oe Cewed ones . a << : . Seren Creeterstuewen enchning ance . core wie ee bento ars tn a Seep eieetrerts ote pabwaa anna: ‘ ~i wediaeeteritugn trims sxlotwane ton nese soe ag hos Nod weap oerng wc rarng wee re : ‘ own ee “vo Pemaene: ara WOtrephuiens oe” nde one . tetova at . : S Wr: eB wape a ~ 7 - . oven ° “ - Aste men a - _ ri “ ’ we a Me -- ws “ . papas ; = : > . . noo . [ai Me ae . ¥ 4 2 atom, Soret wen “~ oar + ro au thn rte errs 2 we ? . Sat - : : =a Aint et Repatenand ee te eee r ; . : ; Chpemetheyheds Mi ana —_ . 4 we WN tniag thay ~ - ¥ “ Perri torn " < - - ; : : v4 . : Y ¥ . - iie—_1.... < Ber enema. AP vere ents . - ~ , ol Peg ° eArhrtealatior toe e . - . sere : - ena ¥ a “a + Pied op — eu % : ‘ — . uh Fisthcm aera : Oren : : me ion b . is wa — co . “~ rn vn . oa < echgiiehd aidan ie homie ee ° i ‘ epee tetra eece + "s Natoma TH Od RaD TE tee, ih siedbahaee Ree hntrenwin Ne chee To ete iets SP OPAL AY WP Vote te tabestoty Ae Ba eT t Spdphris a} tp letea, ded? - PORN at oes. on rn ; “ 6 4 ~ ’ DRT At irae os - - Ne? es . 4 See, aoe Y hae me Ar atery enim fF 4 : Vetetyiume — 2 . ; —. , Heart . Saha on 7 ~ - . . “ Y Viglen are Seoolhba BOP WEP NG bce pad x - ota 7 “ : pentoves Span : b , ya neg 7 r ; r Cae, ne - Jo> Aas 6ade% . oe segue 484 ey - . £ ; + , “4, . i et ee esting a - oe in 2009 with funding from University of Toronto itt ‘/www.archive. org/details/journalofexperim2o ies re iiss . C7 THE JOURNAL OF EXPERIMENTAL ZOOLOGY EDITED BY Winu1amM §. Caste Jacqugs LoEB Harvard University The Rockefeller Institute Epwin G. ConxkLin Epmunp B. WILson Princeton University Columbia University Tuomas H. Morgan Columbia University Cuarues B. DAvENPORT Carnegie Institution GrorGE H. PARKER H®RBERT S. JENNINGS Harvard University Johns Hopkins University RAYMOND PEARL FRANK R. LILuiz Maine Agricultural University of Chicago Experiment Station and Ross G. Harrison, Yale University Managing Editor De ks jo fae 22 VOLUME 29 ~go/¢ 1919 THE WISTAR INSTITUTE OF ANATOMY AND BIOLOGY PHILADELPHIA, PA. uO mt: AU Ava CONTENTS No. 1. AUGUST E. R. Hoskins anp M. M. Hoskins. Growth and development of Amphibia as affected by thyroidectomy. Nine plates (one hundred and nine Os oy A | PE Pe ee anna er re ae 1 Heiten Dean Kina. Studies on inbreeding. IV. A further study of the effects of inbreeding on the ie and variability in the body weight of the albino rat. Eight charts.. : 71 Haritey N. Gouup. Studies on sex in fir: ipenweirodite a lese menial plana. III. Transference of the male-producing stimulus through sea- wieniee, CnGae«ie tiie GRA: to laos apm aa obE oe otc elas: No. 2. OCTOBER Gary N. Carxins. Uroleptus mobilis engelm. II. Renewal of vitality through conjugation. One chart and one figure.. a Be Al Lesiig B. Anny AND W. J. Crozier. The sensory Pe aoades Go cman Fourteen figures. . he Ee | se NTS Soe cto rages LOM W. J. Crozier AND leeetcn B. inry. Seore reactions of Chromodoris zebra. Hight figures. . .. 261 Matsuzirno TAKENOUCHI. Studie. on the meputed cud akae aes AS ihe toymens gland (albino rat).@Mewo charts....:.....0ese.-eeees-esess-22. Ol No. 8. NOVEMBER Dwicut E. Mrinnicw. The photic reactions of the Fe eed a melli- fera L. Seventeen figures........ .. 343 Rospert W. HeeGner. The effoai Bile envir ronmental Piciers\ upon ithe Renik able characteristics of Arcella dentata and A. polypora. Seven figures... 427 M. E. Cottetr. The toxicity of acids to ciliate Infusoria. Six graphs.... 443 J. A. Dawson. An experimental study of an amicronucleate Oxytricha. I. Study of the normal animal, with an account of cannibalism. Twenty re ss] wy) lil s* 7 «aw ae e ee = re £ er 4 . Li = é » eA TO Vom yiigyal Baw badisid one ; agin q vibe Wini4 OVE BA! wed) vd boind ee (hs dialed RAd0 ae) 4 s reyes en rhithy| Sialy yh Bly te ad Steer: 0) Ok TUR Dee os aes ae ) eee See pee Fey Foden i ¥ A Aba jes ee a Te rEERO Sy STiciat asada eS mee De taeda io shar ARRe Re) ROU : 1 SMe OY GuNp. st Brtsit s fo, awa ae yak ald at wesePyvirkicAab rr ay ii gems « He a Cee f ; as Maree odd atters: asl ab it at vie a fe ‘sit upea Ty MARooe ) (> HE ge ole mete hie try 1). 8) CT PCT ae! ee ) leew) rLit we og) 7 eee Tea mY 3sih if) } £ . Hany olat Apert icy: LED iT SED Oe Kamae OLY De) aber St ie Ae niioedtie ier 3 ray ie cit eee WETS ABT RE 1) eR “BY iat bg ™ Hider: Oi h+ y PUK 3 Raye eu) eteathinn Oxe Te ate)’ re bs ete) easier tl x4) Lith ‘\ Pad eg Mabe ero torts veal «pieces i 1 ok ae isan hg dot i AO Re ri ieeiion Chaiotth (as ake al ss} aah cicaEretgye veil ep SawGl: Tet, an! re we Lanttgon wan etal” -< SA Dine iy tps Weil ig t he 8 Dk isle ars PIE ey ot | fa d ‘ > : 4 ear Tae : Lj q * ae ids a : j ‘ i pe ‘ \ f wa! Gan Ws . Resumen por el autor, E. R. Hoskins. Universidad de Pittsburgh. Crecimiento y desarrollo de los anfibios afectados por la tiroidectomfa. El autor ha extraido el esbozo de la tiroides en embriones muy pequefios de Amblystoma punctatum y Rana pipiens, des- cribiendo la técnica empleada. Cuando se obtiene éxito en la operacién se impide el desarrollo de la tiroides, pero en algunas larvas la glindula se regenera. Cuando la tiroidectomia se ejecuta correctamente se impide la metamorfosis de los anfibios por lo menos durante dos estaciones (un afio y cuatro meses) siempre que se alimenten con una dieta normal. Las larvas desprovistas de tiroides crecen mas rapidamente que las normales y finalmente adquieren dos o tres veces su tamafio. La cesacién de la metamorfosis se atribuye a un metabolismo defectuoso, especialmente del calcio, puesto que los primeros y mas sor- prendentes efectos se notan en el esqueleto, que se calcifica y osifica menos completamente que el normal. Otros caracteres de las larvas desprovistas de tiroides son los siguientes: Retienen la capacidad de regeneracién de partes perdidas; el cerebro crece despacio y se diferencia atin mds lentamente; el higado se transforma en un Organo relativamente grande y se diferencia parcialmente en su forma; la hip6fisis experimenta hiperplasia; el timo persiste y se hace relativamente grande; los cuerpos epitelioides y los rifiones aumentan relativamente de tamafio en estados posteriores; el bazo no es afectado por la operacién; las branquias internas persisten; los pulmones se desarrollan y hacen funcionales; el intestino retiene su caracter larvario; los ovarios crecen, desarrollandose grandes ovocitos sin tener lugar la maduracién y los testiculos maduran completamente, Pro- duciendo espermatozoos que salen al exterior después de atravesar los rifiones. El autor describe los cambios de tamano de los organos de larvas normales durante la metamorfosis y también la histologfa de las gléndulas endocrinas y las gonadas de las larvas normales y tireodectomizadas. Translation by José F. Nonidez Carnegie Institution of Washington AUTHOR’S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, JUNE 2 GROWTH AND DEVELOPMENT OF AMPHIBIA AS AFFECTED BY THYROIDECTOMY E. R. HOSKINS AND M. M. HOSKINS Osborn Zoological Laboratory, Yale University, and Department of Anatomy, New York University NINE PLATES (ONE HUNDRED AND NINE FIGURES) CONTENTS MeTEMENANTCLTOTE gece S ccic a fhenh denies AM +! oj cehets SORA ST Te eiale Shrslow stash fare 2 Bes Mee it, ba cc ce SE. 00 ee So suebe gap Seat) Results.. bd es MS: 5s ae open eee eee Microscopic euicues Se cis cM - «oe cell Sai Mees brevets o's ta stakere Detatatolels aye, ove 4 LOSE 0 Se ene. = ny Se eae eo SO era 5 Sera Giler DUGY:. ..2s°,.. MEE... cee Lae ey Scteleemiels. ss oa bengethiand volume..:... MMe. -. (eRe tes Sle pee oats Sine wale 22 Gir ee so ha a =e le! CS hae ep eh Se ee =< 8. Pee eee 2 UREEIEVRIGS. iat toc. c's... . MEME. «<> « ya eC To Sage bee a baa Thymus.. Pe So ee Epithelioid (parathyroid) b Biles... ch A ena oe 32 Thyroid... Re Nae, Re eek MIGNGy Sox, . 2s sag TRE... . « APRs bcd oa ky Gales sag eens 36 Ealigiand hitigy. 2. .ee eee ....' 1 aeeee os co. Jas ances emeaeee gor PMGaGiNCs.,.... i) SSC SERan GE. . . . cGh SRR Ree on «Seek iesia’s Pee ee ele 38 Adrenals Eee ee ee One! | Ovaries.. er: «= ee Pere 2 a ais Testes.. Teen We Le Re aS! Summary and SoMebnsions.......... an ee, een: 46 Ue ee OO be 2 E. R. HOSKINS AND M. M. HOSKINS INTRODUCTION One of the experimental methods most frequently used in the study of the thyroid gland is that of extirpation. Mammals have usually been employed in this study and have usually proved unsatisfactory for definite results, especially where data on the relation of the thyroid to growth processes were sought for. In most mammals complete thyroidectomy is followed shortly by death. Moreover, in such animals the gland cannot be removed until after it has functioned fora time. Still further, growth processes are relatively rapid in the higher animals. We thought that if use were made of a lower form wherein growth processes are relatively slow and wherein a larval existence would permit extirpation of the thyroid before it could possibly have become functional, a new approach would be opened to the problem of the relation of the thyroid to growth and to life processes in general. With this in mind, we selected amphibia for our material. An added interest lay in the fact that Guder- natsch (’12, 714) had clearly shown that the opposite sort of experiment, i.e., feeding thyroid substance to larval frogs, alters remarkably their growth processes and hastens metamorphosis. The forms selected were Rana sylvatica and Amblystoma puncta- tum, which are very abundant about New Haven. The first operations were performed by one of us in April, 1916, and in the spring of 1917 and 1918 the experiments were repeated. We learned in November, 1916, that Doctor Allen, of the University of Kansas, had performed similar experiments with R. pipiens at about the time when our first were made. Doctor Allen’s and our own work were done independently of each other, neither of us at the time knowing of the other’s experiments. We wish to thank Prof. R. G. Harrison for the kind interest he has taken in the work and for his very helpful suggestions. Abstracts of different portions of this paper have already been published (Hoskins and Morris, 717; Hoskins and Hoskins, 718). GROWTH OF AMPHIBIA AFTER THYROIDECTOMY 3 TECHNIQUE The best larval (or embryonic) stage for a study of this kind is that just preceding the beginning of the circulation of the blood, as at this time hemorrhage is avoided and the chance of regeneration of the removed tissue is less than if younger stages are used. The R. sylvatica larvae at the time of operation were 5 to 8 mm. in length and the Amblystoma larvae were 10 mm. in length. The animals were anesthetized and were operated upon in chlorotone as recommended by Harrison (’04). The solution used was 0.02 per cent chlorotone in 0.3 per cent sodium chloride. A layer of paraffin blackened with carbon was put into a Syracuse watch-glass and in it were made six grooves, each sufficiently wide to admit an embryo lying ventral side up. After the thyroid gland was removed the embryo was left in position for about an hour or until five other operations had been performed, the first then being removed to make room for the seventh. By that time the wound had sufficiently healed to permit handling with a pipette. Asepsis was found to be unneccessary. The operations were performed under a binocular microscope. A transverse incision was made with fine scissors through the ectoderm posterior to the oral plate, the flap of ectoderm covering the thyroid anlage was pulled caudally, the mesenchyma pushed laterally or removed, and the exposed anlage of the thyroid was removed together with that portion of the ventral wall of the pharynx surrounding its proximal margin. The thyroid being opaque is readily distinguished from the translucent mesenchyma. Care must be taken to avoid injury to the pericardium and heart. Usually the wound appeared to be healed in two or three hours. After the operation (one to two hours) the salt solution was gradually diluted with tap-water. It was found that the em- bryos were harmed by being left too long in the saline. Many sylvatica larvae were kept together in the same aquarium, but the Amblystoma larvae were isolated to prevent their eating one another’s gills. 4 E. R. HOSKINS AND M. M. HOSKINS RESULTS Amblystoma—1916 Growth. There were no constant external changes evident among the thyroidectomized Amblystoma larvae in the three months after the operation during which the animals were under observation. At the end of this time all those surviving were killed for microscopic examination. The gills and legs developed normally. The average length of the seven surviving controls at this time was less than that of the seven experimental larvae, but the number of specimens was too small for final conclusions with regard to growth rate, especially since the largest of the fourteen was one from which the thyroid had been removed. The largest thyroidless larva was 80 mm. long and the largest control 24 mm. One probable source of variation in growth was the food supply, as the animals were isolated and may very likely have had unequal amounts of food. Microscopic changes. Four days after the operation the ani- mals had attained the length of 12.5 mm. Sections. of control animals show that the thyroid at this time is long and narrow, lying ventral to the pharynx and anterior to the heart. In the animals from which the thyroid had been removed the floor of the pharynx healed without regeneration of the gland. The hypophysis was not structurally different from that of the con- trols in this stage. | Three months after the operation there were still no differences to be observed between sections of the control and experimental Amblystoma larvae, except of course absence of the thyroid from the latter. The cartilages and muscles ventral to the pharynx developed normally despite the removal of mesoderm in this region. The thyroid glands, which were lacking in all the experimental animals were well developed in the controls at this time (figs. 1, 2, and 5). The vesicles are found scattered along the ventral aortae. Colloid is abundant, although the cells” have very little cytoplasm, the wall of the follicle being very thin in most cases. The gland is about 0.4 mm. in length and of variable shape. Since none of the thirteen experimental GROWTH OF AMPHIBIA AFTER THYROIDECTOMY 5 larvae which were sectioned showed any thyroid tissue, it is prob- able that the operation was successful in most if not all cases. The hypophysis (figs. 3 and 4) three months after the operation exhibited no peculiarities of structure in the thyroidless animals that could be attributed to the operation. Micrometric deter- minations averaged practically the same as in the control animals when allowance was made for the size of the animals in each case. The hypophysis shown in figure 3 (control) is smaller than that in figure 4 (operated), but the entire animal was also smaller. The cells of the hypophysis are arranged in clusters in both larvae, but there is no other evidence of functional differentiation. R. sylvatica—1 916-1918 A comparison of figures 6 and 7 of R. sylvatica of 6.5 mm. in length shows the longitudinal dimension of the area of ento- derm removed in the operation. This included not only the entire anlage of the thyroid gland, but also a considerable portion of the floor of the pharynx and a part of the surrounding mesen- chyma. The anlage in this stage is comparatively large. It ex- tends caudally and ventrally to the pericardium, but with care it may be removed without injury to the latter. Blood-vessels re- moved during the operation usually regenerated. The hypophysis (fig. 6) is already ventral to the infundibulum. It does not appear in figure 7 because of the orientation of the specimen for sectioning. Figures 8 and 9 of larvae four days after the operation show that the floor of the pharynx and mouth have been completely regenerated. The thyroid which is well developed in the control animal (fig. 9) is entirely missing from the operated larva (fig. 8X). The hypophysis is well formed in both specimens and shows no difference in size or structure at this time, although in older thyroidless tadpoles it is found to have undergone hyper- trophy. No other changes in the organs are noticeable until the time of metamorphosis. In the larva represented in figure 8 and in many other speci- mens the removal of mesoderm with the thyroid anlage had no effect on the development of the muscles and cartilages of this 6 E. R. HOSKINS AND M. M. HOSKINS region. In some cases, however, some of the muscles failed to develop properly, and this has an important effect on the growth of the animal. This will be discussed later. Among the frog larvae from which the thyroid had been re- moved a few animals developed abnormal external gills, the shape of which was variable. Some of these were low and flattened, some long and slender, but irregular, and with a vari- able number of filaments. In other specimens the filaments were club-shaped being enlarged at the distal ends. The circu- lation through these abnormal gills was often sluggish. One experimental frog larva did not develop any external gills, al- though it lived and grew otherwise apparently normally through the period in which external gills normally persist. These bran- chial malformations were considered to be the result of vascular injury inflicted during the operation. Ekman (’13) states that the blood-vessels are necessary for the later development of the external gills. The twisting of the tail which is common among frog larvae kept in aquaria occurred in our 1917 specimens which were not under optimum conditions. It was also noted in the regeneration experiment described below. It did not occur at all in the rapidly growing 1918 larvae. The abnormality is not due to hardness of water, but to food conditions, for the same kind of water was used in both years. Growth. In our 1916 experiments it appeared that the average growth rate was less in the operated animals than in the controls, and it was so reported (Hoskins and Morris, ’17). The reason for this difference in the average rate of growth we later dis- covered to be due to two chief causes, and the real difference is just the opposite from what it appeared to be. In some cases the operation injures the developing lower jaw and its muscles, especially the submaxillary; and these animals have difficulty in eating, and suffer from partial or complete inanition. In other animals the thyroid is not completely removed, and re- generates, and these animals grow at about the same rate as the controls. The experimental larvae may be divided into three groups: those with defective jaws or muscles which grow GROWTH OF AMPHIBIA AFTER THYROIDECTOMY 7 less rapidly than the animals of the control group; those with regenerated thyroids which grow at about the same rate as the controls; and those which have normal jaws and muscles, but are devoid of thyroid, and which grow more rapidly than the normal animals. The growth rate of frog larvae kept in aquaria varies greatly, and the most important factor in this is the kind and quantity of food; but the size of the aquarium is very important, because it is easier to keep large aquaria balanced than small aquaria. The influence of the food and size of the aquarium is shown very clearly by a comparison of the growth of our 1917 and 1918 animals. i In 1917 fourteen series of extirpations were performed with about 700 larvae. These were kept in small aquaria and fed from April to the beginning of June upon a plant obtained from a local aquarium supply dealer, and to this diet were added occasionally cracker crumbs and bits of liver. In June the diet was changed to spirogyra, and the animals were transferred to larger aquaria. The growth rate rapidly increased when this change was made. The average time of metamorphosis of the control animals did not arrive until the first week in August. Under optimum laboratory conditions R. sylvatica undergoes metamorphosis in June or July. In 1918 the animals were exposed to the sun for an hour or two each day. In each aquarium were placed small plants and a layer of mud and old oak leaves from a pond of stagnant water. The diet used in 1918 was composed principally of algae found floating on the surface of a stagnant pond. Taken with the algae were large numbers of small crustaceans which the tadpoles may have eaten after the crustaceans died in the aquaria. Only small amountts of the algae should be put into the aquarium at a time, because if added in large amounts it will die and poison the tadpoles. In our 1918 experiments the animals grew much more rapidly than did those of the previous years. Whereas in 1916 and 1917 the larvae required a minimum of three months to reach their maximum size and begin metamorphosis, the minimum time required by the 1918 larvae for this was only 8 E. R. HOSKINS AND M. M. HOSKINS one month—a time much less than that usually required for R. sylvatica kept in the laboratory. In the former series the maximum time required for metamorphosis to begin was four months, while in the latter series the corresponding time was only seven weeks. The growth of our experimental and control animals seems to have been different from Allen’s (’18), who states that he observed no difference between the controls and thyroidless larvae until the time of metamorphosis. This point we studied very carefully by taking hundreds of measurements of the grow- ing animals. As shown in table 1, we found that in every one of the experiments except one small series (5/5) in which the thyroidless larvae all had defective jaws or muscles, the largest thyroidless larvae grew more rapidly than the largest controls. The same was true also of the average after elimination of those suffering from inanition. For a few days after the operation the larvae grew more slowly than the controls, but passed them at an average size of 18 mm. (range 12 to 23 mm.). At this time there were alive 260 operated and 218 control larvae, so the fact of more rapid growth of the former seems well established. The 1918 series were not measured at this stage. This difference in growth rate could not be accidental or due to the food supply, because it was constant in every series from 1916 to 1918 larvae, the latter of which were especially well fed. By the time of metamorphosis of the controls the largest corresponding thyroidless larvae had attained the length of 58 mm., and in every series those thyroidless larvae without defects were all larger than the controls when the latter had reached their maxi- mum size. One control larva became 53 mm. long before meta- morphosis and one reached the length of 50 mm., but with these two exceptions none of the controls ever grew longer than 48 mm. before metamorphosis, the smallest being only 40 mm. long, and the average 46 mm. before they became transformed into frogs. In general, the 1917 larvae, which grew more slowly than the 1918 animals, became larger than the latter before metamor- phosis, but the largest in 1918 were larger than the average of GROWTH OF AMPHIBIA AFTER THYROIDECTOMY 9 TABLE 1 Showing that the removal of the thyroid from small frog larvae hastens their growth. Thyroidless larvae which do not have defects of the lower jaw or its muscles grow more rapidly than the controls R. sylvatica a °o waEN op.| quar er, | Numman | SG] LeNors ete. ene une EXCEED Pg aie a 2 WAPOE CON Ideas ea Gia Or DATE OF OPER. ae . CON. ‘ A 2 S MET, MM. ig Sata Z Op. | Con.| Op. | Con.| Op. | Con. Op. | Con 1916 nx. nx. 1 nied 1917 RUE ig ets Sgt 5 x ¥ 318 17.0/15.0) 8) 9] O 2 531 PE Te ee. 5 a aio 0 t 18.5]18.0} 10} 20 Ole | gaa aaa TX nx. 2 0} 21! 50 Killed at 56 os 121 Re eae 2 OME S12 ee en 1 42 | 40 LCC Cae eee 15.5|14. 5] 50)" 18}, 9 0) 1) 47 Killed at 62 RM NA oo ce orcienc len «0 7 DIL Ol. eenieeedl eee OV 52, 58 68 PTAOUNMED Soo sinew sess ss 19° 5119.0) Iie sit Nek 55) 45 70 Merb? oo... woes (2L.019.0} 8) 22 J. 074 | OA ey 23.0/20.0) 15) 5) 1 0} 9 | 59 68 Poa eye os. os. -(20.5,19.0] . 34 24) 3 44 2} 49 | 48 67 WA Ue a Oe 19.0/17.0} 40} 30) 7 Sled | O4a ea G2 Wiitvirol teenies. (198. 7116-5) 2ihe22r 5 3 Oo | Sean 68 bt age AS A ae a 17:5/18.0) “S| *6 is AG. <3 Se ete 2925129: 0|--ediee6| 1 6| 1 | 53 | 50 72 1918 0) 4 ie be 30.0/23.0) 9/55) 9] 55) 7 | 55 | 48 | Killed at 65 o/.\70, 1g: >) 2) en 26 nx. nx. 11 | 30} 11 | 56 | 46 | Killed at 56 Wi aigegt ee 8... tk Lee 38.0/32.0} 29} 13] 28 13) 26 | 48 | 44 Killed at 48 Aver. max. { 1916-17... ./18.3]17.2} 260) 218) 43 | 21) 31 or total \ 1916-18.... 91 | 119) 75 | 58 | 531 72 N. B.—The 1918 larvae were not measured at the 18 to 20-mm. stage. The 5/5 group of thyroidless larvae had defective jaws. Abbreviations: Aver., average; Con., control larvae; Op., thyroidless larvae; Nx., not measured; Mazx., maximum; Met., metamorphosis or metamorphosed. 1 Only one of all the control larvae studied became over 50 mm. in length. 