ingestion of its PLEYEL SUUGIES Ome chyetelseieiel: oe. 7 oy -— ates Sbsisscssteegt ete tetiteseiees cate ele tet weresere pees ere ere th Stateless os — 4 6 — 6 ~4— 4 — O46 et 8 isresseit ; petetetecatetatats tetitateys 3 Seiebeiisiebsiss + #32 3 ee sses esi sist sts = Pees Sets Ssssss ess eo sees = pits ere ees > Soere tS z preteweretree scesssaseagtecoesessererstacsesisestscses : SETHE ies sat eees tet set eecrrses acts Ssdsicteisesss = pha el shad sat Soeprperesebe terre rererrrorrs sraee Rode bebals tere tats es tebe be serers- > 5oateseseaeeeles : DIP{ES IETS THREE AES eet ase : = ne Sapetese Tape srersrer we esse Perec cremnen rere es SPsisies hse hitstsis labs tstsbsesbsri sews esranseersrsbarsse earns ~ t eos epesrsesetesepes Bites cane efi tite he Leele tate: Sts ereptrspesebest reseed piecssaesesoeessseseastes atetstececes ere AMMMNMISH ts Seprcoecre seer rere reset pes eeee it Teele tse = ee Peprerpreee te dtare ee reeeee. = pepeseeesen tes. O— 0-5 — Sellers reese see ste site * eek it “Hi heh tics eae pee SSS srsteeee sPsre sass : Srreee tapes sees e sepa pes pepepepeetes + = : Sbstere ethos LE RM OL ELT I OG SM ON ee . 2 yas, PIP a5 Saree “st << Sristpeseet Si etherseeratetebrsebire > Bese eesertr tees estate eee peers SSS SSS SS : pe yipestsbebsistisss. Siresieet pres Sertiscssss = > > Sessse Detsbabpipesesapiprsspapesreds ribeiseses es eehe Srasiseahiates yt bissaees Ssoiesetestse: : = Shandon —O— 4-8-4 rer ee eres a hone adh aneianetean —_ ere be oered ~ . Debsepeprerseersepepisrssrribssere Parte sitets ise SS se “. Syrsoste> hare eer eee a teeta ee oe ten pt pe se se Seppe eee See is esis selstreateleeteseee pores erases Sisecsaeasars its s— es pe Sese sss sos bette eeectete ts Srp lss 5. — jboss ee (baka eee rere peed ened ae heeded and anal soa oad ere rere - 533 SOT Eb lils ts Sahat pr reese sage pesete seer prea yep epe separ ee seer epee eee Pen eee eer EN SOSe ee pes ese ee : Sree eS b 4-00 - 0- 0-9 02s Tt ss Bat eats pre=s > Bebe ep tsetse ers es ese e tp bee sep pe sy ee reeset ~ ‘ Si harpetecscrsetwewprsrser sie sep es rete ter preopl pr peer per eere ete pes sys eseteees : : sap be bers beecre wre gepese peters et nt et epep ee eye pe pease ooo 56 6 5 86 0 tS St 8 Pee a wean we ase = ce Senn tee : : 5270 oa ss Tiltrss Sie ee tee eet es eee eee iaiese + Sst sess shsi5> Fe o5a see bsenrne zt eiaee pepe repr. Sate 3 eieeesh ere $ <2 poang Faseertta ses hietepepeersepe pr site sesrie se : £53333: tab 8— 6-5-8364 5 33 S53533 eee Sees bebelr tra Setetse pepe eres Se teeas, > Sapo eee >=. peesierergrinprsreepesistets. ereiprerprer 35se5 > peiais a ete Peete tete: ate Le tees = seetete Me eats ete eteteewtetr ssi o tet itieeee site tetare see 533 ; pteter ere aters tere ete seee esr eeersrsrersters Estiratgisrgeprissesapesesesey : p< esos! PHS N SUSE PSS ee shart ase 4 -6- pe tae (obetebebenseeeste - - —— ps 64 — 00 bee > =s TSR MSA pepesesessi Seereree p55 58 6-8 36s