10 E. R. HOSKINS AND M. M. HOSKINS 1917. The frogs which formed from the larger larvae were larger than those formed from smaller larvae. In Rana sylvatica | there is thus a great variation in the size of the newly formed frogs. As shown in table 2, the body length of these normal frogs of the same stage of development varies more than 25 per cent and their volume nearly 100 per cent. The mortality in 1917 was very great, but was abo the same in both the thyroidless and control animals. However, 42 of the former and 20 of the latter reached the period of normal metamorphosis. Of these (table 1) the controls all metamor- * phosed normally. Of the operated animals, 12 metamorphosed and 30 failed to do so. By autopsy and sectioning it was later ascertained that every one of the 30 was devoid of thyroid, and in every one of the 12 the gland had regenerated after the oper- ation. In the 1918 work, out of 47 thyroidless animals which reached the period of normal metamorphosis, 44 failed to meta- morphose and in only three experimental larvae did the thyroid regenerate permitting metamorphosis. In the three seasons a total of 91 experimental larvae were reared to or beyond the time of normal metamorphosis, and of these 75 remained in the larval condition. In addition to these, many which were killed in the early stages for purpose of study would doubtless have survived long enough to be added to the 75. From these experi- ments and those of Allen (718) we are justified in the conclusion that removal of the thyroid from young frog larvae will delay if not entirely prevent metamorphosis under ordinary laboratory conditions. Most. of the 75 thyroidless larvae just referred to were killed after it became evident that they would not develop into frogs, but a few were kept alive. Of these, six survived the winter and the second season of normal metamorphosis with- out becoming frogs. It is more or less common knowledge that if Amblystoma larvae kept in aquaria are not properly fed, they may fail to undergo metamorphosis their first season, but do metamorphose the second season. The same thing is true also in nature. Dr. T. G. Lee, of the University of Minnesota, has told us that he has collected in the early spring Amblystoma larvae which should have metamorphosed the previous season. GROWTH OF AMPHIBIA AFTER THYROIDECTOMY 11 Normal Rana sylvatica larvae kept in aquaria and: well fed metamorphose normally, but with our thyroidless frog larvae the second season was passed without their metamorphosing. Concerning the growth of the thyroidless larvae, it may be noted from tables 1 and 2 and figures 10 to 19 that they ulti- mately became much larger than the control larvae. As stated above, the thyroidless animals grow more rapidly than the con- trols, but the difference is not very great until near the time of metamorphosis. When the controls at this period have nearly reached their maximum length, their growth seems to slacken, and their length does not change for a short period of time before it starts to decrease during metamorphosis. It is during this time, as was noted also by Allen (718), that the experimental animals gain very much in size over the controls. In some of our 1918 series measurements of volume of the animals were made at this time and in some cases (table 2) the thyroidless larvae were more than twice as large in volume as the control animals of the same age. The volume was determined by the amount of liquid displaced by an animal. © After the metamor- phosis of the controls (1917) the growth rate of the experimental larvae decreased greatly and growth nearly ceased during the winter, but the larvae grew more rapidly during the second spring and summer. For example, the animal shown in figure 19 became 55 mm. in length by the time the controls had meta- morphosed in early August, 1917. It was 62 mm. in length on September 2 and 66 mm. on September 27 when next measured. On October 19 the length had not increased. The animal began growing again in May, 1918, but it increased its length by only 5mm. by June 25, and on July 5 had reached its maximum length of 72 mm. with a volume of 2.38 cc., or nearly three times that of the normal larva at the time of metamorphosis. In this growth after the first summer the body length (nose-anus) in- ~ creased 2 mm. and the tail 8 mm. the animal thereby becoming relatively long-tailed. Its hind legs increased after the first season about one-half of a millimeter in length, reaching 5 mm. In these miniature legs the normal segments and digits are slightly differentiated. 12 E. R. HOSKINS AND M. M. HOSKINS One 1917 operated larva deserves special mention. Its thyroid regenerated and then hypertrophied. The animal metamor- | phosed about a month before the controls and after metamorpho- sis it was less than one-third the normal size (fig. 20). The length of the controls at this time was 40 mm. (fig. 21). An attempt was made to duplicate this experiment in 1918 by cutting to pieces the thyroid anlage and leaving the pieces in the animal. In other specimens transplantation of the thyroid was performed, but in neither group did these experimental larvae metamorphose earlier than the controls, although a few were smaller than nor- mal. In three 1918 specimens from which the thyroid had been removed, the gland regenerated and the animals metamorphosed at a size about one-half that of the average control larva, but not at an earlier date. Figures 10 to 19 show the changes in body form exhibited by the control and thyroidless animals. Figures 12 and 13 show a normal tadpole in the process of metamorphosis. The hind legs are well developed and the skeleton of the body has begun to acquire the adult shape. In figure 12 it may be noted that the head is flattening and narrowing. The larva shown in fig- ures 14 and 15 is a thyroidless specimen of the same age as that of figures 12 and 13. Its total length is greater than that of the control larva, the hind legs are very small and the body (fig. 14) is more nearly cylindrical than that of the control (fig. 12). The eyes are more laterally placed in the control larva and, although its head is narrower than that of the other (fig. 15), the eyes are further apart. The head of the control speci- men is slightly more pointed than that of the thyroidless larva and is more like that of the frog (fig. 17). The experimental animal retains this larval form for some time, but it slowly ander- goes changes in shape which tend toward those of the controls. Normally, during the period of metamorphosis, the form of the animal changes as follows: The body decreases very slightly in length; the anal canal shortens 2 to 4 mm. (compare figs. 12 and 16); the transverse and horizontal diameters decrease nearly one-half; the tail atrophies; the dorsal and ventral sides flatten; the eyes come actually closer together, but are relatively farther GROWTH OF AMPHIBIA AFTER THYROIDECTOMY 13 apart; the rima oris increases in size and changes in shape; the nostrils become relatively more lateral in position and larger, and the legs increase from one-half to two-thirds (table 2). The thyroidless larvae after the season of normal metamor- phosis still resemble young larvae in general shape, but they TABLE 2 Length and volume of larvae and frogs. Showing the variations in the length and volume of living and fixed larvae. The animals were fixed in picro-formo-acetic and preserved in alchohol. Note the small size of the frogs, which are only one- third to one-half the volume of the larvae from which they are transformed. Vari- ations in shrinkage occur during fixation. THYROIDLESS LARVAE CONTROL LARVAE IN | procs IN ALCOHOL mee fT Alive In alcohol Length Length Length Length Volume|___+— > —) | Volume) 222-2 S22 Volume]. __.s 2. | Volume N-A. Tot. N-A | Tot. N-A.| Tot. N-A. | Tot. mm, mm, cc, mm. mm, cc. mm, mm, cc. mm, mm, cc. 20.0 45.0 | 1.10} 25 | 72.0) 2.38 | 24.5} 68.5) 1.92 | 16.0} 20.0) 0.42 19.0 46.5 | 0.81 25 | 65.5) 2.40 | 23.0] 60.0) 1.80 | 16.0} 18.0] 0.40 19.0 41.5 | 0.76 | 25 | 65.0) 2.45 | 20.0) 54.0) 2.19 | 15.0} 19.0) 0.35 18.5 46.0 | 0.95 | 25 | 65.0) 2.41 | 22.5) 58.0) 1.95 | 15.0] 17.0) 0.39 18.5 46.5 | 0.86 | 25 | 63.0) 2.37 | 23.0) 60.0) 1.80 | 14.5) 15.0} 0.31 18.5 46.5 | 0:79 | 24 | 64.0) 1.98 | 22.0} 57.0] 1.78 | 14.0] 15.0} 0.28 18.0 46.0 | 0.99 | 22 | 63.0] 1.88 | 21.0] 56.0) 1.58 | 14.0} 15.0) 0.23 18.0 42.0 | 0.78 | 20 | 54.0) 1.19 | 17.0) 48.5] 0.98 | 13.5} 16.0} 0.30 18.0 41.0 | 0.68 | 18 | 46.0) 0.85 | 17.0] 42.5) 0.75 | 13.0) 19.0} 0.26 17.0 42.5 | 0.74} 16 | 39.5) 0.60 | 15.0) 36.5} 0.50 | 13.0) 13.5] 0.22 17.0 39.0 | 0.55 | 14 | 38.0] 0.51 | 18.7| 34.5) 0.41 | 12.5] 14.0] 0.22 16.5 40.0 | 0.73 | 13 | 32.5) 0.38 | 12.0) 30.0} 0.32 16.0 39.5 | 0.52 Averages 18 | 43.2 | 0.79 | 21 | 55.6) 1.57 | 17.6] 50.5] 1.33 | 14.2) 16.5] 3.07 N-A., Nose-anus length; Tot., total; Vol., volume. have undergone a few changes (figs. 18 and 19). The tail has become relatively long, the eyes are wider apart, the dorsum of the head is slightly flattened, and the legs have increased very slightly in length. During the winter and the second spring and summer the thyroidless larvae (figs. 10 and 11) make a still nearer approach in form to that of the frog. The legs 14 E. R. HOSKINS AND M. M. HOSKINS increase very slightly in length, the head and back become very flat on account of skeletal changes, and the eyes are relatively more lateral in position than in normal larvae. Other changes noticeable are a relative increase in the length of the tail and a blunting of the anterior end of the head. Allen (’18) states that the head is relatively long, but this is not the case in our specimens. In his comparison Allen (’18, fig. 3) failed to super- impose the centers of the heads of the larvae. At the end of the period of metamorphosis the thyroidless larvae are from three to five times as large in volume as the frogs of the same age and some become ultimately more than six times as large as the average frog after metamorphosis and nearly three times as large as the average control larva before metamorphosis (table 2). Table 2 shows the variation in the volume (and hence weight) of amphibians as compared on a basis of length both in the living and fixed condition. These animals were fixed in Bouin’s fluid and preserved in 70 per cent alcohol. The data here were se- lected from a very large number of such measurements and published in order to demonstrate a source of serious error in attempts to compare relative volumes of organs on a basis either of length or volume of the animal, as well as to show the relative sizes of larvae and frogs. The volume of fixed specimens varies greatly on account of variations in the amount of shrinkage during fixation, and similar variations in the size of fixed organs must also occur. The volume (or weight) of a living larva of a given length is only fairly constant and varies not only with the amount of foreign substances within the intestine, but also on account of variations in shape produced by changes in the tail-body re- lationship. The intestinal contents cannot be remoyed without destroying the gut, and this would prevent the measurement of the volume (or weight) of the entire animal. In addition to loss of water, a considerable portion of the natural decrease in size suffered during transformation from the tadpole to the frog is due to discharge of intestinal contents, and the exact amount of this is difficult to determine. This loss of fecal matter in- GROWTH OF AMPHIBIA AFTER THYROIDECTOMY 15 creases the difference between the relative volume of larvae and frogs. The shortening of the tail and anal canal during meta- morphosis prevents comparison of the larvae and frogs on a basis of the body length (figs. 12 and 16). Given a frog and larva of the same body length (i.e., nose—anus) the latter is two to three times the volume or weight of the former (table 2). If the body length be considered as the distance between the tip of the head and the posterior limit of the body cavity, a larva is more than three times the size of a frog of the same body length. In table 2 the measurements of living larvae are given only in the case of the thyroidless animals, as the variation in the controls is similar. The increase in the volume of the control larvae and frogs on account of growth of the legs is offset by a slight increase in the relative length of the tail in the thyroidless animals. | Regeneration. On April 13, 1918, 12 mm. of the tail and the entire hind legs were removed from a large thyroidless larva from the 1917 series. After fourteen days the tail had increased in length 11 mm. and the body 2 mm. One hind leg had not grown at all, but the other had grown 2 mm. and regenerated a foot that was about half the size of the one removed. The part of the tail that regenerated contained less muscle than normal but was otherwise like the part removed. On April 27 an additional 13 mm. of the tail and the regenerated leg and foot were removed. The tail increased again 14 mm. in length in twenty-eight days. The leg did not again regenerate. The body again increased 2 mm. in length. On June 24 an additional 12 mm. of the tail was removed and it regenerated 7.5 mm. in twenty-three days. During this period a twist developed in the tail near the body. On July 17 an additional 7.5 mm. of the tail was removed. In five days, 2 mm. of the tail had regen- erated when the animal died. During the three months of this experiment none of the other thyroidless larvae had increased more than 6 mm. in total length, 2 mm. of which growth was in body length, while the experimental animal had regenerated 34.5 mm. of tail and its body grew 4mm. Had it not died, the THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 29, No. 1 16 E. R. HOSKINS AND M. M. HOSKINS larva might have grown still more. Only one hind leg showed any sign of regeneration, and it grew to only half the size of the one removed. When this regenerated leg was removed the animal was unable to regenerate it again. The steadily increasing amount of time required to regenerate the removed caudal tissues and the failure in regeneration of removed legs indicates that, although thyroidless larvae are able to grow and regenerate tissues more than a year after the ex- tirpation of the thyroid, the larva will not continue to regener- ate removed tissues indefinitely or at the same rate as the first regeneration it produces. Allen (718) has noted that young thyroidless larvae are able to regenerate removed parts of the body. Zeleny (’09) found that in normal Salamander larvae repeated amputation of the tail causes an increase in the rate of regeneration. This is influenced, however, by additional in- juries to the body, such as removal of one or both hind legs. If the additional injuries are severe, a decrease in the rate of regeneration is to be observed. Forced metamorphosis. An attempt was made to force a thy- roidless larvae into metamorphosis. The total length of the animal was 59 mm.; the nose-anus length, 24 mm.; the length from the nose to the end of the body cavity, 20 mm.; that of the hind legs, 4.5 mm., and the volume was 1.90 cc. This larva was placed in a moist chamber on a bed of wet spirogyra, where it lived for two days. It is probable that with care animals might be kept alive in such conditions much longer. This speci- men breathed at irregular intervals. In the forty-eight hours during which it lived in the moist chamber, its tail shrank 24 per cent, its anal canal about 50 per cent in volume (due mostly to discharge of contents), and its volume decreased 18 per cent. The gut shortened little, if any, but was strongly contracted. The lungs were filled with air. The shrinkage was not due to inanition, for if a larva similar to the one described is kept in water without food it will maintain its volume for a much longer period than that given to the experiment. Moreover, the shrink- age which finally results from inanition is uniform, while in this experiment the tail shrank first of all, and very rapidly, as it does GROWTH OF AMPHIBIA AFTER THYROIDECTOMY 17 in normal metamorphosis. A part of the shrinkage noted in this experiment was due to loss of water from the tissues, as happens to some extent during normal metamorphosis. Discussion. From the results given in the numerous papers dealing with thyroidectomy in higher animals it was to be ex- pected that definite and striking results would be obtained by thyroidectomy in amphibia. Rogowitsch (’89) found that this operation in mammals causes hypertrophy of the hypophysis and Stieda (’90) noted an increase in the number of ‘chief’ cells of this gland. Herring (’08) stated that thyroid extirpation in mammals causes increased activity of the pars intermedia of the hypophysis, bringing about colloid formation. He reported also a proliferation of the cells in the posterior lobe. Hofmeister (94) stated that this operation in mammals produces cachexia, , resulting in abnormal growth, especially of the bones. Ceni (05) stated that thyroidectomy in fowls causes interference with egg production. Gudernatsch (’12, ’14, ’17) and several others have shown that feeding thyroid to larval frogs hastens the time of metamorphosis and checks general growth, hence results somewhat opposite to those were to be expected in our experi- ments. The effects produced by feeding thyroid to young am- phibia are probably not due to any specific metamorphosing function of the thyroid, but rather to a perversion of metabolism, since they have been produced with thyroid substance of mam- mals which do not undergo metamorphosis as do the amphibia used in the experiments. Feeding thyroid to mammals increases metabolism and causes hypertrophy of various organs, including some of the ductless glands (Hoskins, 716). It was also to be expected that the operation we performed might affect the hypoph- ysis, since Smith (’16, 717) noted that after hypophysectomy in young frogs embryos the thyroid failed to develop normally, and Allen (’17) obtained somewhat similar results. The earliest apparent and most extensive differences between the thyroidless and control frog larvae are in the skeletal ele- ments. As Terry (’18) has well described, calcification and ossi- fication (especially the latter) progress less rapidly in the thy- roidless larvae than in the controls, and finally these processes 18 E. R. HOSKINS AND M. M. HOSKINS practically cease before the frog skeleton is laid down and before the other parts of the larval body have ceased growing. It is probable, as in mammals, that thyroid deficiency interferes with calcium metabolism and thus causes the skeletal abnormalities noted here. The failure of the legs to grow beyond mere buds may be due to the failure of the skeletal part to push out into these buds. The beginnings of metamorphosis are to be observed in the skeleton, which changes to accommodate the body of the animal to its future terrestrial existence. Metamorphosis consists of a series of changes which occur in sequence, and when the first — part of the process is prevented (i.e., skeletal change) through faulty metabolism (probably calcium) other steps in the sequence are prevented. It is to be noted that the lungs develop and function before metamorphosis begins, and have nothing to do with this phenomenon. As soon as normal larvae are well started in metamorphosis they may be removed from the water, and the process will continue to its completion. It is of course possible that thyroid secretion is directly neces- sary for the atrophy of tissues which occurs in metamorphosis, but the experiment in forced metamorphosis described above showed that a certain amount of atrophy will take place in a thyroidless larva kept in a moist chamber. The hypertrophy of the hypophysis which occurs after thy- roidectomy may have something to do with the failure of meta- morphosis, but it is more likely that the condition of this gland is responsible for the rapid growth rate of thyroidless larvae. There is no apparent reason for the fact that the size at which frog larvae go into metamorphosis is not constant. Animals from the same egg mass kept in the same aquarium and hence under similar food and temperature conditions may vary con- siderably in the size attained before metamorphosis and in the time required for it. As noted above, the 1917 (slowly growing) larvae averaged larger than the 1918 (rapidly growing) larvae, but the food conditions are not the only determining factor, because the largest 1918 larvae were larger than the smaller 1917 larvae. It may be that about the time the larvae reach their GROWTH OF AMPHIBIA AFTER THYROIDECTOMY 19 maximum length their entire metabolic process is very unstable, and at this time some slight influence, say of the thyroid and perhaps also of the hypophysis, is able to furnish the stimulus necessary to upset this process and so to initiate the series of vital phenomena which we call metamorphosis. From the point of view of age, it may be considered that the growth of the body of the thyroidless larvae as a whole, or of most of their organs is precocious, since in a given time they grow more rapidly than the control larvae. The skeleton as a whole and the brain grow more slowly than those of the controls. If thyroidless larvae are compared with control larvae of the same size regardless of age, then the only very precocious growth noticeable is that of the hypophysis. Organs and method of comparison The following figures representing the condition of the organs of the animals studied are all camera-lucida drawings of the organs removed in autopsies or studied in microscopical sections. ° The number of complete autopsies performed was sixty-nine. In addition there were done about forty partial autopsies. In addition to these, many larvae and frogs were sectioned serially. For comparison of the organs, outline drawings were made and the drawings afterward were compared directly. Measurements were made of the various diameters of those organs of regular contour, but most of the organs are so irregular in outline that such measurements are of little or no value, and the accompany- ing pictures give a much better idea of the conditions obtaining in most of the organs than would any tables, no matter how carefully prepared. It should be noted that the accompanying figures are drawn from specimens which were carefully chosen as representing the condition as near the average as possible. The organs are in most cases too small for direct volumetric determinations. Allen (’18) and Rogers (’18) have used the product of the three principal diameters of an organ to represent its volume, but this is not even approximately correct, because of the great variation in the shape of most organs of larvae and 20 E. R. HOSKINS AND M. M. HOSKINS frogs. If two glands of exactly the same size but of different shapes are measured by this method, the volumes determined might differ greatly, while two others of different shape and size might seem to have the same volume. ‘This irregularity in the form of the hypophysis, for instance, is shown in our own figures and in those of Rogers on page 605. Rogers states that the same relative error would apply to all the glands so measured, but he used so few specimens that no two of the glands were of quite the same shape, and hence different errors have been made in his measurements of different specimens. In far greater error are Rogers and Allen when they attempt to establish the relative volume of an organ by dividing the product of its three principal diameters by the length of the animal’s body, for the volume of a body varies as the cube of a linear dimension. Moreover, in larvae of different sizes the proportions of the body vary considerably and the difference in the size (volume or weight) of two animals is not directly proportional to the differ- ence in their body length. This is especially true where one ‘attempts to compare tadpoles and frogs. It may be noted in figures 12 and 16 that the length of the real body of the larva (excluding the anal canal) is nearly the same as that of the re- sulting frog and yet the larvae are from two to three times as large as the frogs produced from them (table 2). It may be noted also that the larger thyroidless larvae have relatively long tails, and hence their size is not directly comparable with that of control larvae on the basis of length of body. Still further, larvae contain relatively much more water in their tissues than do frogs. Another source of error in work of this sort is that the animals vary in the amount they shrink during fixation. Given two tadpoles fixed in the same fixative, one may shrink nearly twice as much as the other, and hence with a small number of speci- mens large errors may be introduced in attempting to estimate volumes of structures. Where different fixatives are used, still greater differences in shrinkage result. These animals are so small that unless they are fixed, many of the organs cannot be properly dissected or easily handled, and will undergo post-- GROWTH OF AMPHIBIA AFTER THYROIDECTOMY 21 mortem changes that entirely alter their volume and shape. There is no practical method by which the exact volumes of these small organs can be determined. One could section them and then calculate the volume of each section by use of a plani- meter, and in this way, except for large errors due to shrinkage, estimate approximately the volume of the organs. However, in the case of most of the organs the variability is so large and the actual differences between the control and thyroidless groups are so slight that the result to be obtained would not justify the effort required. We are certain that only large differences in size of glands of these larvae can be detected by any known practical method, and that any small differences noted are as likely to be due to errors of measurement as to the experiment itself. Brain. Figures 22 to 27 show the general character of the brain in the various animals. The brain in the thyroidless animals appears to be relatively small and undifferentiated when compared with the brain of controls of the same age. Com- parison of figures 24 and 25 shows the difference that exists between most of the control and thyroidless animals. In the former the brain is practically the same shape and size as that of the frogs immediately after metamorphosis (fig. 23), whereas the brain of the thyroidless larva, especially in the telencephalon, thalamus, optic lobes, and cerebellum, is shaped very much like that of a very small control larva (fig. 22). This undeveloped condition of the brain of the thyroidless larvae may be noted in older and larger specimens. Gradually, however, the brain of these specimens tends to become differentiated (fig. 26) and finally it assumes a condition practically the same as that of the young frog (fig. 27) except for the shape of the anterior part of hemispheres. In some of the large thyroidless larvae (fig. 26) the optic lobes are relatively short, and of a consequence the cerebellum bulges dorsally and hence appears relatively larger than in the other figures shown here, but the difference in size is apparent rather than real. In the still older large thyroidless larvae (fig. 27) the optic lobes are relatively longer than in the former (fig. 26) and overhang the cerebellum, partially 23 E. R. HOSKINS AND M. M. HOSKINS hiding it and pressing upon it so that it becomes more flattened and appears to become smaller. Allen (’18) noted this early lack of differentiation of the brain of the thyroidless larvae, but did not observe the changes which occur later. It is to be noted that this differentiation of the brain is independent of the changes that occur normally during metamorphosis. ‘The reason for this slow growth of the brain of the thyroidless larva is not evident. In control and thyroid- less larvae of the same size the brain in the latter is usually, though not always distinctly smaller, but it is always less highly differentiated in the thyroidless than in the control larva during and for some time after metamorphosis of the control. The brain during metamorphosis, loses very little, if any, in size and hence changes greatly in relative size; so that, relatively, a nor- mal larva has a much smaller brain than a frog, and a thyroidless larva a still smaller one. Eyeballs. The eyeballs of the metamorphosing controls are of about the same size as those of the corresponding thyroidless larvae. During metamorphosis the eyeballs remain about the same size as before or perhaps gain slightly, so that a frog has relatively larger eyes than a control or thyroidless larva. In the latter, however, the eyeballs continue to grow and this differ- ence is somewhat decreased. During metamorphosis the head narrows and the eyeballs approach each other medially, but as the brain loses but little if any in width the optic nerves must be obliged to shorten (compare figs. 13 and 17). Heart. The heart (figs. 28 to 31) has about the same propor- tions in the control and thyroidless larvae, but varies in size as seen in autopsies on account of varying amounts of enclosed blood. During metamorphosis in the normal animal the heart loses a little in size, but less than the body as a whole. In the older thyroidless larvae, the heart increases in size while the body is growing. Liver. At the time of metamorphosis the liver is very irregular in outline (figs. 32 and 33). 160. 75 Thyroid of precocious frog (fig. 20). This thyroid regenerated and hy- pertrophied. Compare with figure 74 from a normal larva of the same age. x 160. 76 Spleen of small normal larva (23 mm.) showing the two types of splenic cells. XX 670. 77 Spleen of control larva at maximum length before metamorphosis. X 670. 78 Spleen of thyroidless larva of same age as that of figure 77. X 670. 62 GROWTH OF AMPHIBIA AFTER THY ROIDECTOMY PLATE 6 E. R. HOSKINS AND M. M. HOSKINS es 2 oF 4 29 Ske, = * aie Sante? So ¢ ; v4 PLATE 7 EXPLANATION OF FIGURES R. sylvatica. Representative specimens from sixty-nine complete and forty pactial autopsies. All removed organs were drawn with a camera lucida. Abbreviations: Ov., ovary; 7'., testis; S., spleen; K., kidney; Ff. B., fat body. Other abbreviations and dates as in plate 2. Ovaries, kidneys, and spleen. X 8 79 Control larvae. 4/14 to 6/2/18. Alive: Tot., 45 mm., B., 18.5 mm.; fixed: Tot., 42 mm., B., 17.5 mm. 80 Thyroidless larva. 5/2 to 6/8/18. Alive: Tot., 45 mm.; fixed: Tot., 42 mm., B., 17.5 mm. 81 Control frog. 4/14 to 6/5/18. Alive: Tot., 15 mm., B., 14 mm.; fixed: Tot., 14 mm., B., 13 mm. 82 Thyroidless larva. Alive: Tot., 53 mm., B., 22 mm.; fixed: Tot., 53 mm. B., 21 mm. 83 Thyroidless larva. 5/3 to 10/17/17. Alive: Tot., 65 mm., B., 22 mm.; fixed: Tot., 60 mm., B., 21 mm. 4 Thyroidless larva. 4/28 to 12/21/17. Alive: Tot., 54 mm., B., 22 mm.; fixed: Tot., 50 mm., B., 21 mm. 85 Thyroidless larva. 4/28/17 to 7/10/18. Alive: Tot., 68 mm., B., 22 mm.; fixed: Tot., 56 mm., B., 22 mm. 86 Control frog. 4/7 to 8/12/17. Alive: Tot., 14.5 mm., B., 14 mm.; fixed: Tot., 14 mm., B., 13.5 mm. 87 Thyroidless larva. 4/14 to 7/13/18. Alive: Tot., 65 mm., B., 24 mm.; fixed: Tot., 60 mm., B., 23 mm. ’ 0.2) Testes. X 8 88 Thyroidless larva. 5/2 to 6/8/18. Alive: Tot., 45 mm., B., 18 mm.; fixed: Tot., 42 mm., B., 17.5 mm. 89 Control larva. 4/18 to 5/28/18. Alive: Tot., 45 mm., B., 18.5 mm.; fixed: Tot., 42.5 mm., B., 18 mm. 90 Control frog. 4/14 to 6/5/18. Alive, 15 mm., fixed, 14 mm. 91 Thyroidless larva. 4/14 to 7/15/18.. Alive: Tot., 54 mm., B., 20 mm.; fixed: Tot., 48.5 mm., B., 17.5 mm. 92 Thyroidless larva. 5/3 to 10/17/17. Alive: Tot., 68 mm., B., 23 mm.; fixed: Tot., 63 mm., B., 21 mm. ; 93 Thyroidless larva. 4/28/17 to 7/6/18. Alive: Tot., 72 mm., B., 25 mm.; fixed: Tot., 68.5 mm., B., 24.5 mm. GROWTH OF AMPHIBIA AFTER THYROIDECTOMY E. R. HOSKINS AND M. M. HOSKINS PLATE 7 (o>) cr PLATE 8 EXPLANATION OF FIGURES R. sylvatica. Ovaries, X 36 94 Half-grown normal larva (fig. 21). Oocytes, 50 u. 95 Precocious frog (fig. 20). Oocytes, 75 pu. 96 Young normal frog (1917). Oocytes, 120 pz. 97 Young thyroidless larva (1918). Large as control larva at metamor- phosis. Oocytes, 50 w (smaller oocytes than those of 1917 larvae of same size). 98 Thyroidless larva (1918) (5 weeks old). Much younger than that of figure 97, but of much larger size, oocytes (60 ») are relatively of the same size. 99 Thyroidless larva one-half the size of larva of figure 98, but more than twice as old (eleven weeks) Oocytes (130 u) are relatively about four times as large in diameter as those in figure 98. 100 Thyroidless larva (1918). Very little older than larva of figure 99 (twelve weeks), but nearly four times as large. Oocytes (180 ») are relatively smaller than those of figure 99. 101 Thyroidless larva (6 months old), killed in October, 1917. Oocytes, 200 py. 102. Thyroidless larva killed December 21, 1917 (eight months old). Oocytes, 250 pn. 103. Thyroidless larva fifteen months after thyroidectomy. Oocytes, 300 u. Measurements represent the diameter of the largest oocytes. 66 GROWTH OF AMPHIBIA AFTER THYROIDECTOMY PLATE 8 E. R. HOSKINS AND M. M. HOSKINS 67 PLATE 9 EXPLANATION OF FIGURES R. sylvatica. Testes 104 A, T. S. Control larva at beginning of metamorphosis (July 29, 1917). < 34. B, same, showing beginning of tubule formation, but no synapsis. > 240. 105 A,T.S. Young controlfrog. X34. B,same. X 240. 106 A, T. S. Thyroidless larva larger but of same age as larva of figure 104. > 34. B, same, showing advance in tubule formation, but no synapsis. x 240. 107. A, T.S. Large thyroidless larva, killed four months after thyroidectomy (August 20,1917). xX 34. B, same, spermatogenesis begun. X 240. 108 A,T.S. Large thyroidless larva, six months after thyroidectomy. X 34. B, same, showing spermatozoa in tubules. 240. 109 Large thyroidless larva one year older than that shown in figure 106. Shows fully matured testis. V. #., efferent tubule leading into kidney and con- taining spermatozoa. X 34. 68 GROWTH OF AMPHIBIA AFTER THYROIDECTOMY PLATE 9 E. R. HOSKINS AND M. M. HOSKINS 69 Resumen por la autora, Helen Dean King. Estudios sobre ‘“‘inbreeding.”’ IV. Nuevos estudios sobre los efectos de ‘‘inbreeding”’ sobre el crecimiento y variabilidad de peso de la rata albina. Los datos publicados en el presente trabajo demuestran el crecimiento y la variabilidad de peso de mas de 600 ratas albinas pertenecientes a las generaciones comprendidas entre la 16 y 25 generacion de un tronco ‘inbred’? hermano con hermana, per- tenecientes ambos a la misma cria. Los principales puntos de interés son los siguientes: 1. El ‘‘inbreeding’’ continuo no ha producido efecto perjudicial alguno en el tronco albino original en lo referente a la marcha y extensién del aumento de peso del cuerpo, ni tampoco ha alterado la forma de la grafica de creci- miento de los dos sexos. 2. Las relaciones normales de peso de los sexos no se han alterado después de 25 generaciones de ‘“inbreeding.”’ 3. La variabilidad de los pesos de estos animales es relativamente alta en todas las edades y no decrece cuando el “inbreeding” avanza. 4. Una comparacién de la variabilidad de los pesos de diferentes series de albinos del mismo tronco con los de las ratas ‘‘inbred”’ indica que el aumento de variabilidad en las tiltimas se debe a la accién del medio ambiente y a la accion de la nutricién, no al “inbreeding.”’ Translation by José F. Nonidez Carnegie Institution of Washington AUTHOR’S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, JUNE 30 STUDIES ON INBREEDING Iv. A FURTHER STUDY OF THE EFFECTS OF INBREEDING ON THE GROWTH AND VARIABILITY IN THE BODY WEIGHT OF THE ALBINO RAT HELEN DEAN KING The Wistar Institute of Anatomy and Biology EIGHT CHARTS— In order to complete the series of records for the first twenty- five generations of inbred albino rats, data showing the growth and variability in the body weights of individuals belonging in the sixteenth to the twenty-fifth generations are given in the present paper. Five litters from each generation of the two inbred series (A and B), comprising a total of 296 males and 310 females, were used for this study. The rats in these litters were selected in the same manner, and they were weighed at the same age periods, as were the individuals of the seventh to the fifteenth generations for which body-weight records were taken (King, 718). The data for the animals in the different generations of the inbred strain are therefore strictly comparable. During the past three years, when most of the weighings were taken, it was not possible to rear the animals under environ- mental and nutritive conditions that were as favorable to growth and to fertility as those existing previously. Owing to economic conditions incident to the war, it became necessary to make a radical change in the character of the food that the rats received. The ‘scrap’ food (carefully sorted table refuse), on which the animals of the earlier generations seemed to thrive exceedingly well, had to be replaced by a ration that consisted, for the most part, of oats and corn, with the occasional addition of various kinds of vegetables and a little meat. Some of the available 71 THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 29, No. 1 ia HELEN DEAN KING substitutes that from time to time were added to the diet in order to vary it, such as alfalfa, linseed and cottonseed meal, proved very injurious to the rats and very materially affected their growth and fertility. For some time, therefore, the food given the animals has been largely in the nature of an experiment, and it has not even yet been. possible to work out a ration that produces as rapid and vigorous growth and that is as favorable to reproduction as was the ‘scrap’ food given previously. Extremes of temperature, either of heat or of cold, have a very marked effect on the body growth of the rat, as they have on that of mice (Sumner, 709), and many of the animals in the later generations of the inbred strain suffered considerably from this cause. During the excessive cold of the winter of 1917-1918 it was impossible to keep the colony house above the freezing point for days at a time, and in consequence the rats ceased growing at a normal rate and many of them developed pneu- monia. The periods of intense heat experienced during the summer of 1918 also had a very deleterious effect on the vitality and on the body growth of the rats. Asa result of the combined action of these various factors, all inimical to growth as well! as to reproduction, the rats of the eighteenth to the twenty-fifth generations were severely handicapped, and they did not increase in body weight as rapidly, nor did they attain as great a maxi- mum body weight, as did the individuals of the earlier gener- ations. That this decrease in the size of the inbred animals was caused by unfavorable conditions of environment and of nutrition, and not by continued inbreeding, is shown conclusively by the fact that the body weights of hundreds of rats in the outbred-stock colony were just as seriously affected by these adverse conditions as were those of the inbred rats, as will be shown later. Data showing the average body weights at different ages of 179 males and of 130 females belonging in the sixteenth to the twenty-fifth generations of the A series of inbred rats are given in table 1 and in table 2: similar data for 117 males and for 180 females belonging in the same generations of the B series of inbreds are given in table 3 and in table 4. TABLE 1 Showing, by generations, the average body weights at different ages of 179 males belonging in the sixteenth to the twenty-fifth generations of the A series of inbred rats GENERATIONS AGE 13 Hoel Sh “TOs [eae ZO k19"| A174" AG IoAsg 30 44} 44) 49] 41] 42 | 45] 46] 44] 39] 43 60 1S ae Oe els ie a tei ht OG. |. Saar iOd 90 188 | 186 | 192 | 142 | 186 | 164 | 165 | 126 | 121 | 163 120 232 | 228 | 223 | 197 | 237 | 211 | 203 | 169 | 159 | 200 151 255 | 253 | 259 | 232 | 259 | 244 | 226 | 207 | 188 | 233 182 274 | 268 | 286 | 262 | 280 | 271 | 250 | 231 | 216 | 252 212 296 | 277 | 309 | 286 | 289 | 295 | 276 | 243 | 231 | 277 243 302 | 288 | 328 | 298 | 297 | 311 | 291 | 254 | 246 | 279 273 3821 | 310 | 352 | 298 | 308 | 310 | 308 | 264 | 265 | 292 304 317 | 321 | 356 | 301 | 309 | 305 | 305 | 279 | 272 | 303 334 327 | 322 | 365 | 305 | 313 | 311 | 313 | 290 | 276 | 316 365 333 | 319 | 378 | 308 | 324 | 322 | 320 | 288 | 284 | 325 395 336 | 332 | 394 | 304 | 334 | 318 | 306 | 298 | 287 | 327 425 331 | 339 | 376 | 295 | 340 | 319 | 293 | 297 | 290 | 332 455 320 | 3382 | 361 | 295 | 357 | 316 | 296 | 289 | 293 | 322 Number rats weighed RNA 7) SAY GS 20: DR Mot |ot TABLE 2 Showing, by generations, the average body weights at different ages of 130 females belonging in the sixteenth to the twenty-fifth generations of the A series of inbred rats GENERATIONS AGE 13 TO Ait ea” Che: kee) bee ey Gee ee 30 41} 44; 48/| 39} 40! 44] 43] 42] 38] 41 60 99 |} 105 | 110} 94; 97 | 95) 99} 81] 79} 94 90 141 | 163 | 148 | 184 | 153 | 135 | 142 | 107 | 109 | 131 120 170 | 178 | 179 | 163 | 177 | 166 | 170 | 135 | 137 | 153 151 187 | 197 | 195 | 182 | 195 | 192 | 179 | 163 | 160 | 177 182 206 | 211 | 208 | 197 | 205 | 203 | 183 | 184 | 178 | 180 212 210 | 214 | 219 | 205 | 215 | 209 | 202 | 190 | 185 | 191 243 214 | 225 | 222 | 212 | 221 | 213 | 207 | 194 | 187 | 194 273 230 | 221 | 224 | 214 | 222 | 218 | 208 | 197 | 190 | 204 304 225 | 233 | 224 | 221 | 216 | 230 | 206 | 206 | 199 | 212 334 229 | 232 | 220 | 225 | 215 | 236 | 215 | 209 | 189 | 215 365 240 | 231 | 221 | 223 | 214 | 242 | 217 | 211 | 194 | 223 395 242 | 227 | 224 | 219 | 210 | 239 | 217 | 209 | 195 | 234 425 235 | 231 | 224 | 214 | 216 | 231 | 216 | 208 | 190 | 240 455 243 | 236 | 223 | 216 | 212 | 229 | 215 | 204 | 186 | 230 Number rats weighed 13 11 12 13 13 14 14 13 13 14 74 HELEN DEAN KING Tables 1 to 4 are inserted mainly for reference, but a compari- son of the data for the males and females in the various genera- tions brings out clearly the relation between the two sexes as regards their relative body weights at different age periods. In some few instances the average body weights of the males and of the females in a given generation were the same when the animals were thirteen or thirty days old, but after this age the males were the heavier at each period for which records were taken. A similar relation between the body weights of the sexes was also noted for the inbred animals of the seventh to the fifteenth generations (King, 718; tables 1 to 4). Investigations in which large series of stock Albinos were weighed at stated periods (Donaldson, ’06; Jackson, 713; King, ’15; Hoskins, 716) have shown likewise that, with few exceptions, the average body weight of the males exceeds that of the females at each weighing period. Since the data for all generations of the inbred strain is in full accord with that for various series of stock Albinos, it is evident that inbreeding through twenty-five generations of brother and sister matings has not changed the normal relative body weights of the sexes at any age period for which records have been taken. For the purpose of analysis and to facilitate a comparison between the growth in body weight of the individuals in the later generations of the inbred series with those in the earlier generations, the body-weight data for the animals belonging in the sixteenth to the twenty-fourth generations of each inbred series were combined in groups of three generations each: the data thus combined are shown in tables 5 to 7. In each of these tables the data for the individuals of the twenty-fifth generation are given separately in order to show the status of the animals at the end of this period of inbreeding. Data indicating the growth in body weight of males and of females belonging in the various generation groups of the A series of inbreds are shown in table 5. As a graphic representation of series of data greatly facili- tates their comparison, the body-weight data for various groups of albino rats, given in tables 5 to 11, have formed the basis TABLE 3 Showing, by generations, the average body weights at different ages of 117 males belonging in the sixteenth to the twenty-fifth generations of the B series of inbred rats GENERATIONS AGE 13 Deke |e 204 O20 AS. 49. [20 | 19! | 18 | 1g 30 52 | 44] 51] 48] 41] 43] 48] 48] 42] 42 60 119 | 111 | 119 | 123 | 109 | 111} 96/114] 95] 114 90 178 | 165 | 181 | 175 | 161 | 168 | 159 | 154 | 143:| 147 120 213 | 222 | 225 | 197 | 203 | 220 | 188 | 180 | 182 | 180 151 237 | 245 | 261 | 227 | 233 | 245 | 229 | 212 | 224 | 221 182 266 | 266 | 284 | 251 | 263 | 267 | 263 | 235 | 228 | 231 212 284 | 279 | 297 | 277 | 272 | 277 | 279 | 252 | 240 | 252 243 289 | 289 | 310 | 304 | 279 | 305 | 284 | 258 | 259 | 279 273 296 | 297 | 319 | 323 | 285 | 312 | 293 | 265 | 272 | 289 304 304 | 303 | 323 | 324 | 289 | 317 | 300 | 276 | 271 | 299 334 310 | 319 | 336 | 337 | 298 | 323 | 301 | 291 | 292 | 312 365 322 | 317 | 322 | 337 | 295 | 328 | 315 | 290 | 305 | 322 395 337 | 326 | 311 | 352 | 296 | 322 | 309 | 293 | 303 | 339 425 304 | 354 | 305 | 364 | 290 | 320 | 305 | 300 | 298 | 352 455 340 | 299 | 299 | 350 | 279 | 322 | 299 | 297 | 287 | 349 Number rats weighed Po hethe TT lo BR Di ie | a fe ey TABLE 4 Showing, by generations, the average body weight at different ages of 180 females belonging in the sixteenth to the twenty-fifth generations of the B series of inbred rats GENERATIONS AGE days 13 19s) 16:| We! 20 a7 | Se 19. heed Ont Aaa 18 30 47| 41} 48 | 44] 39); 41 44; 45] 41 41 60 102 | 938] 107 | 101 | 92] 101} 85 | 97); 838] 98 90 151 | 125 | 157 | 143 | 133 | 149 | 183 | 183 | 119 | 127 120 166 | 165 | 179 | 167 | 162 | 192 | 151 | 149 | 148 | 152 151 189 | 180 | 199 | 183 | 179 | 188 | 177 | 166 | 167 | 166 182 199 | 194 | 212 | 194 | 198 | 199 | 194 | 177 | 177 | 178 212 211 | 205 | 214 | 205 | 201 | 200 | 199 | 187 | 185 | 187 243 215 | 214 | 216 |.212 | 216 | 210 | 199 | 196 | 191 | 191 273 221 | 226 | 218 | 216 | 220 | 210 | 203 | 196 | 194 | 194 304 224 | 217 | 217 | 221 | 218 | 211 | 200 | 198 | 196 | 202 334 225 | 224 | 215 | 225 | 217 | 209 | 207 | 208 | 203 | 204 365 233 | 224 | 220 | 282 | 214 | 212 | 205 | 208 | 211 | 214 395 239 | 216 | 216 |. 2382 | 212 | 210 | 207 | 212 | 201 | 213 425 243 | 238 | 218 | 233 | 207 | 203 | 203 | 210 | 199 | 220 455 — 243 | 253 | 217 | 228 | 201 | 202 | 202 | 211 | 200 | 218 Number rats weighed 16; 14) 18 | .J9-) gear 18 | 18) 19] 19) 20 76 HELEN DEAN KING for the construction of the graphs shown in figures 1 to 8. The graphs in figure 1 show the growth in body weight of four gener- ation groups of male rats belonging in the A series of inbreds (data in table 5). In this, as in some of the other figures, the graphs should properly run very close together or overlap in various places. If, however, the graphs had been drawn in this manner, it would be difficult to follow their course, and therefore TABLE 5 Showing the average body weights at different ages of inbred rats of the A series, separated into groups according to the generation to which the individuals belonged MALES FEMALES eh: Genera- | Genera- | Genera- | Genera- | Genera- | Genera- | Genera- | Genera- tions 16-18|tions 19-21|tions 22-24] tion 25 |tions 16-18 tions 19-21)tions 22-24] tion 25 days 13 19 18 17 18 18 Le 16 ibys 30 46 43 42 43 44 41 41 41 60 129 110 96 103 105 95 86 94 90 189 163 137 163 150 139 120 131 120 228 214 LUC. 200 175 169 144 153 151 255 244 206 233 192 190 168 172 182 275 271 231 252 208 202 185 180 212 291 290 249 277 214 209 193 191 243 301 302 264 279 220 215 196 194 273 322 306 280 292 225 218 198 204 304 326 305 286 303 227 221 204 212 334 333 310 294 316 227 222 204 215 365 337 317 298 325 232 224 208 223 395 340 316 299 327 251 220 208 234 425 345 313 294 331 229 218 205 240 455 338 316 293 322 231 218 202 230 the space between them has been arbitrarily widened in some places in order to keep the lines distinct. While the general course of all of the graphs in figure 1 is much the same, their relative position clearly shows the progressive decrease in body weight that has resulted from the action of unfavorable conditions of environment and of nutrition. The rats in the sixteenth to the eighteenth generations were fed, for the most part, on ‘scrap’ food, and, as graph A in figure 1 shows, EFFECTS OF INBREEDING ON BODY WEIGHT 77 the males of the A series that belonged to these generations were heavier at all ages than were the males in the later generation groups, excepting at the 243-day period. Rats in the nineteenth to the twenty-first generations were not greatly affected by the change in diet, as for some months it was possible to give them ‘scrap’ food part of the time. The males of this generation eEESen - ; Tt 2585! : Ht 2ESSEGSERERSEREESS __ =o Titi = PTrtiilir tii wg. ~ sseReeeaee SSS SS eee Hf = = Z coo @GG8 5588 ‘ annm ——7-— saa > a Perr rrr SESS SSeSSeee! H nanueeae SESSERESESEUEES: Fic.1 Graphs showing the increase in the weight of the body with age for males belonging to various generation groups of the A series of inbred rats. A, graph for males of the sixteenth to the eighteenth generations, inclusive; B, graph for males of the nineteenth to the twenty-first generations, inclusive; C, graph for males of the twenty-second to the twenty-fourth generations, inclusive; D, graph for males of the twenty-fifth generation (data in table 5). group, as the position of graph B indicates, were nearly as Jarge as were those of the earlier generation group during the adoles- cent period, but in the adult state their body weights fell off rapidly. Individals in the twenty-second to the twenty-fifth generations of the inbred strain suffered most severely from the altered food conditions as well as from extremes of temperature, and the males of the A series were very inferior in body weight 78 HELEN DEAN KING to those of the preceding generations, as graph C and graph D in figure 1 show. Since the number of weighed individuals in a single generation was comparatively small, it is not surprising that the course of graph D should be rather erratic. At its beginning this graph runs very slightly higher than graph C, but at the 90-day period it begins to rise rapidly, and at 334 days it crosses graph B and subsequently runs above it until the final weighing. In the A series of inbreds the males of the rH coo) «Growth in body weight Albino Rat HH Body weight in grams H oo 340 360 380 400 Fig. 2 Graphs showing the increase in the weight of the body with age for. females belonging to various generation groups of the A series of inbred rats (data in table 5; lettering as in fig. 1). twenty-fifth generation were, as a group, superior in body weight to the males of the generation preceding. The superiority of these individuals can be attributed in part to an improvement in the nutritive conditions and in part to the fact that the ma- jority of animals in this generation were born at the time of year that experience has shown is most favorable for body growth in the rat, i.e., the winter months. Graphs showing the growth in body weight of females belong- ing to various generation groups of the A series of inbreds are EFFECTS OF INBREEDING ON BODY WEIGHT 79 shown in figure 2. The data from which these graphs were constructed are given in table 5. In general the relative position of the graphs in figure 2 is much the same as that of the graphs in figure 1. Graph A, represent- ing the body weight increase with age for females of the sixteenth to the eighteenth generations, runs higher than any of the other graphs for the greater part of its course, while the position of the other graphs indicates that there was a gradual decrease in TABLE 6 Showing the average body weights at different ages of inbred rats of the B series, separated into growps according to the generation to which the individuals belonged MALES FEMALES AGE a | Genera- | Genera- | Genera- | Genera- | Genera- | Genera- | Genera- | Genera- tions 16-18)tions 19-21|tions 22-24] tion 25 |tions 16-18|tions 19-21|tions 22-24] tion 25 days 13 19 19 19 18 17 18 18 18 30 49 44 46 43 46 42 43 41 60 116 115 102 114 102 98 89 93 90 175 168 152 147 146 141 128 127 120 220 207 184 180 170 173 148 152 151 249 235 216 211 190 183 170 166 182 271 260 243 231 202 197 183 178 212 286 275 258 252 211 202 190 187 243 295 297 268 279 215 213 195 191 273 303 308 277 289 221 215 198 194 304 309 dll 283 299 220 216 198 202 334 319 319 296 312 222 217 206 203 365 320 319 304 322 227 219 207 210 395 326 319 301 339 230 217 207 213 425 328 323 301 352 234 213 204 220 455 321 318 296 349 236 211 204 218 the body growth of the animals as inbreeding advanced. The females of the twenty-fifth generation (graph D) were, on the whole, slightly heavier than were the females of the preceding generation group (graph C). Table 6 gives data showing the average body weights at differ- ent age periods of males and of females belonging to various generation groups of the B series of inbreds. 