tet > RRR Ra eee Ree eee ee pees ese sere ss Dei Se ee owe rorees rt Pere tele e sleet eee es eee ewe ees esate ebeeeiee nese oS 9 3 9 6 8 3 5 = STgeebareret rites sefetetetsists tet etet eset are teteret state sertsirse she beers ep earls ecbt ase epee ree ae eae 0-004 44-0 -o- 4-5-6 ee ae re ra rnd be eee awe ree it ait Pte el) Sie ear oes eee hee ee tee ee tee ee een OTe a Se eee ee wie ; - a ewe a Swe w ee pape eee eee 3 —2- eBosteeesseseenenee * Tithee }ejere base = SEerat HIST bebrritsisesscsy i> 2ss Ss ep isi pretest crest, ir eri Se toer eee : - i + rats pret heh hh rots i + jet 352535 rtasetseses =. = 6: £7 + s Riresttistes figiseresr esetetis sisiseshpepeshsepessse ‘ eit ie ajateiaie — PROT ert ert ee set 6 o— a Pets leie Sette ee. 8 >i tt 6 64 Tap ree Oder tee eevee eee Ree eee swe eee Mest ee ee S55 Stee *. n6 —t—0= = Sp by bee ae eee we weweee Peele eee tee ee ese at ecabebeee pee eeaes Fae epeeee es > spabe to Lite ne le Better entepedere <3 Sst ty SNS et = bereheerrere elitstatt iste terete teseseta te satt es RESESOPste beter ereeeterrteerie ears =e pe" =n ae ae eee pana e. eases ey) Hatt REPL OLDER STR a ne Oe so pap bese ner ee enone we leer “ ene eee ee! ayer eeteteresrsperegerrsrsreree Sarerersrereer ests == a on dened oiedtendhendbened ere Sereprbebetieateteretere $ gH sit { $ + eeperes She bebe Serer t °—s77- S555 ——s>3 t i$ 6r$ 078-528-825 os Dre ee ete esa =Se ree gesee p83 — oo - ms 327 eek a re see aie ee ne een rns peer ee pep ewe ee eee eee <7 ae pe SOS SOC SS O88 8 Oe te ee ee ee eee ea enensebetebebeseteteaetee Ss ort tse pt ai ats sees : : wesboepe recess repr pep eer r etree pests area pee es eee ee eaeeen se etre TE hibseshsep ep eeee pbs epigepee bert eteios street ries ethers ene n esae Stee Sey eS SeEeES Eee ene se Seer Sebo bee beet see etre n ne ate th pepe pate Pe bn Bs Le pe ee Ree we we ee eee eee ere ee a a rn re ee rn an ne ee re a ne a Perea ee ee Oe peer ew ea ph eben snes e Shel ils les neste eetenane peneeerr tse sit iets sete sesh rene ern ene nw ween ews peers enesteasherecestscsleesbieserises resp hecsestoesessy. Pacepere sree reels rere ree erases Sabi hsesbetirsesesierstpteberstsratsts eh sesase Sele bicerineseepisesee = Pirate tercbersisigehcetetsbabsia re Deb eeeee Petgiiisistetstsipistessegerepepe pps > TSSSSs5 Sp tpipipep pists cPabrro sere rs Sebsbasseetsrpes (3355555555555 5555555 Siphiipipisterser trepepepeseyey Rete e a see eeeeteesecestsest sect ees: Sbphsetbe sts sbs tie sbresitesrpetes ee eeties ese eeererer Sigbserhsrewsbsts eebure Bets beh bs hsra rs csrsrarsspeeree prsipicipighst: Se bsrars tse Sos tiwgrersrcrsbprererprererpeseperry wetbereeeeette ss Dyess a banadantedobe-dad oe +: Ferree Peet ere ee re eee Shits Seebsret ests tsisr eess poreseseseres saree 3355. poser esse erise cesses 2s She sas kalibperceepsy 7o>e? eee sess PTS Tt atts tse pesierese sat: t SEs ofS SS SESE eee Saabs 3 ; : 4 Hi a i) + tite te e sehr pore! fisted eh +5 Spaperee Shs apesry ties bebs beri rs rsoy eeeeeeet epee rteeet=+-,- psp Se rebstelpreeseteceesprsrscpesey Sheep bee Behsbtistebsrsegizescshers His sets i 4 rt ‘} +F z u ty $: ry r + rT + eke ep wees a ekee pete hess te pete ee ee S232 25 eeireseses Soeerer = edn oe ds Fs bi ay Si ees oe ee tee .