80 HELEN DEAN KING The data given in table 6 served as the basis of construction for the graphs shown in figure 3 and in figure 4. A comparison of the graphs in figures 3 and 4 with the corre- sponding graphs in figures 1 and 2 shows that there was very little difference between the two inbred series (A and B) as regards the pody-weight increase with age in the animals of the various generation groups. In the B series, as in the A series, males and Growth in body weight. Albino Rat SeriesB Males 5 Fig. 3 Graphs showing the increase in the weight of the body with age for males belonging to various generation groups of the B series of inbred rats (data in table 6; lettering as in fig. 1). females in the sixteenth to the eighteenth generation groups (graph A) were heavier animals at any given age than were those of subsequent generations; while the rats of the twenty-second to the twenty-fourth generation groups showed a much less vigorous growth than did the animals in the earlier groups. The rats in the twenty-fifth generation of the B series increased in body weight very slowly during the adolescent period, as the position of graph D in figures 3 and 4 indicates; but in the adult EFFECTS OF INBREEDING ON BODY WEIGHT 81 state their growth was much more vigorous, and their body weights, especially those of the males, compare favorably with the weights of the animals in the group comprising the rats of the sixteenth to the eighteenth generations (graph A). An examination of figures 1 to 4 brings out one fact of con- siderable interest: all of the graphs have the same general form, although they vary somewhat in height. As the form of these * graphs is practically the same as that of the growth graphs for 5 EH Coo Pri ss uv 20 40 60 80 100 120 140 160 180 200 220 240 260 260 300 320 340 360 380 400 420 440 460 460 Fig. 4 Graphs showing the increase in the weight of the body with age for females belonging to various generation groups of the B series of inbred rats (data in table 6; lettering as in fig. 1). stock Albinos as determined by Donaldson (’06) and others, it follows that close inbreeding, continued through many genera- tions, does not alter the character of the growth graph for the albino rat. Theoretically, it might be expected, perhaps, that long-continued inbreeding would cause a slowing up of the growth processes, since the animals totally lack the stimulus to growth that a condition of heterozygosis seems to give in many cases (East and Hayes, 12; Jones, 718). The body weights of the animals in the sixteenth to the twenty-fifth generations of the inbred strain tended to lag somewhat during early postnatal 82 HELEN DEAN KING life (figures 7 and 8, graph B), but this was undoubtedly due to the action of environmental and nutritive conditions, not to inbreeding. Any agency influencing growth, whether it be bene- ficial or detrimental, naturally produces its greatest effect during the period when growth is normally most rapid and vigorous. Since unfavorable conditions of environment and of nutrition unquestionably limited the extent of body growth in the animals of the later generations of the inbred strain, it is very probable that these factors also lessened the rate of growth during the early life of the individuals. If body growth in the inbred rats of future generations is retarded during the adolescent period, although the environmental and nutritive conditions under which the animals live are such that they produce rapid and vigorous growth in outbred stock Albinos, the change in the rate of growth can be ascribed to the effects of inbreeding. As far as the ex- periment has gone at present, the evidence does not warrant the conclusion that inbreeding per se has altered the form of the growth graph to any appreciable extent. The body-weight data for the animals in various generation groups of the two inbred series, as given in table 5 and in table 6, were combined in order to show the weight increase with age in the individuals of the inbred strain as a whole. The combined data are shown in table 7. The data in table 7 are not presented graphically, since there was such a close agreement between the corresponding records for the various generation groups of the two series that graphs constructed from the combined data would not differ materially from those given for the separate series (figs. 1 to 4). Table 8 gives data showing the increase in the weight of the body with age for all of the individuals in the sixteenth to the twenty-fifth generations of the A series of inbreds for which growth records were taken; table 9 shows similar data for in- dividuals of the B series. A comparison of the data in table 8 with corresponding data in table 9 shows that the rats in the two inbred series were much alike as regards the rate and extent of their growth in body weight. To show this similarity more clearly, weight data for the males of the two series are presented graphically in figure 5. TABLE 7 Showing the average body weights at different ages of inbred rats of the two series (A, B) separated into groups according to the generation to which the individuals belonged MALES FEMALES AGE = SSS SS | Genera- | Genera- | Genera- | Genera- | Genera- | Genera- | Genera- | Genera- tions 16-18|tions 19-21|tions 22-24) tion 25 |tions 16-18/tions 19-21|tions 22-24) tion 25 13 19 19 18 18 18 18 17 17 30 47 43 44 43 45 41 42 41 60 124 112 98 107 103 OF. 88 94 90 183 165 142 157 148 140 125 129 120 225 211 179 193 173 171 146 152 151 253 240 210 225 191 186 169 168 182 274 266 235 244 205 199 184 178 212 289 283 252 268 212 205 192 188 243 298 299 265 279 217 214 196 192 273 313 307 279 291 223 217 198 198 304 318 308 285 301 222 218 200 206 334 326 314 294 314 224 219 205 208 365 328 318 300 324 229 221 208 215 395 312 318 300 332 231 218 207 222 425 336 317 297 339 232 215 205 229 455 329 317 294 302 234 214 203 223 TABLE 8 Showing the increase in the weight of the body with age for 179 males and for 130 females belonging in the sixteenth to the twenty-fifth generations of the A series of inbred rats « MALES FEMALES AGE Body weight Number Body weight Number of in- 3 was yee 2 Ota Average | Highest | Lowest | dividuals) average | Highest | Lowest | Tividuals days grams grams grams grams grams grams 13 18 24 14 179 if 24 14 130 30 43 54 36 179 42 57 33 130 60 109 205 58 179 95 158 62 130 90 161 268 92 179 135 187 80 121 120 203 294 128 179 162 218 108 125 151 232 321 163 178 182 235 133 123 182 256 361 196 178 196 238 162 122 212 274 382 192 170 203 246 159 116 243 285 404 199 162 208 268 157 110 273 298 432 215 148 212 268 178 100 304 302 413 213 141 215 279 169 95 334 308 410 213 127 216 273 174 91 365 314 418 223 116 220 283 168 86 395 308 421 231 103 220 305 164 80 425 313 485 227 85 218 298 169 68 455 311 447 223 75 214 269 162 58 84 HELEN DEAN KING The relative position of the graphs in figure 5 shows that during the early growth stages males of the B series of inbreds were ‘slightly heavier at any given age than were the males of the A series; in the period from 100 to 300 days the advantage in body weight was with the males of the A series; beyond this age males of the B series were again the heavier. In the adult state the space between the graphs represents a difference of only about TABLE 9 Showing the increase in the weight of the body with age for 117 males and for 180 females belonging in the sixteenth to the twenty- jifth generations of the B series of inbred rats MALES FEMALES AGE BODY WEIGHT Number BODY WEIGHT Number OLin=)) [\ a ee Average | Highest | Lowest | viduals! 4 verage | Highest | Lowest | tividuals days grams grams grams grams grams grams 13 19 24 15 117 18 22 14 180 30 46 62 36 117 44 60 34 180 60 111 147 64 117 96 137 63 180 90 163 230 110 117 136 188 98 161 120 201 281 153 117 162 218 122 169 151 230 326 165 117 179 236 136 164 182 255 358 189 116 186 247 143 176 212 270 367 195 115 199 250 157 163 243 285 392 219 113 205 261 169 163 273 294 415 227 106 208 277 168 148 304 300 410 236 104 209 290 172 148 334 311 459 258 96 213 287 181 136 365 315 460 259 85 216 280 180 126 395 317 449 239 78 216 293 Ure 114 425 319 455 246 64 215 293 171 99 455 315 450 238 56 213 279 168 78 2 per cent in the average body weights of the two groups of animals. Graphs showing the increase in the weight of the body with age for females of the two inbred series are shown in figure 6. These graphs are based on data given in table 8 and in table 9. In figure 6, as in figure 5, the graphs lie very close together throughout their entire course. Females in the B series of in- breds were slightly heavier animals than those in the A series Fig. 5 Graphs showing the increase in the weight of the body with age for males belonging in the sixteenth to the twenty-fifth generations of the two series (A and B) of inbred rats (data in table 8 and in table 9). Fig. 6 Graphs showing the increase in the weight of the body with age for females belonging in the sixteenth to the twenty-fifth generations of the two series (A and B) of inbred rats (data in table 8 and in table 9). 85 86 HELEN DEAN KING during early life, but in the adult state this relation was reversed and the females in the A series were about 2 per cent heavier, as the graphs in figure 6 indicate. In the seventh to the fifteenth generations of the inbred strain, also, the animals of the two series had about the same average body weight at corresponding age periods, although, as a group, the individuals of the B series were slightly heavier (King, 718; tables 11 and 12). Throughout the period of over nine years that this experiment has been in progress, therefore, body growth in the individuals of the one inbred series has closely paralleled that of the individuals in the other series. If the varying con- ditions of environment and of nutrition to which the animals of the inbred strain have been subjected have had any influence on the heritable factors on which growth depends, it is evident that they have acted on the animals of both series in a similar way. Iam strongly inclined to the opinion that environmental and nutritive conditions do not influence genetic growth factors directly, but that they act by either stimulating or retarding the growth processes. Body-weight data for a total of 606 individuals, 296 males and 310 females, belonging in the sixteenth to the twenty-fifth genera- tions of the inbred strain are given in table 10. Reference. to this table, which is a combination of the data in table 8 and in table 9, will be made later. In connection with another problem I have recently taken a series of body-weight records for a second group of outbred stock Albinos. Supposedly these rats represented the best stock in our colony at the time that the investigation was begun (1916), as care was taken to select for breeding the largest and apparently the most vigorous individuals from the large number available for this purpose. These stock Albinos were reared simultane- ously with, and under the same environmental and nutritive conditions, as the inbred rats of the twenty-first to the twenty- fifth generations. The body-weight data for these animals are given in table 11. A comparison of the body-weight data for the stock Albinos (table 11) with that for the inbred group (table 10) shows that EFFECTS OF INBREEDING ON BODY WEIGHT 87 the inbred rats, both males and females, were much heavier than the stock rats at every age for which records were taken. Not only were the animals in this stock series very inferior in size to those in the first stock series reared in 1913 to 1915 as controls for the inbred strain (King, 715; table 3), but their average body weights during adult life were no greater than those of the rats in the first six generations of the inbred strain which suffered severely from malnutrition (King, 718; table 3). TABLE 10 Showing the increase in the weight of the body with age for 296 males and for 310 females belonging in the sixteenth to the twenty-fifth generations of the inbred rats (Series A and B combined) MALES FEMALES 2cH BODY WEIGHT Number BODY WEIGHT Number 2 of in- of in- Average | Highest | Lowest dividuals Average | Highest | Lowest dividuals days grams grams grams grams grams grams 13 18 24 14 296 18 24 14 | 310 30 44 62 36 296 43 60 33 310 60 110 205 58 296 95 158 62 310 90 161 268 92 296 136 188 80 282 120 202 294 128 296 162 218 108 294 151 232 326 163 295 180 236 133 287 182 255 361 189 294 187 247 143 298 212 272 382 192 285 201 250 157 279 243 285 404 199 275 206 268 157 273 273 296 432 215 254 210 277 168 248 304 301 413 213 245 211 290 169 240 334 310 459 213 223 214 287 174 227 365 314 460 223 201 218 283 168 212 395 312 449 231 181 218 305 164 194 425 315 485 227 149 216 298 169 167 455 312 450 223 131 213 293 162 136 To facilitate a comparison between the body growth of inbred rats belonging in various generation groups and that of outbred stock Albinos, graphs showing the weight increase with age in two groups of inbred rats and in two groups of stock rats are given in figure 7 and in figure 8. Growth graphs for various series of male rats are shown in figure 7. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 29, No. 1 88 HELEN DEAN KING In figure 7, graph A runs considerably above all of the other graphs, except at the thirteen-day period, thus showing that the growth of the males in the seventh to the fifteenth generations of the inbred strain was exceptionally vigorous. Males in the sixteenth to the twenty-fifth generations were relatively small: in the adult state their average body weights were about 9 per TABLE 11 Showing the increase in the weight of the body with age and the coefficients of vari- ability for 165 males and for 139 females belonging to a series of stock albino rats that were reared under the same environmental and nutritive conditions as the inbred rats belonging in the twenty-first to the twenty-fifth generations MALES FEMALES ace Average : Number | Average : Number ha gariabilty individuals Bi 8 “Variability indsviduals days grams grams 13 15 15.8+0.92 165 17 16.0+0.84 139 30 40 18.4+1.01 165 39 17.6+1.04 139 60 94 21.3+0.83 150 83 20.2+0.83 131 90 126 20.0+0.97 149 116 17.4+0.75 122 120 173 19.6+0.76 149 137 14.9+0.66 118 151 195 18.1+0.71 149 152 12.0+0.52 120 182 213 15.9+0.63 147 164 11.3+0.51 111 212 226 18.0+0.71 143 171 13.7+0.65 102 243 232 17.6+0.61 137 174 13.1+0.61 105 273 239 18.38+0.76 129 185 13.0+0.74 101 304 243 19.6+0.87 116 186 14.4+0.71 94 334 247 17.9+0.89 108 189 14.3+0.72 7 365 254 15.8+0.77 94 188 14.4+0.74 86 395 258 15.6+0.76 15 195 15.4+0.86 71 425 263 19.1+1.24 54 192 14.8+0.93 57 455 269 18.0+1.35 39 195 15.2+1.02 49 18.0+0.85 14.8+0.76 cent less than those of the males in the earlier generations, as the position of graph B indicates. A comparison of graph B with graph C in figure 7 shows that the body-weight increase with age in the males of the later gener- ations of the inbred strain was, on the whole, very similar to that in the males of the series of stock Albinos reared in 1913 to 1915 as controls for the inbred strain: stock males grew somewhat EFFECTS OF INBREEDING ON BODY WEIGHT 89 more vigorously during the adolescent period, but they were not as heavy as the inbred males in adult life. Since the inbred males were fully as large as the males in the stock series that had been reared under much more favorable conditions of environ- ment and of nutrition, it is evident that continued inbreeding ey 400 Tg fan if SESSSEReeeeeeee 0 0 40 60 80 100 120 140 160 180,200 220 240 260 280 500 320 340 360 Fig. 7 Graphs showing the increase in the weight of the body with age for males belonging to four series. A, graph for males of the seventh to the fifteenth generations of the inbred strain (series A, B); B, graph for males of the sixteenth to the twenty-fifth generations of the inbred strain (series A, B); C, graph for males of the selected series of stock Albinos reared in 1913 to 1915 as controls for the inbred strain; D, graph for males of the stock series reared simultaneously with the individuals of the twenty-first to the twenty-fifth generations of the inbred strain (data in table 10 and in table 11 of the present paper and in table _ 13 of ‘Studies on inbreeding I;’ King, ’18). has not produced a deterioration in the original stock as regards the normal weight increase with age. The males in the seventh to the fifteenth generations of the inbred strain were much su- perior in body weight to outbred stock males reared under simi- lar environmental and nutritive conditions (compare graph A with graph C in figure 7). Likewise, inbred males of the six- Bf 90 HELEN DEAN KING teenth to the twenty-fifth generations, living for the mest part under the handicap of inadequate nutrition, were considerably heavier at all ages than the males in a stock series that were reared simultaneously with them, as a comparison of graph B with graph D in figure 7 shows. The space between these graphs, at the 200-day period, indicates a difference of about 17 per cent in favor of the males of the inbred group. Growth graphs for various groups of female rats are shown in figure 8. seuusesescauscsusesssss Growth in body weight Albino Rat ‘To co HH Hy 1 Body weight in grams [4 . Fig. 8 Graphs showing the increase in the weight of the body with age for females belonging to four series (data and lettering as in table 7). The growth graphs for various groups of females, shown in figure 8, have the same relative positions as have the graphs for the corresponding groups of males (fig. 7), but they lie some- what closer together. Inbred females of the seventh to the fifteenth generations, as graph A shows, were heavier at all ages (except thirteen days) than the females of the other groups; in the adult state their average body weights were about 2 per cent greater than those of the inbred females belonging in subse- quent generations (graph B). Body weight increase with age EFFECTS OF INBREEDING ON BODY WEIGHT 91 in females of the sixteenth to the twenty-fifth generations of the inbred strain closely followed that of the females in the first series of stock controls (compare graph B with graph C in figure 8). The animals in both of these latter groups were about 14 per cent heavier in adult life than the females in the stock series reared during the past two years (graph D). In explanation of the remarkably vigorous growth of the animals in the seventh to the ninth generations of the inbred strain it was suggested in the first paper of this series (King, ’18) that: “favorable nutritive conditions following a period of semi- starvation greatly increased metabolic activity and so stimulated the growth impulse that the animals attained an unusually large size. After the maximum effect of the stimulus had passed there was a gradual decline to more normal conditions of metabolism and a corresponding decrease in the average size of the individ- uals.”” Rats seem to be particularly sensitive to changes in food conditions, more so than is generally supposed, and only by feeding them constantly on a proper diet can their normal weight and fertility be maintained. In light of the valuable researches of McCollum (’18) and his associates, it is evident that the ‘scrap’ food that the rats received during the period when they exhibited their maximum growth and fertility not only furnished a well-balanced ration as regards the basic food stuffs, but that it also gave a sufficient quantity of the essential accessory foods, ‘fat-soluble A’ and ‘water-soluble B,’ to greatly stimulate the growth processes. The experimental diets recently used in our colony have very evidently been deficient in ‘fat- soluble A.’ As a result the rats have shown marked evidence of malnutrition, although they have received an abundance of food. By rectifying the mistakes of the past and feeding the animals on a properly balanced ration, it is hoped that body growth will again respond to the stimulus of adequate nutrition and that it will be possible to obtain inbred animals that are as large as those in the seventh generation. As after twenty-five generations of brother and sister matings the animals in the in- bred strain were fully as large as were the best stock animals obtainable, it is evident that close inbreeding does not inevitably 92 HELEN DEAN KING cause a decrease in body size, as Darwin (’75, ’78), Crampe (’83), Ritzema-Bos (’93, ’94), and others have asserted. Inadequate nutrition, seemingly, is far more detrimental to body growth than is close inbreeding, even when continued over many generations. VARIABILITY IN THE BODY WEIGHTS OF INBRED RATS At the end of fifteen generations of brother and sister matings the rats in the inbred strain were over 96 per cent homozygous, according to the calculations of Fish (714). Animals of the later generations, which had attained a degree of homozygosity prob- ably greater than that ever before reached by any group of labora- tory mammals, might be expected, perhaps, to show a very great uniformity in body weight at different age periods, if the body weight increase with age in the rat is entirely dependent on the action of genetic growth factors. But just as the rate and extent of body growth in this animal seems to be largely a matter of environment and of nutrition, so also the variations in body weights at different age periods are apparently greatly influenced by these conditions. As it is impossible, at present, to distinguish the variability due to environmental and nutritive action from that resulting from a difference in the genetic factors for body growth, one can only calculate the total amount of variability in given groups of animals and then, by comparison, determine the relative variability of the groups. No very defi- nite conclusions can be drawn regarding the effects of close inbreeding on the variability in the body weight of the rat until the animals can be kept under environmental and nutritive conditions that are so uniform that their effect is practically constant and therefore negligible. In order to obtain some idea regarding the relative extent of variability in the body weights of the animals in various genera- tions of the inbred strain, coefficients of variability, with their probable error, were determined for the body weights of the individuals in the sixteenth to the twenty-fifth generations of each of the two inbred series and for the weights of the animals in the two series combined (A, B). These coefficients, with EFFECTS OF INBREEDING ON BODY WEIGHT 93 their probable error, were calculated from the data summarized in tables 8, 9, and 10 according to the formulae given by Daven- port (’14); they are shown in table 12. During early postnatal life, as the coefficients in table 12 show, the females in both inbred series were slightly more vari- able in body weight than were the males, but after thirty days of age the males, as a rule, were the more variable. Variability TABLE 12 Showing the coefficients of variation, with their probable error, for the body weights at different ages of the two series of inbred rats (sixteenth to the twenty- fifth generations, inclusive) SERIES A SERIES B COMBINED SERIES (A, B) AGE Males Females Males Females Males Females days 13 12.6+0.45/138.5+0.56/12.2+0.54/12.2+0.44/12.4+0.36/12.9+0.33 30 12.1+0.43)11.9+0.50)14.3+0.63|14.4+0.52/13.5+0.37/13.3+0.36 60 22.9+0.82/18.0+0.75|16.3+0.71/16.4+0.60/20.6+0.57|17.1+0.46 90 20.1+0.72|15.8+0.69)14.8+0.65)14.0+0.53)18.4+0.51/15.5+0.44 120 19.1+0.68/13.7+0.58}14.3+0.63/11.9+0.44/16.1+0.45)12.7+0.35 151 14.5+0.52}/10.2+0.45/13.7+0.60/10.7+0.39}14.2+0.39)10.5+0.30 182 13.30.48} 9.1+0.39)12.6+0.59)12.3+0.44/13.1+0.36/10.4+0.29 212 13.3+0.49| 9.6+0.43/12.0+0.53} 8.7+0.33/12.8+0.43} 9.2+0.39 243 13.30.50} 9.8+0.45)11.8+0.51| 8.5+0.32)12.4+=0.36| 9.3+0.27 273 12.6+0.49) 9.9+0.47|10.9+0.50) 9.1+0.35)11.8+0.35| 9.4+0.29 304 11.3+0.45/10.8+0.53/10.2+0.48) 8.6+0.34)11.2+0.34| 9.7+0.30 334 12.3+0.52)10.1+0.50/11.2+0.55|) 7.9+0.32/11.7+0.37) 8.9+0.28 365 12.3+0.54/10.3+0.53)12.0+0.62) 9.0+0.38)12.2+0.45| 9.6+0.31 395 12.9+0.61)11.3+0.62)12.4+0.67| 9.8+0.44)12.7+0.45)10.5+0.36 425 14.2+0.73)10.2+0.59}/12.9+0.77/10.5+0.50)13.8+0.54/10.7+0.39 455 14.3+0.79|11.7+0.73/13.6+0.69]11.5+0.62/14.1+0.59|11.6+0.47 Average........ 14.4+0.58)11.6+0.55/12.8+0.60/10.9+0.43)13.8+0.43/11.3+0.35 was at its maximum for both sexes at the sixty-day period, and then tended to decrease with advancing age for some time. In table 12 the average coefficient for the male group in each of the two inbred series, taking all ages together, exceeds that for the corresponding group of females by over two points. Since this difference is over three times the probable error, it is sufficiently large to indicate that the males had a greater range of vari- 94 HELEN DEAN KING ability in body weight than had the females. Coefficients of variability for the body weights of the individuals in the earlier generations of the inbred strain (King, ’18; table 15), and also those for various series of stock Albinos (Jackson, 713; King, ’15), all show that the males are more variable than the females. Such a relation between the sexes as regards the variability in their body weights would seem to be a characteristic of the albino strain of rats in general, and from the results obtained in the present study it is evident that this relation has not been changed by twenty-five generations of close inbreeding. Males in the sixteenth to the twenty-fifth generations of the A series of inbreds had a somewhat greater range of variability in body weight than had the males of the B series, judging from the relative size of the coefficients for the two series as given in table 12. Between the average coefficients for the two series there is a difference of 1.6 points in favor of the males of the A series; a similar relation between the two series existed also at an earlier period (King, 718; table 15). Throughout all genera- tions of the inbred strain, therefore, the range of variability in body weights was greater in the males of the A series than in those of the B series. This difference persisted even during the periods when body growth and variability were greatly influ- enced by environmental and nutritive conditions. A comparison between corresponding coefficients for the fe- males of the two inbred series (table 12) shows that, as a rule, the females of the A series were slightly more variable in body weight at different age periods than were the females of the B series, but, taken as a whole, the one group of females was about as variable as the other, since the difference between the average coefficients for the two groups is only 0.7 point. As the study of variability in the females of the earlier generations of the inbred strain led to the conclusion that “the range of variability in body weights was practically the same for the females of the two inbred series,”’ it is evident that long-continued inbreeding has not altered the relative variability of the females in the two inbred series any more than it has that of the males. EFFECTS OF INBREEDING ON BODY WEIGHT 95 Table 12 shows that in each inbred series the coefficients of variability for both sexes decrease in size with advancing age until the animals attained an age of about 300 days, and then tend to become somewhat larger; a similar change in the size of the coefficients at various age periods was also noted for the animals in the earlier generations of the inbred strain as well as for those in the two stock series reared as controls. After reach- ing the height of their reproductive activity at the age of from seven to ten months, certain individuals, especially males, tend to accumulate an excess of adipose tissue; while other individuals, even members of the same litter, will show little change in body weight for a period of several months, or they may even decline steadily in body weight although they are apparently in good physical condition. The increased variability in the body weights of older rats is, therefore, due in great part to the ac- cumulation of a greater or less amount of adipose tissue; it is not a growth phenomenon comparable to that shown during early postnatal life. In order to make a closer analysis of the relative variability in the body weights of animals in successive generations of the inbred strain, coefficients of variability were calculated from the body-weight data for the animals in three generations combined as summarized in table 7. This series of coefficients is shown in table 13. In table 13 the average coefficients for the male groups com- prising the individuals of the sixteenth to the twenty-fourth generations vary by less than one point, so it is evident that in the later generations of the inbred strain the variability in the body weights of the males did not decrease with the advance of inbreeding, as was the case in the earlier generations (King, ’18; table 16). The series of coefficients for the males of the twenty- fifth generation are, as a rule, smaller than the corresponding coefficients for the males of the preceding generation group. But the difference between the average coefficients for the two groups is less than three times the probable error, ‘so it cannot be considered as significant, especially as the number of body- weight records used in calculating the coefficients for the animals HELEN DEAN KING 96 Se Se See 9209'S 66° 09'S c0'1+6°6 00° T+T TT 99°02 2 8o'0FEe'9 TZ 0F8'L 69 06'S c9 OF TL 99° 0*9'L 99°0+2'9 LL: 0*8'8 L9°0*Z'8 ¥6' OT IL 02° 1T+9 FI GG WOT}BIIU9r) Go'O+F OT 19°0+ 2°01 99°0+9'°6 GG 0F8'8 SF 09'S tr O18 cr OFT'S 88° OF TZ OF 00'S €F 09'S €F 0*9'8 9F 0+0'6 69°09 TT 820 L°ST 68°0*€'8T 99° 0*9' ST 29 08 ZI F2-GE suor}B18uer) TS 026 GL OFF 6 89° 0*E'8 99° 0*6'8 €o'0OF8'8 br OF9 L Gc’ 0=T 8 8§'0*6'9 88°0-0'2 Ge-0F8'9 €€'0FF'9 68° 0F6'2 oo’ OE OL $9 OF € CL L9°OF LET Lo‘0*8 IT L9°0+*8'&T _ 18-61 SUO1}B.10U2F) Sa IVNGL 19°08 OL Go 16 TT 08 086 92° OF OT 99° 0*2'6 690 €'6 09'°0*9°6 69 08 '6 6F OF F'8 Lv 0F'8 Tg 0*?'6 FG OFZ OL 69° 0% Tit 9 OFT IT 08 08ST 6L 02 FI cL OF GET 8-91 SUOI}B.IOUI) 98°06 OL cf T*¥0'ST IG 13 GL 46°08 01 86°09 OT 66° 0*6 OT 00° T+0'IT ¢3°0*6'6 G8 0*6'6 €8°O=T OL 08 '0*9'6 S803 OT 18°09 OT 00° TT al 91°08 '8T Lo°0*6'9 19 OF LL 4 worpe19Uer) £9 OP SI 8L0FL 11 G9 09°01 86° OFF 01 19°0*E 11 9° 09°01 zo OF 01 99° OFF TT 69° OFZ ZI 79 0FZ'E1 So OF T'ZI 1¢°0*0°S1 890+ FI 8 0FG LI LOFT ST 79 OFF EI 19'0+8 ZI ¥E-ZE suor}e10udy 91° OF 8'°CE CL“Tsat 66 T*8'ST GSO F al 18 0*9 GI TZ 08 TT 69°0+*T OL 9 OF € TT 09°06 OL 89°02 01 69° OF TTT €9°0*6 IT €2Z'0+6 '&T 6809 ST 06°00 LT 69°O+T I GL OF 2 €1 1Z-61 SUOT}BIOUIY) Sa IVI 61°08 SI a ak i | GO'IT +3 IT LE‘T+6'ST ¢3°O+?F IT 69 OF OL 19 0*6'6 €9'0*€ OL 69 OF OL 19-0+T TE 19° OF IT G90 ZI €9'0*8'TT GL'0*36 FI ¥6'0*6 LT 0L°0*€ €T 99° OG cI 81-91 SUOT}RIOUEL) se eee - er VdVIDAY pabuojag sppnprarpur ay} yryn 07 WoLDLAUab ay} 07 Burpsoran sdno.b opur pajpundas (g ‘Y) $a2uas om) ay) fo swou pauque fo sabp yuasaffyp yo syybram fipog ay} wof ‘10.109 apqngoud w19y} yzun ‘uoynrupa fo syuaraifaoa ay, Buynoys el WIAVL EFFECTS OF INBREEDING ON BODY WEIGHT 97 of a single generation was only about one-third of that used for a group of three generations. The average coefficients for the three groups of females com- prising the animals in the sixteenth to the twenty-fourth gener- ations of the inbred strain are all lower than those for the corre- sponding groups of males (table 13), and they also fail to show a significant decrease in size as inbreeding advanced. The average coefficient for the body weights of the females in the twenty-fifth generation is considerably smaller than that for any of the three generation groups, but here also no definite conclusion seems warranted, since the small number of records on which the coefficients are based may be responsible in great measure for the result. The animals in the seventh to the fifteenth generations of the inbred strain lived under environmental and nutritive conditions that were fairly uniform and seemingly very favorable to growth and to fertility. The body weights of these individuals showed a slow decrease in variability with the advance of inbreeding, as the relative size of their coefficients of variability indicates (King, "18; table 16). During early life the rats in the sixteenth and seventeenth generations lived under the same environmental and nutritive conditions as the animals of the preceding generations, and at this time they were all seemingly somewhat less variable in body weight than were the individuals in the fifteenth gener- ation. Before the weight records for these rats were completed, a change in diet became necessary, as ‘scrap’ food of the required quality and quantity could no longer be obtained. The effects of the change in food became very apparent in the course of a few weeks, and, as individual rats responded differently to the altered conditions of nutrition, there was a marked increase in the variability of the body weights in the animals of all ages. When the coefficients of variability were calculated from the series of body-weight data obtained for the animals in the six- teenth to the eighteenth generations, they were found to be somewhat larger than those for the animals in the fifteenth generation, as was expected from the observed appearance of the animals. The animals in the later generations of the inbred 98 HELEN DEAN KING strain have shown a variability in body weights considerably greater than that found in any group of inbred animals since the tenth generation. By comparing the corresponding coefficents for the two series of outbred stock Albinos that were reared in the colony on differ- ent diets, one can determine whether the variability in the body weights of these animals was influenced by the nutritive con- ditions under which they lived. By a further comparison of these coefficients with those for the animals in the later generations of the inbred strain, it will be possible to determine whether the increase in the variability of the inbred animals was due to al- tered conditions of nutrition or to the effects of long-continued inbreeding. All of the stock Albinos reared in 1913 to 1915 as controls for the inbred series were fed on ‘scrap’ food. As has already been recorded (King, 715; table 4), the coefficients of variability for the body weights of the fifty males in this series range from 10.2 to 17.0, with an average of 13.6 for the entire group, taking all ages together; coefficients for the fifty females vary from 8.9 to 15.7, with an average of 11.5 for the entire group. The second series of stock controls was reared in 1916 to 1918 simultaneously with the inbred rats of the twenty-first to the twenty-fifth generations, and they, as the inbred rats, were fed on various experimental diets. These stock Albinos came from the same general stock colony that furnished animals for the first series of controls, so the coefficients for the two series are strictly comparable. An examination of the coefficients for the body weights of the rats in this control series, as given in table 11 of the present paper, shows that all of them are much larger than the corresponding coefficients for the animals of the first stock series, while the difference between the average coefficients for the two series is over four times the probable error. It is evident, therefore, that the rats in the second series of stock controls were much more variable in their body weights at all age periods than were the animals in the first stock series. Since both of these stock series were outbred, the increased variability in the animals of the second series cannot be attributed to the EFFECTS OF INBREEDING ON BODY WEIGHT 99 effects of inbreeding; nor can it be ascribed to a difference in the genetic constitution of the two series of animals, since no new ‘blood’ was introduced into the general stock colony from 1913 to 1917. From the evidence given, one seems warranted in assuming that the marked difference in the variability of the two series of stock animals was due, in great part, to the effects of changed conditions of nutrition which so greatly influenced the body growth of the individuals in the second series. It is probable also that the extremes of temperature to which many of these rats were subjected also affected their variability in body weight to some extent, although the effects of temperature changes were very much less than those of nutrition. Since the variability in the body weights of outbred stock Albinos was seemingly greatly affected by nutritive and environ- mental factors, one would naturally conclude that these factors would likewise influence the variability in the body weights of inbred animals reared simultaneously with and under the same conditions as the stock Albinos. The increased variability in the inbred animals of the sixteenth to the twenty-fifth gener- ations is, on this assumption, the result of environmental and nutritive action, and it cannot be cited in support of Walton’s (15) contention that continued inbreeding tends to increase variability. _ It is interesting to note in this connection that a comparison between the average coefficients for various groups of inbred rats and those for stock Albinos indicates that changed conditions of nutrition produced a much greater effect on the variability in the body weights of stock Albinos than it did on that of the animals in the later generations of the inbred strain. In this experiment, owing to the action of environment and of nutrition, it is impossible to determine the changes, if any, that inbreeding per se produced on the variability in the body weights of the animals in the later generations of the inbred strain. This study of variability is of value, therefore, mainly because it shows that in the later generations of inbreds there existed between the two series (A and B), and between the two sexes, the same relative variability in body weights as that found in the earlier generations. Twenty-five generations of brother 100 HELEN DEAN KING and sister matings have not, seemingly, altered the relative variability in the strain, whether the total amount of variability has been influenced by inbreeding cannot be determined until it is possible to rear a number of generations of these animals under uniform conditions of environment and of nutrition. GENERAL CONCLUSIONS As a whole, this experiment has shown that the closest form of inbreeding possible in mammals, the mating of brother and sister from the same litter, is not necessarily inimical either to body growth, to fertility, or to constitutional vigor, provided that only the best animals from a relatively large number are used for breeding purposes. Selection, seemingly, is able to hold in check any tendency that inbreeding may have to bring out the undesirable, latent traits inherent in the strain. In the course of this investigation it has been shown that adverse conditions of environment and of nutrition produce far more detrimental effects on growth and fertility in the albino rat than does inbreeding. These factors, apparently, do not alter the genetic constitution of the individual, since the animals soon resume their normal growth and fertility when environ- mental and nutritive conditions are again favorable. The sex ratio in the rat is seemingly a character that is amen- able to selection, since through this process the inbred strain has been separated into two lines: one line (A) showing a high sex ratio, the other line (B) showing a low sex ratio. The effects of selection on the sex ratio seem to be limited, however, since there has been no cumulative effects of the selection, although the two lines have been kept distinct for eighteen successive generations. Whether: it will be possible to change the sex ratio in the two lines by reversing the selection is the chief prob- lem in view in the continuation of this work. Throughout the entire course of this investigation there has been a great similarity between the two inbred series as regards the variability in the body weights of the animals at different age periods. In the earlier generations the variability in body EFFECTS OF INBREEDING ON BODY WEIGHT 101 weights seemed to decrease with the advance of inbreeding, but in the later generations the variability was greatly influenced by environmental and nutritive conditions. Until these latter factors can be controlled, it will not be possible to draw any definite conclusions regarding the effects of inbreeding per se on the variability in body weights. SUMMARY 1. The data given in the present paper show the growth and variability in the body weights of 296 males and of 310 females belonging in the sixteenth to the twenty-fifth generations of two series (A and B) of albino rats that were inbred, brother and sister from the same litter. ; 2. Owing to economic conditions, many of these rats were not reared under very favorable conditions of environment and of nutrition, and in consequence they did not grow as rapidly nor did they attain as great a maximum body weight as did the individuals in the earlier generations of this inbred strain. 3. In every generation from the sixteenth to the twenty-fifth the males were heavier than the females at all age periods after thirty days (tables 1 to 4). This result agrees with the finding for the inbred rats of the earlier generations, and also with that for various series of stock Albinos. Apparently, therefore, long- continued inbreeding has not changed the normal body-weight relations of the sexes at any age period for which records have been taken. 4. In the A series of inbreds the rate and extent of growth in body weight were much the same as those in the B series of inbreds: in the adult animals there was a difference of only about 2 per cent in the average body weights of corresponding groups of males and females in the two series (tables 8 and 9; fig. 5 and 6). 5. Close inbreeding for twenty-five generations has not altered the form of the growth graph for the albino rat to any extent. 6. Rats belonging to the later generations of the inbred strain were not as heavy at any age period as were the animals in the earlier generations, but they were much superior in body weight to stock Albinos reared under similar conditions of environment and of nutrition (figs. 7 and 8). 102 HELEN DEAN KING 7. Individuals in the sixteenth to the twenty-fifth generations of the inbred strain had about the same average body weight at different age periods as had the individuals of the stock controls reared in 1913 to 1915 under favorable conditions of environ- ment and of nutrition (figs. 7 and 8; compare graph B with graph C). Seemingly, therefore, inbreeding has as yet pro- duced no deterioration in the original Albino stock as regards the rate and extent of growth in body weight. 8. Variability in the body weights of the animals in the later generations of the inbred strain followed the same general trend as that in the animals of the earlier generations and in those of the two stock series studied: in both sexes it increased from birth to sixty days, and then decreased steadily until the animals were about 300 days of age, tending to rise again in older rats (table 12). 9. In the later generations of the inbred strain the males were more variable in body weight than the females. This result agrees with the finding for the animals of the earlier generations and for various series of stock Albinos. 10. In the inbred animals of the sixteenth to the twenty-fifth generations variability in body weights was relatively high, and it did not tend to decrease with the advance of inbreeding as in the earlier generation (table 13). . 11. Outbred stock Albinos, reared simultaneously with and under the same environmental and nutritive conditions as the inbred rats of the twenty-first to the twenty-fifth generations, showed a variability in their body weights at all ages much greater than that in the animals of the earlier stock series reared under more favorable conditions of nutrition. It appears, there- fore, that the increased variability in the body weights of the animals in the later generations of the inbred strain was due to the action of environment and of nutrition, not to the effect of continued inbreeding. EFFECTS OF INBREEDING ON BODY WEIGHT 103 LITERATURE CITED IN STUDIES ON INBREEDING I TO IV AnLFELD, Dr. 1876 Uber den Knabeniiberschuss der iilteren Erstgebirenden - nebst einem Beitrage zum Hofacker-Sadler’schen Gesetze. Arch. Gynikologie, Bd. 9. ALLEN, Ezra 1917 Spermatogenesis in the albino rat. Abstract in Proc. Amer. Soc. Zoologists. Anat. Rec., vol. 11. 1918 Studies on cell division in the albino rat (Mus norvegicus var. alb.) III. Spermatogenesis: the origin of the first spermatocytes and the organization of the chromosomes, including the accessory. Jour. Morp., vol. 31. AnpErson, W. S. 1911 A study in heredity. Report sixth annual Ky. State Farmer’s Institute. B. 1906 Ein Beitrag zur Wirkung des Geschlechtsverkehrs zwischen Blutsver- wandten. Kosmos, Bd. 3. Batrzer, F. 1914 Die Bestimmung des Geschlechts nebst eine Analyse des Geschlechtsdimorphismus bei Bonellia. Mittheil. Zodl. Sta. Neapoli, Bd. 22. Banta, A. M. 1916 Sex intergrades in a species of Crustacea. Proc. Nat. Acad. Sci., vol. 2. LO Basser, G. C. 1914 Habit formation in a strain of albino rats of less than normal brain weight. Behavior Monographs, vol. 2. Bipper, F. 1878 Uber den Einfluss des Alters der Mutter auf das Geschlecht des Kindes. Zeitschr. Geburtshilfe u. Gynakologie, Bd. 11. Born, G. 1881 Experimentelle Untersuchungen iiber die Entstehung der Geschlechtsunterschiede. Breslauer Arzt. Zeitschr., Bd. 3. Burcx, W. 1908 Darwin’s Kreuzungsgesetz und die Grundlagen der Bliiten- biologie. Rec. Trav. Bot. Néerl., Bd. 4. CastLe, W. E. 1896 The early embryology of Ciona intestinalis Flemming (L). Bull. Mus. Comp. Zool., vol. 27. 1903 The heredity of sex. Bull. Mus. Comp. Zool., vol. 40. 1916 Genetics and eugenics. Cambridge, Mass. 1916 a Size inheritance in guinea-pig crosses. Proc. Nat. Acad. Sci., vol. 2. 1916 b Variability under inbreeding and crossbreeding. Amer. Nat., vol. 50. CastLe, W. E., CarPENTER, F.W., et al. 1906 The effects of inbreeding, cross- breeding and selection upon the fertility and variability of Drosophila. Proc. Amer. Acad. Arts and Sci., vol. 41. Casrip, W. E., anp Puruuips, J.C. 1914 Piebald rats and selection. Carnegie Inst. Pub. no. 195, Washington, D. C. Castie, W. E., anp WricHT, SrwaLL 1916 Studies of inheritance in guinea- pigs and rats. I. An expedition to the home of the guinea-pig and some breeding experiments with material there obtained. Carnegie Inst. Pub. no. 241, Washington, D. C. CHapgaurouGcE, A DE 1909 Einiges iiber Inzucht und ihre Leistung auf ver- schiedenen Zuchtgebieten. Hamburg. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 29, No. 1 104 HELEN DEAN KING Cote, L. J., anv Kirkpatrick, W. F. 1915 Sex ratios in pigeons, together with observations on the laying, incubation and hatching of the eggs. Bull. no. 162, Agri. Exper. Sta., R. I. State College. Conrow, 8. B. 1915 Taillessness in the rat. Anat. Rec., vol. 9. 1917 Further observations on taillessness in the rat. Anat. Rec., vol. 12. : Corrman, 8. Monckton, AND Parsons, F. G. 1904 Observations on the sex in mice. Preliminary paper. Proc. Royal Soc., vol. 73. Cramer, H. 1883 Zucht-Versuche mit zahmen Wanderratten. I. Resultate der Zucht in Verwandtschaft. Landwirtschaftliche Jahrb., Bd. 12. 1884 Zucht-Versuche mit zahmen Wanderratten. II. Resultate der Kreuzung der zahmen Ratten mit wilden. Landwirthschaftliche Jahrb., Bd. 13. Cusnot, L. 1899 Sur la determination du sexe chez les animaux. Bull. Sci. de la France et de la Belgique, T. 32. Darwin, Cuartes 1875 The variation of animals and plants under domesti- cation. Second edition, London. 1878 The effects of cross and self fertilization in the vegetable king- dom. Second edition, London. Davenport, C. B. 1900 Review of von Guaita’s experiments in breeding mice. Biol. Bull., vol. 2. 1914 Statistical methods with special reference to biological variation. John Wiley & Sons, New York. Donatpson, H. H. 1906 A comparison of the white rat with man in respect to the growth of the entire body. Boas Anniversary volume, New York. 1912 The growth of the brain. New York. 1915 The rat. Data and reference tables. Mem. Wistar Institute of Anat. and Biol., no. 6, Philadelphia. Doncaster, L. 1913 On an inherited tendency to produce purely female fami- lies in Abraxas grossulariata, and its relation to an abnormal chromo- some number. Jour. Genetics, vol. 3. 1914 The determination of sex. Cambridge. 1914. Chromosomes, heredity and sex. Quart. Jour. Micr. Sci., vol. 59. Diistnc, Karu 1883 Die Factoren welche die Sexualitit entscheiden. Inaug. Dissertation, Jena. 1884 Die Regulierung des Geschlechtsverhiltnisses bei der Ver- mehrung der Menschen, Tiere und Pflanzen. Jen. Zeitschr. Natur- wiss., Bd. 17. 1886 Die experimentelle Priifung der Theorie von der Regulierung des Geschlechtsverhiltnisses. Jen. Zeitschr. Naturwiss., Bd. 19. 1887-92 Die Regulierung des Geschlechtsverhiltnisses bei Pferden. Landwirtschaftliche Jahrb., Bds. 16, 17, 21. East, E. M., anp Hayes, H. K. 1911 Inheritance in maize. Bull. no. 167, Conn. Agri. Exper. Station. 1912 Heterozygosis in evolution and in plant breeding. Bull. no. 243. U.S. Dept. Agri. EFFECTS OF INBREEDING ON BODY WEIGHT 105 Ewart, J.C. 1910 Principles of breeding and the origin of domesticated breeds of animals]. Bureau Animal Industry, Washington, D. C. Fasre-DomencugE, P. 1898 Unions consanguines chez les Columbins. L’inter- médiare des Biol., T. 1. Ferry, E. L. 1913 The rate of growth of the albino rat. Anat. Rec., vol. 7. Fiso, H. D. 1914 On the progressive increase of homozygosis in brother-sister matings. Amer. Nat., vol. 48. Gentry, N. W. 1905 Inbreeding Berkshires. Proc. Amer. Breeders Assoc., vol. 1. Gottp, H. N. 1917 Studies on sex in the hermaphrodite mollusk Crepidula plana. Jour. Exper. Zodl., vol. 23. GrizsHemm, A. 1881 Uber die Zahlenverhiltnisse der Geschlechter bei Rana fusca. Arch. ges. Physiol., Bd. 26. Guaita, G. von 1898 Versuche mit Kreuzungen von verscheidenen Rassen der Hausmaus. Ber. naturf. Gesellsch. zu Freiburg, Bd. 10. 1900 Zweite Mittheilung itiber Versuche mit Kreuzungen von ver- schiedenen Hausmausrassen. Ber. naturf. Gesellsch. zu Freiburg, Bd. 11. Guyer, M. F. 1909 On the sex of hybrid birds. Biol. Bull., vol. 16. 1910 Accessory chromosomes in man. Biol. Bull., vol. 19. Hammonp, J. 1914 On some factors controlling fertility in domestic animals. Jour. Agri. Sci., vol. 6. Hatar, S. 1907 Effect of partial starvation followed by a return to normal diet on the growth of the body and central nervous system of albino _ rats. Amer. Jour. Phys., vol. 17. Hayes, H. K., anp Jones, D. F. 1917 The effects of cross- and self-fertilization in tomatoes. Report Conn. Agri. Exper. Sta. Heare, W. 1899 Abortion, barrenness and fertility in sheep. Jour. Royal Agri. Soc., vol. 10. P . 1908 Note on the proportion of the sexes in dogs. Proc. Cambridge Phil. Soe., vol. 14. 1909 The proportion of the sexes produced by white and coloured peo- ples in Cuba. Phil. Trans. Royal Soc., vol. 200. Hertwia, R. 1906 Untersuchungen iiber das Sexualitaétsproblem. Verhandl. deutsch. zool. Gesellsch. 1907 Weitere Untersuchungen iiber das Sexualitaitsproblem. Ver- hand]. deutsch. zool. Gesellsch. Hirscu, M. 1913 Uber das Verhiltnis der Geschlechter. Centralbl. fiir Gyn- ikologie, Bd. 37. Hoskins, E. R. 1916 The growth of the body and organs of the albino rat as affected by feeding various ductless glands (thyroid, thymus, hypo- physis and pineal). Jour. Exper. Zodél., vol. 21. Hours, A.H. 1887 The marriage of near kin. 2nd Edition. London. Hype, R. R. 1914 Fertility and sterility in Drosophila ampelophila. I. Ster- ility in Drosophila with special reference to a defect in the female and its behavior in heredity. Jour. Exper. Zodél., vol. 17. 