* Pee eeseese soso cesses FF pare ere eres eee se site sees 22 ee +09 — 83a 90> 08 Seek e nee a wee ee te Te SN SeSe eT Nhe tet eeee rant eh aneee Tebibyrseprs pepe pease eee ra wees esreee ASSESS EE Sei beba wep ep ep aces warren esr se wares et rsh anpor ee ee bane eeeery Deets SSS SESS TSE t= Srp ase see ep wer sesreeerne ee Petr ishtisitssssi en bse tales t : hie sth + {tits ihe oyeytt z aie to} th ! +) ha ier trisbsishrsstprisse ses S858 sbseprsisepesrsig pe peaeseee spapreeprsepepeseprsnpes= stp pet peer ener pe - Bibs atts SSS t ets ts tates sist eteteseetetete : pn ae ae e ee eerie te etitetetete Tosh slabs bsbabshsspeerseprcirenrpese Sopot ee te pee cece cabs escncsey ps rnwe deer tresribersbereree . : SESESESESE Nore tier es bere endarsspesteseeseyetree ‘ ; Sareerstabslsnsepeeesesbsry ep essaeaee berate ose =! a3: Sisisitestpspreepeee pee + ; - Parsee —? re: ‘ Sesh ere reeersrp rere bars revirsspsensesesrget rarer : Dsbieststieterer arses pebsisieretirersseratsbsbeereteespeseres tte = ; eee sibeseeseseree eee eteaealeetle. sos: Vere. ; Sbsbibnbshsbsbrrssseatssaniepspspaceresepisseepeprseansetates eee pears 4 3 Pedsestariestererersesrses ee te ea tsia tessa = astsisers : Seba otestiethssinrsiayesesopapesisis tele heen rears Seti teste tetets ; esrstsisectrshisstsrpesrerseietisrrttierstsiprsisrsrtestieteiestensreeteprse ses biesbsesbsrsratersr SSS SSS SSS SS SSS SSS SSS se : : eS53e2322: Sabsbslstereesterstsaeberestes tt ssrissesste es os. abs Sher > << thst are no S928: cS 535 +3 zy ote i setter Suet ¥ Sirbceetetetrtcerersr eet eee esecere 3 Sepestseseseeeaeaeeeseeerepipe ptpheree pert sestseeeere x FsbcBeechsee reece sts Ssexbeletasere ser: Riba tebarsrsbirsrsssherets Peete sesteaee et : ele lo! sheer Whi apeey 4 be rpacer tears bieitin : eratgapretrternarrcrsa pepe phpecraeeeesenespresrapesieapesasapere peteee ate > chs Sor aebebireepeees retary tere betebrectescleesherses soararsseeseses *: rarer tT SESS ESS SSS == eababrbebebetere tebe retetens Sbshse = Stet etetceceeere cereale = > bitsty $etstirsse SESESESTEDR IR SREK SEE 35 : Biestetiteerrereteeris Spritehphtetrieparssrsepesepapeeepepapetepetays Sits ae SEEN Ot eeleee tees bec seeaee Spar hess coreaer weep aeneaee SEs Cat ieee > oaees = veep eeees wee te aoe seeree ae Syetsestsrsre eee tess esesbpesiaests. - + oe eat ee hee ete 3 sity Loses e132 +42 s tenes ree EweeesEses? SPEER SR ESSE SeEC ESS Srbceey ~ = >: _ Seses. 2-4 Sesscesere; passsesssestesetrsssstressqesessstgeatstesssesstessestssststitstsesegestesststetssatscststssssssecesiseiissstetiesss <7 eieieie « 2~4—@ eee reser To! yeatet= pete re) rae es “« 3 <3 = 333