1914 Fertility and sterility in Drosophila ampelophila. II. Fer- tility in Drosophila and its behavior in heredity. Jour. Exper. Zodl., vol. 17. 106 HELEN DEAN KING Issakowi1TscH, A. 1905 Geschlechtsbestimmenden Ursachen bei den Daphni- den. Biol. Centralbl., Bd. 25. Jackson, C. M. 1912 On the recognition of sex through external characters in the young rat. Biol. Bull., vol. 23. 1913 Postnatal growth and variability of the body and of the various organs in the albino rat. Am. Jour. Anat., vol. 15. 1915 Changes in the relative weights of the various parts, systems and organs of young albino rats held at constant body weight by under- feeding for various periods. Jour. Exper. Zodl., vol. 19. JenninNGs, H. 8. 1916 Heredity, variation and the results of selection in the uniparental reproduction of Difflugia corona. Genetics, vol. 1. JOHANNSEN, W. 1909 Elemente der exakten Erblichkeitslehre. Jena. Jones, D. F. 1918 The effects of inbreeding and crossbreeding upon develop- ment. Conn. Agri. Exper. Station, Bull. no. 207. Jorpan, H. E. 1911 The spermatogenesis of the opossum. Arch. Zellforsch., Bd. 7. Kine, Heten Dean 1907 Food as a factor in thedetermination of sex in am- phibians. Biol. Bull., vol. 13. 1911 Studies on sex determination in amphibians. IV. The effects of external factors, acting before or during the time of fertilization, on the sex ratio of Bufo lentiginosus. Biol. Bull., vol. 20. 1911 a The sex ratio in hybrid rats. Biol. Bull., vol. 21. 1912 Studies on sex determination in amphibians. V. The effects of changing the water content of the egg, at or before the time of fertilization, on the sex ratio of Bufo lentiginosus. Jour. Exper. Zo6l., vol. 12. 1915 Growth and variability in the body weight of the albino rat. Anat. Rec., vol. 9. 1915 a On the weight of the albino rat at birth and the factors that influence it. Anat. Rec., vol. 9. 1916 On the postnatal growth of the body and of the central nervous system in albino rats that are undersized at birth. Anat. Rec., vol. 11. 1916a The relation of age to fertility in the rat. Anat. Rec., vol. 11. 1918 Studies on inbreeding. I. The effects of inbreeding on the growth and variability in the body weight of the albino rat. Jour. Exper. Zoél., vol. 26. 1918 a Studies on inbreeding. II. The effects of inbreeding on the fertility and on the constitutional vigor of the albino rat. Jour. Exper. Zo6l., vol. 26. 1918 b Studies on inbreeding. III. The effects of inbreeding, with selection, on the sex ratio of the albino rat. Jour. Exper. Zoél., vol. 27. Kinc, HELEN Dean, AND SToTSENBURG, J. M. 1915 On the normal sex ratio and the size of the litter in the albino rat (Mus norvegicus albinus). Anat. Rec., vol. 9. Kirxnam, W. B., anp Burr, H. S, 1913 The breeding habits, maturation of eggs and ovulation of the albino rat. Am. Jour. Anat., vol. 15. Ko.azy, J. 1871 Uber die Lebensweise von Mus rattus, varietas alba. Ver- handl. zool. bot. Gesellsch., Wien. EFFECTS OF INBREEDING ON BODY WEIGHT 107 Kraemer, H. 1913 Uber die ungiinstigen Wirkungen naher Inzucht. Mittheil. der deutsch. Landwirtsch. KuscHakEwitTscH, 8. 1910 Die Entwicklungsgeschichte der Keimdriisen von Rana esculenta. Festschr. zum sechzigsten Geburtstag Richard Hertwig 11. Jena. Lantz, D.E. 1910 Natural history of the rat. Bull. Public Health and Marine Hospital Service of the U.S. Washington, D.C. Luorp, R. E. 1911 The inheritance of fertility. Biometrika, vol. 8. 1912 The growth of groups in the animal kingdom. London. Lors, Leo 1917 The experimental production of hypotypical ovaries through underfeeding. A contribution to the analysis of sterility. Biol. Bull., vol. 33. McCivne, C. E. 1902 Notes on the accessory chromosome. Anat. Anz., Bd. 20. 1902 a The accessory chromosome-sex determinant? Biol. Bull., vol. 3. McCoutivm, C. V. 1918 The newer knowledge of nutrition, MacMillan Co., New York. MarsHaL., F. H. A. 1908 Fertility in Scottish sheep. Trans. Highland Agri. Soc., vol. 20. 1908 a The effects of environment and nutrition on fertility. Science Progress, vol. 2. 1910 The physiology of reproduction. London. Mitter, Newron 1911 Reproduction in the brown rat (Mus norvegicus). Amer. Nat., vol. 45. Minot, C.S. 1891 Senescence and rejuvenation. 1. On the weight of guinea- pigs. Jour. Phys., vol. 12. MircHett, A. 1865 On the influence which consanguinity in the parentage exercises on the offspring. Edinburgh Med. Jour., vol. 10. Mitcuett, C. W. 1913 Sex-determination in Asplanchna amphora. Jour. Exper. Zodl., vol. 15. Moenxuaus, W.J. 1911 The effects of inbreeding and selection on the fertility, vigor and sex-ratio of Drosophila ampelophila. Jour. Morph., vol. 22. Monrcomery, T. H. 1908 The sex ratio and cocooning habits of an aranead and the genesis of the sex ratio. Jour. Exper. Zodl., vol. 5. 1911 The cellular basis of the determination of sex. Internat. Clin- ies, vol. 1. Morean, T. H. 1904 Self-fertilization induced by artificial means. Jour. Exper. Zool., vol. 1. 1905 Some further experiments on self-fertilization in Ciona. Biol. Bull., vol. 8. 1911 An alternation of the sex-ratio induced by hybridization. Proc. Soc. Exper. Biol. and Med., vol. 8. 1914 Heredity and sex. Second edition. New York. 1914 a Two sex-linked lethal factors in Drosophila and their influ- ence on the sex ratio. Jour. Exper. Zool., vol. 17. Morean, T. H., Payne, F., anp Browne, E. N. 1910 A method to test the hypothesis of selective fertilization. Biol. Bull., vol. 18. 108 HELEN DEAN KING Morean, T. H., Sturtevant, A. H., er au. 1915 The mechanism of Mendelian Inheritance. New York. Newcomes, 8. 1904 A statistical inquiry into the probability of causes of the production of sex in human offspring. Carnegie Institution, Wash- ington, D. C. Nicuots, J. B. 1907 The numerical proportions of the sexes at birth. Mem. Amer. Anthropological Assoc., vol. 1. Osporne, T. B., anpD Menpet, L. B. 1914 The suppression of growth and the capacity to grow. Jour. Biol. Chem., vol. 18. : 1915 The resumption of growth after long continned failure to grow. Jour. Biol. Chem., vol. 23. 1916 Acceleration of growth after retardation. Amer. Jour. Phys., vol. 40. OsBorNE, T. B., MENDEL, L. B., AND Ferry, E.L. 1917 The effects of retarda- tion of growth upon the breeding period and duration of life of rats. Science, vol. 45. PApANIcoLAou, G. 1915 Sex determination and sex control in guinea-pigs. Science, vol. 41. , PEARL, R. 1912 Mendelian inheritance of fecundity in the domestic fowl. Amer. Nat., vol. 46. 1912 a The mode of inheritance of fecundity in the domestic fowl. Jour. Exper. Zool., vol. 13. 1913 A contribution towards an analysis of the problem of inbreeding. Amer. Nat., vol. 47. 1915 Modes of research in genetics. MacMillan & Co., New York. 1917 The selection problem. Amer. Nat., vol. 51. 1917 a Studies on the physiology of reproduction in the domestic fowl. XVII. The influence of age upon reproductive ability, with a description of a new reproductive index. Genetics, vol. 2. Peart, R., anp Parstey, H. M. 1913 Data on sex determination in cattle. Biol. Bull., vol. 24, 1913. Peart, R., AND Peart, M. D. 1908 On the relation of race crossing to sex ratio. Biol. Bull., vol. 15. Prart, R., AND SauaMan, R. N. 1913. The relative time of fertilization of the ovum and the sex ratio amongst Jews. Amer. Anthropologist, vol. 15, 1913. PreaRL, R., anp Surrace, F. M. 1909 Data on the inheritance of fecundity obtained from the records of egg production in the daughters of ‘200- egg’ hens. Me. Agri. Exper. Sta. Bull. no. 166. Pearson, K., Len, A., AND Bramiey-Moore. 1899 Mathematical contri- butions to the theory of evolution. VI. Genetic (reproductive) se- lection: Inheritance of fertility in man, and of fecundity in thorough- bred racehorses. Phil. Trans. Royal Soe., vol. 192. Pearson, K., Scuuster, E.H.J., AnD WELDON, W.F.R. 1903 Assortive mating inman. Biometrika, vol. 2. PFrLtcer, E. 1881 Einige Beobachtungen zur Frage iiber die das Geschlechts- bestimmenden Ursachen. Arch. ges. Physiol., Bd. 26. EFFECTS OF INBREEDING ON BODY WEIGHT 109 Puiuuirs, J.C. 1914 A further study of size-inheritance in ducks, with obser- vations on the sex ratio of hybrid birds. Jour. Exper. Zool., vol. 16. Pixs, F.H. 1907 A critical and statistical study of the determination of sex, particularly in human offspring. Amer. Nat., vol. 41. Porenor, P. 1917 An experiment in long-continued inbreeding. Jour. Heredity, vol. 8. Powys, A. O. 1905 Data for the problem of evolution in man. On fertility duration of life and reproductive selection. Biometrika, vol. 4. Punnetr, R. C. 1903 On nutrition and sex-determination in man. Proc. Cambridge Phil. Soc., vol. 12. Punnett, R. C., anp Battey, P.G. 1914 On inheritance of weight in poultry. Jour. Genetics, vol. 4. QuaKENBUSH, L. S. 1910 Unisexual broods of Drosophila. Science, vol. 32. Ravuser,’A 1900 Der Uberschuss an Knabengeburten und seine biologische Bedeutung. Leipzig. Rawtis, E. 1913 Sex ratios in Drosophila ampelophila. Jour. Exper. Zodl., vol, 24. Rippie, O. 1914 The determination of sex and its experimental control. Bull. Amer. Acad. Med., vol. 15. 1916 Sex control and known correlations in pigeons. Amer. Nat., vol. 50. 1917 The control of the sex ratio. Jour. Washington Acad. Sci., vol. 7. 1917 a The theory of sex as stated in terms of results of studies on pigeons. Science, vol. 46. RirzeMA-Bos, J. 1893 Onderzoekingen aangaande de gevolgen van de teelt in Bloedverwantschap. Handelingen van het vierde nederlandsche natureren geneeskundig Congres te Houden te Groningen. 1894 Untersuchungen iiber die Folge der Zucht in engster Blut- verwandtschaft. Biol. Centralbl., Bd. 14. RomMELL, G. M., anp Puiuuips, E. F. 1906 Inheritance in the female line of size of litter in Poland China sows. Proc. Amer. Phil. Soc., vol. 49. ScutecuTer, J. 1884 Uber die Ursachen welche das Geschlecht bestimmen. Biol. Centralbl., Bd. 4. Scuutrzp, O. 1902 Was lehren uns Beobachtung und Experiment iiber die Ursachen mannlicher und weiblichen Geschlechtsbildung bei Tieren und Pflanzen? Sitzungsber. phys.-med. Gesellsch. zu Wiirzburg. 1903 Zur Frage von den geschlechtsbildenden Ursachen. Arch. mikr. Anat., Bd. 43. SuHaMEL, A. D. 1905 The effects of inbreeding in plants. Yearbook U. S. Dept. Agri. Suuit, A. Franxuin 1910 Studies in the life cycle of Hydatina senta. 1. Artificial control of the transition from the parthenogenetic to the sexual method of reproduction. Jour. Exper. Zodl., vol. 8. 1912 Studies in the life cycle of Hydatina senta. II. Internal factors influencing the proportion of male producers. Jour. Exper. Zodl., vol, 12. 110 HELEN DEAN KING Suuuy, A. Franxurn 1913 Inheritance in Hydatina senta. 1. Viability of the resting eggs and the sex ratio. Jour. Exper. Zodl., vol. 15. Suu, A. FRANKLIN, AND Laporr, 8. 1916 Factors affecting male-production in Hydatina. Jour. Exper. Zodl., vol. 21. Suuut, G. H. 1910 Hybridizaton methods in corn breeding. Amer. Breeders Magazine, vol. 1. SLtonaKeR, J. R. 1912 The effects of a strictly vegetable diet on the spontane- ous activity, the rate of growth and the longevity of the albino rat. Pub. Leland Stanford Jr. Univ. 1912a The normal activity of the albino rat from birth to natural death, its rate of growth and the duration of life. Jour. Animal Be- havior, vol. 2. Stonaker, J. R., anp Carp, T. A. 1918 The effect of omnivorous and vegeta- rian diets on reproduction in the albino rat. Science, vol. 47.+ Smitu, G., anD THomas, Mrs. Hata 1913 On sterile and hybrid pheasants. Jour. Genetics., vol. 3. Stevens, N.M. 1905 Studies in spermatogenesis with special reference to the ‘accessory chromosome.’ Pub. Carnegie Institution, Washington, Die: 1911 Preliminary note on heterochromosomes in the guinea-pig. Biol. Bull., vol. 20. 1911 a Heterochromosomes in the guinea-pig. Biol. Bull., vol. 21. Stewart, C. A. 1916 Growth of the body and of the various organs of young albino rats after inanition for various periods. Biol. Bull., vol. 31. SrorsENBURG, J. M. 1915 On the growth of the fetus of the albino rat from the thirteenth to the twenty-second day of gestation. Anat. Rec., vol. 9. Strout, A. B. 1916 Self- and cross-pollinations in Cichorium intybus with reference to sterility. Mem. New York Bot. Garden, vol. 6. Sumner, F. B. 1909 Some effects of external conditions upon the white mouse. Jour. Exper. Zool., vol. 7. 1915 Some studies of environmental influence, heredity, correlation and growth in the white mouse. Jour. Exper. Zool., vol. 18. Tuury, M. 1864 Uber das Gesetz der Erzeugung der Geschlechter bei den Pflanzen, den Thiere und den Menschen. Leipzig. Ursurikawa, N. 1917 Temperamental differences between outbred and in- bred strains of the albino rat. Jour. Animal Behavior, vol. 7. Watton, L. B. 1915 Variability and amphimixis. Amer. Nat., vol. 49. Warren, D. C. 1918 The effect of selection upon the sex-ratio in Drosophila ampelophila. Biol. Bull., vol. 34. Watson, C. 1906 Observations on diet. The influence of diet on growth and nutrition. Jour. Phys., vol. 34. Watson, T. B. 1905 The effects of bearing young upon the body weight j;and the weight of the central nervous system of the female albino rat. Jour. Comp. Neur., vol. 15. Wentworth, E. N. 1913 The segregation of fecundity factors in Drosophila. Jour. Genetics, vol. 3. Wentworth, E. N., ano AuseL, C. E. 1916 Inheritance of fertility in swine. Jour. Agri. Research, vol. 5. EFFECTS OF INBREEDING ON BODY WEIGHT 111 Waite, F.W. 1914 Variation in the sex ratio of Mus rattus associated with an unusual mortality of adult females. Proc. Royal Soc., vol. 87. Wuitney, D. D. 1914 The influence of food in controlling sex in Hydatina senta. Jour. Exper. Zool., vol. 17. 1914 The production of males and females controlled by food con- ditions in Hydatina senta. Science, vol. 39. 1916 The control of sex by food in five species of Rotifers. Jour. Exper. Zool., vol. 20. 1917 The relative influence of food and oxygen in controlling sex in Rotifers. Jour. Exper. Zool., vol. 24. Wickens, M. 1886 Untersuchungen iiber das Geschlechtsverhiltniss und die Ursachen der Geschlechtsbildung bei Haustieren. Biol. Centralbl., Bd. 5. 1886 a Untersuchungen iiber das Geschlechtsverhiltniss und die Ursachen der Geschlechtsbildung bei Haustieren. Landwirthschaft- liche Jahrb., Bd. 15. Witson, E. B. 1905 Studies on chromosomes. II. The paired microchromo- somes, idiochromosomes and heterotropic chromosomes in Hemiptera. Jour. Exper. Zodl., vol. 2. 1910 Selective fertilization and the relation of the chromosomes to sex-production. Science, vol. 32. WILSON, JAMES 1912 The principles of stock-breeding. London. WINIWARTER, H. von, 1912 Etudes sur la spermatogenese humaine. Arch. Biol., Bd. 27. WopsEDALEK, J. E. 1913 Spermatogenesis of the pig with special reference to the accessory chromosomes. Biol. Bull., vol. 25. Wooprurr, L. L. 1911 Two thousand generations of paramecium. Arch. Protistenkunde, Bd. 21. Wriept, C. 1916 What they say about inbreeding in Europe. Jour. Heredity, vol. 7. YerKeES, A. W. 1916 Comparison of the behavior of stock and inbred albino rats. Jour. Animal Behavior, vol. 6. YerKES, R. M. 1913 The heredity of savageness and wildness in rats. Jour. Animal Behavior, vol. 3. ZELENY, C., AND Faust, E. C. 1915 Size dimorphism in the spermatozoa from single testes. Jour. Exper. Zodl., vol. 18. Resumen por el autor, Harley Nathan Gould. Universidad de Pittsburgh. Estudios sobre el sexo en el molusco hermafrodita Cre pidula plana. III. Transmisi6n del estimulo productor de machos por el agua de mar. E] molusco gaster6podo Crepidula plana pasa durante su vida por una fase de macho, una fase de transicién y una fase de hembra. La fase de macho es inestable y se presenta solamente como resultado de un estimulo susministrado por un individuo de la misma especie mds grande que el individuo estimulado. E] aislamiento completo de los individuos pequefios no desarrol- lados sexualmente, durante largos periodos, demuestra que bajo tales condiciones no tiene lugar mds desarrollo de los caracteres machos que la formacién de unas pocas espermatogonias. En su debido tiempo aparecen los caracteres de la hembra. Los individuos pequefos y no desarrollados sexualmente confinados a distancias fijas de 4 a 7 mm. de hembras grandes, impidiéndose de este modo todo contacto, desarrollan en la mayor parte de los casos caracteres del macho en varios estados de madurez sexual. Bajo tales condiciones se producen menos machos y peor desar- rollados que cuando los animales pequefios estin mas cerca del origen del estimulo. Los individuos grandes de Crepidula forni- cata, una especie préxima a Crepidula plana, no inducen desar- rollo alguno sobre los individuos pequefos de esta ultima especie, excepto en unos cuantos casos dudosos. El estimulo que provoca el desarrollo de machos acttia de tal modo que indica que es una substancia que sale de los cuerpos de los individuos grandes de Crepidula plana, la cual substancia es difusible en el agua de mar, pero es muy inestable. Translation by José I’. Nonidez Carnegie Institution of Washington AUTHOR’S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, AUGUST 4 STUDIES ON SEX IN THE HERMAPHRODITE MOLLUSC CREPIDULA PLANA Il]. TRANSFERENCE OF THE MALE-PRODUCING STIMULUS THROUGH SEA-WATER HARLEY N. GOULD Marine Biological Laboratory, Woods Hole, and School of Medicine, University of Piitsburgh ONE TEXT FIGURE The second paper of this series! described a number of experi- ments showing the instability of the male phase in the marine gastropod Crepidula plana. In common with other members of the family Calyptraeidae, C. plana passes through a sperm-pro- ducing phase during the early part of its life while it is small (up to about 15 mm. in length) followed by a transitional phase (15 to 20 mm.) and later by an egg-producing phase (20 to 40 mm.). Growth goes on with varying degrees of rapidity during life. The functional females are the largest and oldest. The species has a peculiarity in that the development and mainte- nance of the male phase requires a stimulus from the outside, which is furnished by the presence of a larger individual, usually transitional or female, in the immediate vicinity of the potential male. The animals are most commonly found in colonies adhering to the inner surface of shells inhabited by hermit crabs. The younger, smaller Crepidulas have various degrees of male devel- opment, those directly attached upon the shells of the large fe- males as a substratum, or close beside them, being nearly all fully developed males, while those at a distance of 5 mm. or over are more likely to have only partially developed male or- gans; the degree of development being less in the specimens far- ther from the source of the stimulus, i.e., the large individuals of the colony. 1 Gould, 1917, II. 113 114 HARLEY N. GOULD In a group in which there are no females and all the members are less than 10 or 12 mm. in length, there are seldom any adult males; the majority being, instead, sexually undeveloped (neu- ter); but often the smaller members of such a group have a rudi- mentary male development, evidenced by the presence of many spermatogonia in the sex gland, even some spermatogenesis and a rudimentary penis. In fact, wherever two members of the species are attached close together, however insignificant the difference in size between them, the smaller tends to begin male development. ISOLATION OF NEUTERS The adult male stage is never developed in isolated animals, nor can it be maintained after removal of a male from the colony. Wishing, however, to determine whether any partial develop- ment of male characters would take place in completely isolated specimens, the writer allowed young neuter animals to attach themselves to the inner surface of glass vials, one to each vial. These were all kept in salt-water aquaria. Selection of the speci- mens for the experiment was made with care from hermit shells containing only a new small C. plana. Each was examined with a lens, and only those quite devoid of rudimentary male charac- ters were used. After isolation a few specimens were taken from time to time, examined, then fixed and sectioned for study of the gonad. At the beginning all were from 5 to 12 mm. in length, and were thus at the size when male development can easily be induced. They grew during the period of isolation, and the last lot, taken at fifty days, were much larger. Slides were made from twenty-four specimens; two at twenty-two days’ isolation,. four at twenty-four days, three at twenty-six days, five at thirty- three days, five at forty-three days, and five at fifty days. The results may be summarized as follows: External male characters: In three animals only, two twenty-four days and one at thirty-three days, there was a very small stump at the spot where the penis forms. No other exter- nal signs of the male condition appeared. STUDIES ON SEX IN CREPIDULA 115 Gonad: In three cases there were a few spermatogonia in the sex gland; one at twenty-two days, one at twenty-four days, and one at forty-three days. None of these corresponded with any one of the three having a rudimentary penis. Jn sixteen cases the gonad was inactive (containing only primordial male and female cells). The remaining five were the animals sectioned after fifty days’ isolation. They had passed from the neuter to the incipient female condition, having various stages in early growth periods of oocytes, and had grown considerably in size, being now from 14 to 23 mm. in length. A similar record was made of males removed from colonies and kept isolated in vials. As was shown in a former paper, all males lose their male characters after removal from the colonies. Four samples were taken from the vials after thirty-six days, four after forty-six days, four after fifty-three days, and four after sixty days. There was no resumption of spermatogenesis or redevelopment of external male organs after the degeneration in any case. The only hint of any such activity was the presence of a few dividing spermatogonia in the gonad of one isolated forty-six days. It should be recalled that previous experiments demonstrated the ability of degenerate males to reassume the functional male state under stimulus from larger individuals. It is thus indicated that the gonads of isolated small specimens may produce a few spermatogonia, but proceed no further toward spermatogenesis; and the spermatogonia so formed later degener- ate, as sections show. ‘The isolation experiment is meant to clear the way for others, i.e., to show, in cases where partial male de- velopment is induced under weak stimulus, how much of this is due to internal causes. The writer concludes that rapid sperma- togonial multiplication, formation of spermatocytes, or any later stage of spermatogenesis is an indication of an external stimulus. In previous experiments where they developed male charac- ters under observation, the neuters were placed as closely as possible to the larger animals. Only in this way could the stimu- lus be clearly shown. The writer failed to find positive evidence 116 HARLEY N. GOULD of a stimulating secretion thrown into the sea-water. The ques- tion arose whether physical contact is necessary for the trans- ference of the stimulus. STIMULUS WITHOUT CONTACT A simple apparatus (fig. 1) was devised to hold a large female Crepidula at a definite distance from a small neuter without al- lowing them to touch or to move farther apart. The female was removed from the inner surface of a hermit crab’s shell and al- lowed to attach herself to the concave surface of a watch crystal. The small neuter was placed on the floor of the flat-bottomed de- pression in a hollow-ground slide. Mosquito netting was fas- tened over the depression to prevent the neuter from escaping. Fig. 1 Diagram showing method of preventing contact between specimens. a, large female; b, small neuter; c, depression slide; d, watch crystal. The depression slide was inverted and fastened over the watch crystal containing the large female; leaving the neuter animal, imprisoned in its cell, at a distance of from 4 to 7 mm. from the top of the female’s shell. There was no possibility of contact, yet there was little hindrance to diffusion currents in the sea- water between the two. The variation in the distance between female and neuter was due to irregularities in the curvature of the watch crystal and in depth of depression of the slide. The average distance was 6 mm. After various periods, samples a the originally neuter C. plana were fixed and sectioned. The results are tabulated below (table 1). Those specimens the gonads of which showed any male de- velopment beyond the mere presence of spermatogonia are marked ‘male.’ ‘Inactives’ are specimens with primordial germ cells only, or with these plus spermatogonia. ‘Females’ are animals where some development of oocytes could be detected. STUDIES ON SEX IN CREPIDULA 117 TABLE 1 NUMBER OF NUMBER OF NUMBER OF NUMBER OF BE STION SPECIMENS MALES INACTIVES FEMALES days | 14 74 4 3 0 15 20 11 3 6 17 21 14 3 4 21 19 14 5 0 ete ot. 67 43 14 | 10 Thus, forty-three out of sixty-seven, or about 64 per cent, showed spermatogenetic activity of some sort more than isolated neuters show. Classifying these forty-three with regard to de- gree of male development, we have: fully developed testis, twenty- five; testis containing sperm, but with some missing stages of spermatogenesis, four; testis developed as far as spermatids, eight; spermatogonia and spermatocytes, two; spermatogonia in multiplication period, four. The occurrence of incipient female development in some of the specimens will be understood if we assort them all in the order of their size, indicated by the length of the shell in millimeters. This is done in table 2. All those having early stages of developing oocytes (‘female’) are seen to be among the larger animals used for the experiment. TABLE 2 LENGTH NUMBER OF MALES NUMBER OF INACTIVES | NUMBER OF FEMALES | _ oOo © a _ Co Sooo orFrwunQnoan ae ee We DOK OOO SC SOON WrRWeH NNO 118 HARLEY N. GOULD Female development is much slower than male, and it is likely that the most of these animals were already in the course of fe- male differentiation when selected as neuters. The percentage of ‘inactives’ is also greater among the larger specimens. It has been evident to the writer from many observations that the tend- ency to male development under stimulus gradually wanes as the period approaches when female development may set in. It is, however, sometimes possible to superimpose male on early female development, as shown in the former paper. In the watch-crystal experiment forty-three animals out of a possible sixty-seven showed some degree of male activity in the sex gland, twenty-five of them being fully developed males. Compare this with the result obtained when neuters were placed on and closely around females. In the latter case (from records in previous paper) fifty-one out of a possible fifty-three showed some degree of male development, and thirty-four of them were adult males. It is clear that more males develop when the neuters are close to the source of the stimulus than when sep- arated by several millimeters; and furthermore, the difference in the results of these two experiments cannot be adequately set forth in tabular form. Examination of the gonad under the mi- croscope shows it more strikingly. Many marked ‘adult testis’ in specimens from the watch-crystal experiment are only a fraction of the size of the gonads developed in those placed close to or on the large females. There are often signs of arrested development in the former, shown by the paucity or absence of some stages of spermatogenesis. An examination of the small individuals in a large number of normal colonies shows about 62 per cent adult males (determined from external characters). By placing neuters on and close to females, about the same percentage of adult males was obtained, and this could have been raised considerably by rejecting all those specimens which had moved several millimeters from the females during the course of the experiment. In the watch- crystal experiment only about 38 per cent became adult males. The development of the male phase by neuters imprisoned in depression slides thus shows that the male-producing stimulus is STUDIES ON SEX IN CREPIDULA 119 ‘able to act in the absence of physical contact and through several millimeters distance in sea-water. A comparison with other ex- periments indicates that fewer and less fully developed males are produced under such conditions than when the stimulus acts more directly. The writer has tried several times to determine whether a large female of Crepidula fornicata, another species of the same genus, could furnish the stimulus for male development in a small neuter Crepidula plana. The experiment has been difficult to carry out, as the little C. plana neuters were generally crushed by the twisting and turning movements of the great C. fornicata before sufficient time elapsed to make the experiment valuable. The writer has, however, slides made from thirty-two C. plana selected as neuters and kept near the C. fornicata for various periods. Of these, twenty-one remained entirely neuter and six became incipient females. The remaining five show traces of male development. Two of these must be counted out because the microscopic appearance of the gonad shows that the few prod- ucts of spermatogenesis there must have been formed and fur- ther activity must have ceased before the experiment began. This leaves only three which seem to have developed any male characters during association with C. fornicata, and they are as follows: a. Penis partly developed and small testis as far as spermatids, not very active. ‘Time, seventeen days. b. No penis. A few spermatogonia and spermatocytes. Time, eleven days. c. Nopenis. Spermatogonia and a few spermatocytes. Time, eight days. Thus, there are no adult males developed out of twenty-six neuter specimens (leaving out of consideration those which had begun female differentiation), but there are three with partial male development during the experiment. This result is rather perplexing. One would naturally expect either an appreciable proportion of males, if the C. fornicata exerted any influence, or none at all, if they did not. However, we may draw the conclu- sion that the male-producing stimulus is not due to any general THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 29, No. 1 120 HARLEY N. GOULD change in the medium (sea-water) caused by C. plana which would be similarly caused by other species. The C. fornicata females used for the experiment were larger than the largest C. plana females, and would be expected to throw into the sea- water at least as much of the general katabolic products, for in- stance, as the latter, yet they had almost no effect in stimulating development of the testis. It should be emphasized that the power of large animals of the species C. plana to stimulate spermatogenesis in the smaller is not limited to females. A number of unusually large males were removed from a colony and imprisoned in a watch crystal with nine small neuters. In eighteen days five of the nine showed some degree of male development, mostly immature. The large males were in the meantime losing their male characters. They were kept forty days after this losing all signs of maleness and growing larger. A second lot of nine small neuters was placed with them. In sixteen days eight of the nine had some degree of male development, averaging nearer the mature male phase than the first nine. The numbers are too small to speak for the relative effectiveness of large males and large transitionals, but show the ability of both to produce the stimulus. SUMMARY The stimulus passing from larger to smaller Crepidula plana, causing the latter to assume and retain the male phase, can be transmitted for several millimeters through sea-water, though its effectiveness is reduced at this distance. Indication that the stimulus may be. given faintly by Crepidula fornicata, a related species, was given in only three out of twenty-six cases. The stimulus acts in such a manner as to suggest that it is a specific substance given off from the bodies of the animals, dif- fusible in sea-water, but very unstable. BIBLIOGRAPHY Goutp, H. N. 1917 Studies on sex in the hermaphrodite molluse Crepidula plana. I. History of the sexual cycle. Jour. Exp. Zodl., v. 23, no. 1. II. Influence of environment on sex. Jour. Exp. Zoél., v. 23, no. 2. hetiy nf i oo svemiel sl 4 - re a ’ a. tlw se Saya aga olnmeracy ty? alent ey re RTL AUR: 1 CE LAS i . . af - oranihn ts es 9’. P a int tht 30 HU p. F Aiteis iRise Bho ») be) PAT ee ere | an il fae et ae) aby AVESEROO Ott Le Rea ars . yale: Bb Si Hititoaet tos “ay, eons! > Ce dnl ini usps hes 1 PORSTIE it DRT I «rst eT tee wil tt eIVINDOC Ree tO aes “Beng ( feb Phas a hii ee a OvU A Ab au! ) ; ‘ae 8d ak) anh he ia wey Let ne thy Be! . areal ree ace Oh wires le) bp wom bavibn “(Oat lab ng> . _— rag . ey . * . a. 5 anes ri 74. (Cans eer aa {Si OD Pas ee: Na, ret, ae shy . 4 . ’ ’ ’ " A t 7 aire £4 fhe 1} ie Panos Aa EOLA ISS Ci SOAOIRIVAD 0b fa : | BA c _— Wiss 1 Sty "Pl ry | , ii aie! tT in Lin 3! titer Ab Be - f “id @e.#) ats “if wD Jak ate ORM { Mis oth! er 8) eae 5 f f oGroi) A} Oh, re te ea Laie ‘e Pee a ial} V8 pk fans pis (ous a *taalon | Hie 3, a Mit neve mao . BAe: | ete Mia Mees WL en Mn mrNaNe rh Wliises exer SSI nie | Sh eee txt tt : " ] ee ( 2 f ny Ph hth) “Lie ny ( on , ys am, aye ey Me hyskohtiy piv cl) y Ue ae aay : 4) weet) ie pie Tamed Ag 1 As oe w ud, a me OR patel ie ype ) chav. ipsiat py Prins eee is aryl ‘; ‘ Woke 7 a wow | ait Ribot sa mt) ong, Telesis i) ia + oP ae Bi iisi9r a : ot a. To Shi tee ay sl eet) | sf aa m0" al Lal hol das OL etl AR is ae uy swan, bondi, h bebiiaes aurea a hs aan 201 8 tin) eae estos tamer J} 71) AES): y ayy hi ieee sae ppv vane ‘ sl oe i. ess bes ee. Resumen por el autor, Gary N. Calkins. Universidad de Columbia, Nueva York. Uroleptus mobilis Engelm. II. Renovacién de la vitalidad por la conjugaci6n. El presente trabajo contiene experimentos que prueban que la conjugacion en los protozoarios rejuvenece a estos animales. E] autor ha empleado el método de cultivo aislado ordinario, usando un medio alimenticio tipo y conservando todas las series bajo las mismas condiciones. Un solo ex-conjugante de Uroleptus mobilis, al cual se impide toda conjugacién y endomixis, pasa por 300 + generaciones originadas por divisién, disminuyendo la cantidad de divisiones de un modo continuo hasta que el proto- plasma muere a causa de su edad avanzada, al cabo de siete a nueve meses. Los individuos de tal serie, de parentesco proximo, se conjugan. Tales ex-conjugantes invariablemente producen una cantidad 6ptima de divisiones sin relacién alguna con la vitalidad de la raza del progenitor en el momento de la conju- gacion. El periodo que cubre los primeros 60 dias, en todos los ex-conjugantes presenta una vitalidad uniforme, representada por 17.4 + divisiones en 10 dias. Si la vitalidad de la serie de que procede el progenitor es grande (por ejemplo, 15 divisiones en 10 dias) el aumento de vitalidad en la serie filial es pequefio (2.4). Si la vitalidad de la serie progenitora es baja (por ejem- plo, 0.2 de divisién en 10 dias) el aumento de vitalidad es grande (asi por ejemplo, la serie J de exconjugantes procedentes del individuo A 311 present6é un aumento de vitalidad que se tra- dujo en la existencia de 17.6 divisiones en 10 dias). El autor ha estudiado 14 de estas series, teniendo lugar las conjugaciones que las produjeron en todas las fases de vitalidad de la serie progenitora. En todos los casos observados los resultados fueron los mismos—vitalidad inicial elevada, disminucién de la cantidad de divisiones y finalmente muerte de los individuos. La reorganizacién asexual (partenogénesis o ‘‘endomixis’’) tiene lugar durante el enquistamiento. Los individuos que emergen de los quistes presentan la misma cantidad de divisiones y dura- cién de la vida que los individuos que han experimentado la conjugacién. La partenogénesis por consiguiente, lo mismo que la conjugaci6n, rejuvenece a los individuos que la experimentan. Translation by José F. Nonidez Carnegie Institution of Washington AUTHOR’S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, August ll UROLEPTUS MOBILIS ENGELM. II. RENEWAL OF VITALITY THROUGH CONJUGATION GARY N. CALKINS Columbia University, New York City. ONE CHART AND ONE FIGURE In a previous paper I have described the morphology and the cytology of division and conjugation stages of the rare hypo- trichous ciliate which forms the subject of the present paper.! Lending itself admirably to the cultural method which has been employed, Uroleptus mobilis is the most satisfactory organism for experimental work I have yet encountered. Paedogamous conjugation, in epidemic form, occurs readily, under the proper conditions, in the culture medium. Ex-conjugants, upon iso- lation, live and thrive in this culture medium in practically 100 per cent of cases—a rare phenomenon among the hypotrichs. Asexual reorganization, or parthenogenesis, called by Woodruff and Erdmann ‘endomixis’, occurs at fairly definite periods, un- der the protection of a cyst membrane. Such reorganizations, therefore, are advertised by the form assumed by the organisms and do not interfere with the study of comparative vitality throughout the life cycle. Starting with a single individual which was isolated imme- diately after conjugation, on November 16, 1917, I have followed the life history of thirteen different series, all beginning as ex- conjugants of pairs of normally conjugating individuals, and all were progeny of the original ex-conjugant which was isolated November 16, 1917. Three additional series, derived from encysted individuals, have also been studied in similar isolation cultures. The different series were started at various periods 1 Calkins, 1919. Uroleptus mobilis, Engelm. I. History of the nuclei during Division and Conjugation. Jour. Exp. Zoél., vol. 27, no. 3, p. 293. 121 122 GARY N. CALKINS of the life history of parental series and at different stages of vitality, so that I have abundant data for the study of com- parative vitality of parent and offspring. It is with pleasure, based upon admiration for the genius of that gifted pioneer in this field of research, Edouard Maupas, that I can say these data convincingly prove the truth of his conclusions that conjugation in ciliates restores vitality and prevents the phenomena accom- panying ‘old age’ which we include under the terms senescence and natural death. METHODS AND RECORDS On October 3, 1917, a rich culture in an old hay infusion was found to contain a large number of individuals of Uroleptus mobilis. The normal structure of this rare ciliate is described and its systematic position given in my earlier paper. It is illustrated again in figure 1 of the present paper. Attempts were immediately made to cultivate the organism on the usual hay- infusion culture medium. A better medium was obtained by mixing boiled flour water, hay infusion, and spring water, and this was used until the middle of January, 1918. It was in this medium that the first pairs of conjugating individuals were found and isolated, giving the ex-conjugant destined to form the first, or A series. A still better medium was substituted on Jan- uary 18th. This was obtained by boiling 100 mg. of chopped hay with 130 mg. of flour in 100 cc. of spring water, and diluting this, when twenty-four hours old, with an equal part of fresh spring water. This standardized medium, madefresh each day, has been used exclusively throughout the experiments. Series and lines As used here, the term ‘series’ is applied to an aggregate of individuals, all derived from a single individual and representing its protoplasm. An ideal way to study vitality of such proto- plasm at different age periods would be to follow all of the progeny throughout the life cycle. As this is obviously impossible, I follow as many individual representatives of that protoplasm as time and space will permit. My practice is as follows: When REJUVENESCENCE IN UROLEPTUS MOBILIS 123 an initial individual divides, each of the two cells is isolated as the beginning of a ‘line;’ when these divide, two more lines are started, and one more is added at the next division. Each day a record is made of the number of divisions during the twenty- four hours in each line, and a single individual is isolated with a capillary pipette from each line and transferred to fresh culture medium. The vitality is measured by the division rate, the average number of divisions per day in all five lines for a given period representing the vitality of the series for that period. I have daily records, extending from November 16, 1917, to date, of sixteen series and eighty lines of Uroleptus. Conjugation tests In every line of a series at the time of the daily isolation there are two or more individuals in the culture dish according to the number of divisions that have occurred. After one is transferred the unused individuals are either thrown away or placed in a similar culture dish containing about 1 cc. of the fresh culture medium. Representatives of all five lines of a series are collected in this way and stored in a moist chamber as ‘stock.’ Here they accumulate by division until a large number are present. Once a week this stock is washed into a Syracuse dish contain- ing several cubic centimeters of fresh culture medium and set aside as a conjugation test. ‘The number of individuals increases rapidly until, in three or four days, there are many hundreds or even thousands of organisms. These Syracuse dishes are ex- amined carefully every other day for a period of two weeks and without any fresh medium being added. During these two weeks the limited food is gradually exhausted, and by the end of the period conjugation would have occurred if the internal condi- tions of the organisms were suitable for it. The date of the appearance of conjugations is recorded and the extent of conju- gations, up to epidemic frequency. At the end of the period the individuals are small, inactive, and starved, and, other tests being under way, they are discarded. 124 GARY N. CALKINS Encystment tests Encystment, with its accompanying asexual reorganization, has never occurred in my isolation culture dishes. It does occur, however, at certain periods of the life cycle in the Syracuse dishes during the conjugation tests. Once encysted, the organism can- not be coaxed out until after a longer or shorter period in a dried state. Such encystments are particularly abundant just prior to the first epidemic of conjugation in a series, and, in some tests, practically all of the individuals encyst without conjugation. In such cases the Syracuse dishes are set aside and the culture medium is allowed to evaporate. Such Syracuse dishes with the dried cysts are then stacked away for future experiments. After ‘several weeks or months of storage, fresh culture medium is added to the Syracuse dish containing the dried cysts. In some cases the reorganized individuals emerge by the end of a week; in others, two or three weeks may be required. Series B and M were derived from such encysted individuals. The cytology of reorganization during encystment will form the subject of a later paper. The measure of vitality Reproduction by division is an indication that cell structures are functioning normally, and the rate of division is an index of the condition of metabolic activities of the protoplasm, or a numerical index of the protoplasmic vitality of a series for a given period. The relative activity of a given protoplasm at different periods or of parent and offspring protoplasm for the same period may be obtained by averaging the number of divisions per day in a series for successive periods of the same length of time, e.g., five days, ten days, or sixty days. If we compute the average division rates in this way for successive ten-day periods from start to finish of a series and plot the results, we obtain a curve indicating the relative vitality of the protoplasm of a series at different periods throughout the life cycle. Furthermore, since all series and all lines are fed daily at the same time and with the same standardized culture medium, if we plot a number of curves representing different series which have originated at different REJUVENESCENCE IN UROLEPTUS MOBILIS 125 times for identical calendar periods, we can tell whether fluctu- ations in vitality are due to conditions of the environment or to inherent vitality of the protoplasm. If due to environmental conditions, a correction may be made which will indicate, approx- imately, the comparative vitality, had the environmental con- ditions been normal. Again, if we plot the curves for different series and start them all from the same ordinate, we have a means of measuring the vitality of different series at similar stages of the life cycle. Such curves enable us to find the cyclical incidence of important phases of the life history, such as conjugation and encystment, and to establish the limits within which they occur. These methods have been used in working out the results described in the following pages. In maintaining the cultures and keeping up the records, as well as in working up the statistical data, I have been fortunate in having the assistance from time to time of Miss Mabel L. Hedge and of my colleague Prof. Louise H. Gregory, whose helpful interest I gratefully acknowledge. GENERAL HISTORY OF THE CULTURES The fact should be emphasized at the outset that these ex- periments deal, in the main, with one bit of protoplasm which emerged from the processes of conjugation on November 16, 1917, reproduced abundantly by division, underwent paedoga- mous conjugation repeatedly, underwent encystment, and is still living with a vigor equal to that at the beginning. Some of this protoplasm has been maintained in isolation cultures whereby conjugation has been prevented and in which encystment does not occur. Such protoplasm invariably dies, the phenomenon of natural death being the last stage of a decreasing vitality—a decrease which begins to show early in the life cycle. Other parts of this protoplasm have been allowed to conjugate among themselves, the subsequent isolation ¢ultures showing the effects of such conjugation upon such protoplasm. Still other parts of this protoplasm have been allowed to encyst and to undergo processes of asexual reorganization within such cysts, and the effects of such reorganization have been ascertained. 126 GARY N. CALKINS Also, the fact should be emphasized again, that throughout the entire history of the protoplasm, with the exception of a few weeks at the outset, the same standardized culture medium, made fresh each day, has been used. Waning vitality cannot be attributed to deleterious food conditions, for, on the same day with the same food, one portion of the protoplasm may be at the lowest ebb of vitality, while other portions are in the full swing of metabolic vigor, all portions being equally old in point of time. The difference in vigor between them cannot be due, therefore, to environmental or external conditions, but must be attributed to the internal conditions following conjugation. This protoplasm has been studied in fourteen different series as follows: The A series, or parent race, started as an ex-conju- gant from a pair of ‘wild’ Uroleptus mobilis on November 16, 1917, and died on September 18, 1918, in the 313th generation. The C series, or first filial series (F,), started on February 4, 1918, as an ex-conjugant from a pair of individuals in approximately the 78th generation of the A series. It died out on December 30, 1918, in the 348th generation. The D series, or second filial series, started as an ex-conjugant on March 9, 1918, from the A series in the 137th generation and died out on October 13th in the 271st generation. An E series was also started at the same time from the same source, but was discarded after the 50th generation. The F series, or first F, series, started as an ex- conjugant on March 25, 1918, from the C series in the 86th generation (the grandparent A series was in the 155th genera- tion). This series died December 21, 1918, in the 317th gener- ation. The H series, or third F, series, started as an ex-conju- gant on April 24, 1918, from the A series in the 237th generation, and died out on January 16, 1919, in the 277th generation. The I series, or first F; series, started as an ex-conjugant on July 7, 1918, from the F series in the 148rd generation (the F; grand- parent C series was in the 224th generation and the great-grand- parent A series was in the 278th generation). This series is still living in the 321st generation, but has stopped dividing and will die shortly. The J series, or fourth F; series, started as an ex- conjugant on August 20, 1918, from the A series in the 311th REJUVENESCENCE IN UROLEPTUS MOBILIS 127 generation, and is still living in the 254th generation. The L series, or first F; series, started as an ex-conjugant on November 3, 1918, from the I series in the 199th generation (the grand- parent F series was in the 294th generation, the great-grand- parent C series was in the 347th generation, and the great-great- grandparent A series was dead). This series, also, is active, and in the 196th generation. The M series started from a cyst of the F series which had encysted on April 27th while in the 45th generation, remained dry for more than six months, and was recovered from the cyst on November 18, 1918. The N series _ was started on December 12, 1918, as an ex-conjugant from the J series in the 188th generation. The O series was started as an ex-conjugant on January 10, 1919, from the M series in the 105th generation. The P series was started on January 12, 1919, as an ex-conjugant of the L series in its 115th generation. The Q series was started on January 19, 1919, as an ex-conjugant of the I series in its 316th generation. Finally, the R series was started as an ex-conjugant on January 19, 1919, from the J series in its 245th generation. In addition to these fourteen series representing one proto- plasm, two other series, representing a different initial proto- plasm, have been followed throughout the life history. Of these the B series started from an encysted ‘wild’ Uroleptus mobilis. This individual encysted on November 9, 1917, remained dry from December Ist until January 24th, when it was recovered from the cyst. A filial series, the G series, started as an ex- conjugant from the B series in the 115th generation on March 23rd, and died out on January 5th, 1919, in the 291st generation. In computing the division rate of a series for a given period, e.g., ten days, the number of divisions in all five lines for the ten days are added and the sum divided by five. The value of such an average depends somewhat on the extent of variation in number of divisions in each of the five lines. In table 1 the indi- vidual line records for six consecutive ten-day periods from June 27th to August 25th and for all the series under observation during this sixty-day period are given. Since the different series were started from ex-conjugants at different times, this sixty- 128 GARY N. CALKINS TABLE 1 Actual numbers of divisions in ten-day periods in all lines of series A, C, D, F, H, and I PERIOD AVERAGE 3 = DIVISION come PR | S| S| Selene cee aq rh = 5 is = IN 10 Days Soul bt eh aaa Hee ka Rafe 1 6 5 10 3 2 0 26 2 9 10 12 4 2 1 38 ANGELES ih. sede 3 11 8 12 5 2 0 38 5.86 4 9 9 10 7 4 0 39 5 10 6 13 5 1 0 35 ia ON Oe CS a ef 2 11 12 19 6 10 10 68 © periess . ssp. 3 12 13 17 10 9 9 70 |/10.90 4 8 12 18 11 10 9 68 5 8 10 18 * 9 10 62 12.83 orwWwWN Re —_ ts b— ou — ~] [or] = oO —_ bo ~I ra FF’ geries::.2.. 2. 13.16 or WN Re _ TS _ or iw) oO _— _ _ bo —_ _ io) Ww IHBSOries', <2. 0. 15.26 ewd —_ TX —_— for) bo nee —_ bo —_ wo — lor) c or qn | w _ ~I bo _ —_ —_ Jo) _ ~J Ieseriesee. ss 17 105 20 24 22 13 17 LAs 20 23 20 13 17 17 | 110 17.16 | | i | | | _ bo — ies) i) (an) ~I _ oO : fod iw) ~J ee SS 2’ SS a | ——— —— or WN Fe _ io4) bo wo bo So = bo — or REJUVENESCENCE IN UROLEPTUS MOBILIS 129 day period covers almost every stage of a typical life cycle. It includes a late stage of the A series and the initial stage of the I series, while the others are intermediate. The five different lines of a series thus give consistent records in number of divisions in the same calendar period, so that an average of all five lines gives a fairly accurate idea of the vitality of the protoplasm of a series for that period. Table 2 is a list of such averages for all series in the same consecutive calendar ten-day periods. The averages given in table 2 are based on the actual records of daily divisions in all series, the individuals being isolated daily and fed on the same fresh, standardized culture medium. In some periods, notably in periods 20 and 21, and again in 25, the averages are conspicuously out of proportion with those before and after. These low averages, occurring in all series at exactly the same time, but in different phases of the life cycles, are obviously due to external conditions. The 20th and 21st periods fell on May 28th to June 17th. On the 29th of May the cultures were transferred from New York to Woods Hole, where a dif- ferent natural water was used in making the culture medium, and the temperature also was considerably lower than it had been in New York. This unusual variation in the sequence of averages makes no difference in the life cycle of a given series, but if we wish to compare similar stages in the cycles of all series, some of which include this period of adverse conditions while others do not, it is obvious that a correction of the lower averages is imperative. Such corrections are made for all series in the 20th and 21st periods by averaging the 18th and the 22nd to obtain the 20th, and the 19th and 28rd to get the 21st. In making such corrections the records of the individual lines are used andthe rate for each line is corrected; these corrected line averages are then averaged to obtain the corrected average for the series. The same method is employed for other periods in which cor- rections are necessary. In table 2 such corrected averages are included in brackets,’ and the corrected averages are used in plotting curves of the life cycles and for comparison of different series in similar phases of the cycle. 130 OONIMBHNLWNWe | 10-DAY PERIODS A SERIES 9: 6. 4. 5. as 1s 8. OoOownmr OD © 20.8 18.0 12.6 13.2 14.8 14.6 13.2 15.0 13.6 17.0 12.8 7.4 4.0 (11.0) 5.6 GARY N. CALKINS TABLE 2 Average division rates, all series, in the B SERIES FROM WILD CYST C SERIES FROM 21.0] 18.6 17.0] 18.8 14.2] 16.2 14.2] 16.8 15.6| 17.2 13.4] 15.6 13.0] 14.2 14.4] 16.4 13.8] 15.6 13.8] 17.4 15.0] 18.0 12.2] 13.4 4.0] 5.4 (14.4 5.6] 6.2 Ua ie (10.6 9.0 9.8 o =) OS CaaS ee > RORRWHOAD 3 |e is rl 3s | eo eee & & & & & & a) |f2| 2. | fy |B Q2e| Qo | am n< Qe | Qe a & o q i] Ll 16.0 18.2 16.0)13.4 17.0/13.8 17.6|16.4/17.4 18.4/19.8|19.4 19.0|20.2/21.0 20.0/19.6)19.6|19.6 15.2/15.8)15.6|16.8 7.6) 6.2] 7.0) 7.6 16.6|16.8|15.8)17.8) 7.2| 9.2} 8.0/10.0 14.4/14.0/13.6)15.8) 13.4)12.8/12.0)14.8 14.6)14.2/12.0/16.0 |18.4 | 18.4,19.8)18.8)21.4 |22.6 11.2/11.0} 7.4)10.0 |12.4 12.6/13.8/11.8/16.4 |19.4) 10.4)15.2)15.0)18.0 |11.8 9.4/12.0/12.6/13.6 |15.0 17.2 3.4/11.8)12.2/15.6 |15.8 16.6 1.4/10.8/12.4/15.4 |14.8 |18.6 0.8/12.0)15.2/17.8 |17.6 17.4 0.4)14.6)16.8)17.8 |18.0 |20.4 0.2} 9.8/11.6/11.0 |11.6 {17.2 0.0} 9.6)13.2/15.4 |16.4 |15.8 0.0/12.0) 9.2}/12.6 |17.4 |17.2 8.0) 3.6) 7.4 |14.0 14.6 same ten-day periods L SERIES FROM 1 199 5.8} 2.4) 2.4 |12.8 12.4 16.0) | | 8.4 13.2 CYsT OF F 45 M SERIES FROM = | = 2 = ° ° ° ° ° = a & a = & = = & & gz n n n nD mo) eS | ea | ee | oe am |a amt | go] aN Qy,i ns] @Q2y/ ay] Qe Za ° mo <4 & REJUVENESCENCE IN UROLEPTUS MOBILIS 131 TABLE 2—Concluded tL | Soe ee le = = = =e En = = = = = 8 6/6 (6 18 |¢ |} 1/8 /$ 18 |Se/e 18 |e 18 Js ee feel eee le fe, eee te Le [mel ee ee We ole ia 2 ae] aad | ae] 2o| 2®q| 2 ah |ad|/2H4/25/a,/]/as] any] @u] 2, =) Pe 2 > a a |s Fs = a = Zz ° ee e |& = Baeiee Bees 0.4 3.8 0.2 1.6 13.8 14.416.217.0 38 |. 0.0 2.8, 0.2/1.6 20.8 20.623.224.8 2 ees ord Oe 0.0 0.8 0.4 15.8 17.018.619.817.2) MPD ce. 22 ee eR |....| 1.2 113.0 12.6116.0118.017.8 hp We ee ee 0.0 14.2 13.016.424.623.6 Bee TS. 2. [OT Se Uae AM ee) Ge 9.6 6.617.418.617.815.216.0 a. ESE ERR Gees Bee eee pee 3.0 | 3.618.618.820.4/22.6/19.8)13.2)19.6 LSS Pe ee eee ee ee 0.0 0.811.213.413.8)16.8 14.8) 5.$/13.8 <7 |. /et.d) conte aguas peal ate ei Die 0.2 0.212.612.413.6116.6|15.0| 1.2/14.0 Incidence of conjugation and encystment Conjugation tests in all series have been carried out at weekly intervals beginning with the sixth ten-day period, during which the standardized culture medium was introduced. The records of these tests give the dates on which pairing occurred. With only one exception, every series derived from an ex-conjugant failed to give any evidence of conjugation before the fiftieth day. The one exception was the I series in which a single pair of con- jugating individuals were observed in a test made during the third ten-day period. In the C and D series the first pairing occurred in the fifth ten-day period; in the H and L series the first pairing occurred in the sixth period, and in the J and F series the first con- jugation occurred in the seventhand eighth periods. In allof these cases, except the I series, conjugations were of epidemic fre- quency, i.e., hundreds of pairs were present in the testing dishes. It is apparent, therefore, that the first sixty days, approximately, represent a period of immaturity so far as conjugation is concerned, although, during this period, vitality as indicated by the division rate is at its maximum. After conjugation begins in a series, the succeeding tests are usually positive unless some adverse condition renders the test inconclusive. During the period of immaturity the history of the conjugation tests was the same for all series. Two or three thousand indi- 132 GARY N. CALKINS viduals would accumulate by division in the Syracuse dishes during the first seven to eight days when food was plentiful. With the transition from rich feeding to starvation, conjugation should occur, provided the protoplasm is ready for it. During this period of immaturity, however, no conjugations take place, and at the end of two weeks the individuals are inactive and greatly reduced in size through starvation. Conjugation tests made during the period of maturity give a very different result. The individuals multiply as before during the first week. Then they begin to collect in a single group until a dense, white agglomeration is formed which sometimes measures half an inch in diameter. Such agglomerations are invariably followed by an epidemic of conjugations. Since each conjugation test is started with a few individuals of the same series that are left over after a daily isolation, the conjugations that occur must be between closely related individuals of the same age and with an identical previous history. In diagram 1 the ten-day averages given in table 2 are plotted for the purpose of comparing the life cycles of the first eight series. The ordinates represent the average numbers of divisions of a single line in ten days. The abscissas represent the successive ten-day periods. The positions of the curves in the first period indicate the average division rate of each line during the first ten days of life after conjugation, while the positions of the curves in the tenth represent the average numbers of divisions between the 90th and the 100th days after conjugation, ete. The period during which conjugations occurred to furnish the filial series is shown on the curve of the parental series by a letter which indi- cates the filial series. Thus the points marked C, D, H, and J on the curve of the A series indicate the periods when filial series C, D, H, and J were started as ex-conjugants from pairs in the A series. The successful and negative conjugation tests are shown in diagram 1 by + and — signs, while the inconclusive tests due to lack of material or to adverse conditions under which the tests were made are indicated by question marks. The dotted line connecting the first positive signs in different series indicates that REJUVENESCENCE IN UROLEPTUS MOBILIS 133 Be eee eet eet SESRhnRnnne aan | | eee RSS er Beep -<-e= pom eee LEE TE a GEBENGEBBEEBBEEBE BeBe 1 i ee CHEESE a fer Cee eee F 1o1— ac USSRgt nae nenecceoeaanee ae © ag a ag a a Misia: | hseet— [| | | | | sonspaeeue te ge nen= se He SHaEnaueeetaaet co yee om rca co PEEEEE ELT rr Pe | 33 IST SiG 17S 19420) 21 23 2% 2 2424 23 3 Sl Diacram 1 134 GARY N. CALKINS the protoplasm of each series represented by the curve to the left of this line was in a condition of sexual immaturity. Sexual maturity, or ability to conjugate, persists after the first sixty days practically to the end of the cycle. This is well illustrated by the A series where conjugations occurred regularly up to one month prior to the death of the race. Usually, however, the conjugation tests in the later periods of the cycle are inconclu- sive, owing to the low division rate and the resulting scarcity of material. The cysts of Uroleptus mobilis are small and spherical and the walls are smooth. When once encysted, the organisms do not emerge from their cyst membranes until weeks afterwards, and then only after an intervening period in a dried condition. Such dried cysts, when placed in fresh culture medium, will give up their contained organisms in from five to twenty days. During such encystment, or prior to it, the protoplasm of the individual undergoes asexual reorganization (‘endomixis’) and, after emerg- ing, begins its life cycle with an initial vigor similar to that of an ex-conjugant. The B and M series, for example, started from cysts. Encystments have never occurred in the isolation cultures and the uniformity of the curves is sufficient evidence to show that no other process of parthenogenesis takes place in these cultures. In conjugation tests, however, encystment seems to be due to the same conditions under which conjugation is possible. In such tests the date of the appearance of cysts is always recorded. Such dates are indicated on diagram I by circles above the symbols for the conjugation tests. As shown in the diagram, encystment usually takes place among some individuals of a conjugation test before the first con- jugating individuals appear (series D, F, G, H, and J). In some cases (G and J) encystment occurs in such tests made during the first ten days of the cycle. In other cases (C and I) encystment and conjugation both appeared for the first time in the same test. I have no comment to offer as to the significance of these results, but I hope to get further light on the subject with continued observations. REJUVENESCENCE IN UROLEPTUS MOBILIS 135 RESULTS OF THE EXPERIMENTS TO DATE The physiological effect of conjugation or fertilization in living things has been interpretated, in the main, along two lines of theory. One of these, and the older, may be indicated by Biitschli’s term Verjiingung and by Maupas’s term rajeunisse- ment—terms indicating that the primary effect of conjugation is to restore the lagging vital activities to an optimum. The other theory, first fully elaborated by Weismann, assumes that the union of germ plasms (amphimixis), brought about by con- jugation, is a source of variation. These two theories are not reciprocally exclusive, and it is possible that both are correct, although neither has yet been conclusively established. Biitschli interpreted conjugation in the protozoa as a means whereby waning vitality is restored to full metabolic activity. The problem thus suggested involves three fundamental ques- tions: 1) Does the protoplasm of a single individual protozoon and its progeny by division undergo a progressive waning of vital activities leading to ‘old age’ and finally to natural death? 2) Does conjugation actually restore such weakening protoplasm to a condition of full metabolic activity? 3) If conjugation accomplishes this, what is the explanation of the result? The first of these questions has been answered in the affirmative by the experiments of Maupas and of subsequent investigators. The second has never been answered conclusively, although strong experimental evidence has accumulated in support of the © affirmative. The third question, obviously, is dependent on the second, and at best can be answered only hypothetically on the basis of our present knowledge. 1. Does the protoplasm of a single individual Uroleptus mobilis and wits progeny by division undergo progressive waning of vital- ity and natural death if conjugation and parthenogenesis are prevented? Table 2 and diagram 1 based upon it show clearly enough that this question is answered in the affirmative. Series A, C, D, F, G, and H, all started as ex-conjugants, show the same initial THE JOURNAL OF EXPERIMENTAL ZOJLOGY, VOL. 29, NO. 2 136 GARY N. CALKINS vitality and a progressively waning vigor leading to death, while other series, obtained from these by conjugation, fed at the same times on the same standardized culture medium, are now liv- ing actively. Counting from the day on which the first division of the ex-conjugant occurred to the day on which the last division occurred, the protoplasm of the A series divided during 267 days; that of the C series during 294 days; the D series, 215 days; the F series, 256 days; the G series, 253 days, and the H series, 245 days. The total number of divisions and the average division rate per individual in each series are shown in table 3. The C series had the greatest vitality in regard to endurance, while the F series had the greatest vitality in regard to division TABLE 3 SERIES - a | © | > T FS Number of division days............... 267 294 OB 956 253 045 Numberiotedivasionss.......2+ seer 313 | 348) 271.| 317 | 2905) s2as Average division rate in any ten day DCTION... cou we tie e soe e neine people mee s (Lb. doe) 21,83) 12.34) 12 eee energy. In the latter case the individuals in each line divided on the average 12.38 times in ten days, while in the C series the average was 11.87, and in the H series it fell to 10.93. The dif- ference between F and C is too slight for comment, but that between F and H or between D and H may have some signifi- cance in connection with the problem as to whether the offspring vary in vitality according to the age of the parents at the time of conjugation. I have not enough data at present to throw much light on this problem, but the data from which table 3 was derived may be further analyzed to show how the differences between the different series are distributed in the life cycles. It is shown above that the first conjugations in a series occur, as a rule, from fifty to seventy days after the first division of the ex-conjugant which gives rise to the series. Sixty days, therefore, may be chosen as approximately the period elapsing before the first conjugation in a series, and a period representing the stage REJUVENESCENCE IN UROLEPTUS MOBILIS 137 of immaturity of the protoplasm. Adopting sixty days as a unit period, it is possible to work out from the daily records the division rates in all series for the first, second, third, and fourth sixty days of each series to date. These are shown in table 4 in which the mean division rate of a series per day and its prob- able error have been worked out by Davenport’s formulae as shortened by Crampton. For comparison with the preceding tables, these rates are divided by five and multiplied by ten (multiplied by two) to give the number of divisions which each of the five lines of protoplasm of a series is capable of undergoing in ten days. The G series is omitted from this table, as it rep- resented protoplasm which did not come from the original A series. Three other series, I, J, and L, are introduced, although only one period of the last is involved. The mean for the first sixty days of the A series is not included, since this represents the period of experimentation with the culture media at the outset of the work; the standard culture media was first used with this series ten days before the beginning of the second sixty-day period. The remarkable uniformity of the division rates during the first sixty days in all series regardless of the source, or calendar period, or age of parent, indicates that every ex-conjugant com- posed of a portion of the original protoplasm derived from A begins its life cycle with a definite optimum division energy indicated by 17.1 to 17.9 divisions per line in tendays. Thisis theaverage rate for sixty days, and the rate for the first or second ten day period may be lower or higher than the average for sixty days. Refer- ring to table 2, we find a higher rate than the mean for sixty days in the first ten-day periods of series C, H, and I; while it is lower than the mean for the first sixty days, in the first ten-day periods of series D, F, J, and L. These averages for the first ten days may be even less than the averages for the same calendar periods of the parental series, a fact which furnishes the kind of evidence that has been used by some experimenters as an argument against rejuvenescence by conjugation. Thus in the first ten-day period of the C series, the division rate was 18.6 while that of the parental A series in the same period was 20.8. Again the F series, with CALKINS GARY N. 138 F1/Z FI/ ZI GT/ OT 8L./2/8 04 6C6Z OF} 9F 9S6L OF} °F SG OF TISy Wo1y G1/Z1/€86'S |999F F |ST/OL/SE8 ST/Z916°2 |9T/8 |S86° LT] 9968 ct PI/T PI/ TT P1/6 Te/ 2 81./ L1/¢ 04 180 OF] 0} G6FS OF} 0} G9IZ OF] 0} Z68T OF LEGY Wl 61/9T|/T Ped|ST/ ITs |9919°O |ST/6 |000°FI|0000°L {9T/2 [990° LTISeES'S |z2/¢ |€e"LT] 999°8 H (ST/O|T Pet) GT/ OT 1810 OF |¢ /6 Dae L/¢ 8L./8/€ 0} |002°0 |00SE*O 0} PLES OF} 0F CLES OF] OF 6PEL OF LELTY Woy ¢ /6 | (ATUO |sXep OF)|Z / 2 [99S TT|FE8L°S (8 /G [EEE 9TI999T'8 [8 /€ |99T ATISE8o"s ad b / Zt 61./ 18|/Z1 Ped|F / OF ¢/8 8/9 8/F¥ SL. H/Z 04 €6S0 OF} 0} O9LT OF} 9} Cr6L OF} 2F 608T OF] °F C8I1Z OF SLY Wold F /OLESh' I |99TZ'O |G /8 |996'2 |FE86'S [8/9 \sEs'cI|99TF'9 [6 /F |€E8'ST\L9I6' ZL |2/% |99% LTIEEE9's 8) "peys![qeyse jou uantp -9Ul o1n}[no L1./ S/ 11 81/6 | (81/6) Peta) {22/2 82/¢ 62/ & L2/ T prepueys ques 0} OOST OF} 9 CC8SL OF} 0F GPOL OF] 2} C861 OF] °F ‘poynd -nfuooxe «PIEM, 12/ L |888°T |2999°0 |82/¢ |006'S | OS’ F |62/€ |99E'FI| E8T°2 |8%/T |99E°9T| E8T°S |SZ/TT} -mI0D §=—4gON V SABp O]T|SeUTT OATT sAep (Q]/Seul] OAT sAep (]|Sseul] 9ATT sABp (| SOUITOATT SABP OJ] SOUITOALT ul oul], “ABp ul oul, “ABp ul 9Ul[ ‘ABp ul oul] ‘Aep Ul OUI] *ABp sezeq |sod 9781) red 0481 |sayeq |10d 0381) 10d 9381 |sozeq |19d 9781/ 10d 0381 |sozeq |10d 03¥1| 10d 0481 |soqyeq ji0d o4¥4) Jod o}¥1 uois | UOISIAIp mors UWOISIAIP uoIs | UWOISTAIp woIs UOISIAIp uwOIs | WOISTAIp NIDTHO ONY. GaTEKS “141d uve “Ald ueeyy “41d uBeyl “AIG use “lAIq UB SAV 09 Hilal Siva 09 HLYaAOd SAVa 09 GUIHL SAiVa 09 GNOOaS SXva 09 L8uld sarwas quvbnluos-xa fo fizyozi aaynuwdwoyg % ATAVL 139 REJUVENESCENCE IN UROLEPTUS MOBILIS 12/21 |Ped €/T 0} 9862 ‘0 F / I100F LT/0002'°8 ¢/T F/I ¢ /6 04 9622'0F| °F 02820] 9} 89F% OF F /Ileg0'Stigotg'z |S /6 |996'STIFE86'L |2/2 |99T LTISese's $2/ IT ¥/ 6 92/ 2 LUG 0} 001Z OF] °F 0691 OF] E1&Z' OF] 09 OZE% OF 2/6 |99L°6 |SE88'F [92/2 |999°ZT/EEEs'9 LZ |99G°STIFE8L°L |8z/ |00%° LT} 009°8 81./ 2/ IT 661T WO au SL/ 1/2 erly Wo1g I 81./ S2/€ 980 Wor A 140 GARY N. CALKINS only 13.4 divisions in its first ten-day period, was considerably lower than that of the parent C series, which was 15.6 for the same calendar period. These fluctuations disappear by using the longer, sixty-day period, where in every case the mean di- vision rate of the filial series is greater than that of the parental series for the same period (table 5). The mean division rates for the second sixty-day periods show a decreased vitality in every series (table 4). Here, again, there is a remarkable uniformity of results, although variations are more marked than in the first sixty-day period. The decrease in the vitality of the H series in this second period is so slight that it would not be noticed were it not for the decrease in all series— a decrease which is especially noticeable in the F and J series. In the third sixty-day period there is a similar decrease in vitality over the second sixty-day period in all series, and the variations in different series become still greater. Compared with the second sixty-day period, the decrease in vitality is again slight, but compared with the first sixty-day periods, the decrease in vitality becomes sufficiently large to indicate an indisputable waning of vitality in all series. In the fourth sixty-day periods the vitality shows a marked drop in all series that have passed through this age, and here the variations in different series are extreme. The D series, for example, drops from an average of eleven and a half divisions to seven-tenths of one division in ten days. On the other hand, the F series dropped only from twelve and six-tenths to nine and seven-tenths divisions in ten days. The precipitous fall in vitality manifested during this fourth period of sixty days is continued into the fifth period. Only a few of the series outlive this sixty-day period, indeed, only the protoplasm of the C series was alive at the end of three hundred days, that of the A series dying during the last week of this fifth period. The decline in vigor and early death were most marked in the F and H series, the former dying four weeks, the latter, three days after the end of the fourth period. Maupas found that thelast individuals of a race were deformed, reduced in size, and degenerate through loss of micronuclei and REJUVENESCENCE IN UROLEPTUS MOBILIS 141 cirri. In Paramecium caudatum (’04) I found that the micro- nucleus is degenerated through hypertrophy, while the cortex showed degeneration mainly through the entire absence of trichocysts. Similarly, the last individuals of a series of Uro- leptus mobilis show evidences of morphological as well as physi- ological degeneration. They become greatly reduced in size and the micronuclei entirely disappear, probably by absorption. The macronuclei do not disappear, but show characteristic changes indicating degeneration. The eight separate nuclei usually remain independent, but show evidence of an attempt to fuse as they do prior to division in a normal individual. Their chromatin contents, however, are quite different from the normal; instead of an equal distribution of uniform granules, it masses together to form an intensely staining nuclear body similar to those which form from the degenerating macronuclei subsequent to conjugation (fig. 1, 2, 3, 4; ef. fig. 87, and 89, Calkins, 719). Unlike Maupas’s hypotrichs, Uroleptus does not show cyto- plasmic changes further than reduction in size: the membranelles and cirri do not disappear and the contractile vacuole keeps up its activity. In this degenerate condition the last individuals of a series are unable to divide—division, apparently, being impossible without a micronucleus. They sometimes show a remarkable tenacity of life, however, in this final phase, indicating that metabolic activities may go on despite degeneration. Thus the last individual of the A series lived for thirteen days after the final (813th) division: the last of the D series for eighteen days; the last of the G series for fourteen days; the last of the H series for fifteen days.. The most remarkable cases of longevity of the single individual occurred in the C and the B series. The last individual of the C series lived in its 348th generation for thirty- six days without dividing again, and the last one of the B series, formed as a product of the 257th division on August 31st, lived, without dividing again, until October 9th, a period of forty days. These last individuals, like all others, were transferred daily to fresh standardized culture medium, and their death was not due to violence. Fig. 1 Uroleptus mobilis. Normal individual and individuals degenerated through old age. Camera drawings, same magnification. 1 Normal individual with typical nuclear complex of eight macronuclei and four micronuclei. 2 Individual from the F series in the 315th generation. Four micronuclei still present, but greatly reduced. The macronuclei have undergone granular degeneration. The F series died out in the 317th generation. 3 A prematurely degenerated individual from the C series in the 316th gen- eration. The micronuclei have entirely disappeared. The macronuclei show the characteristic granular degeneration. The C series died out in the 348th generation. 4 Individual from the A series in the 313th generation (a sister cell of the last individual of the series). All of the nuclei have degenerated, the macro- nuclei by granule formation, the two remaining micronuclei by hypertrophy. 142 REJUVENESCENCE IN UROLEPTUS MOBILIS 143 Bearing in mind the fact that, during these periods of declining vigor and death, other filial series derived from their protoplasm through conjugation were living with full vigor, although handled precisely alike and fed at the same times with the same culture medium, we are led to the conviction that waning vitality and natural death are inevitable attributes which are inherent in Uroleptus protoplasm, and that the conclusions of Maupas and of subsequent investigators in regard to this phenomenon are fully confirmed. 2. Does conjugation between two closely related individuals of Uroleptus of the same age result in checking this waning vitality and in restoring the protoplasm to full metabolic activity? The conjugation tests are made with individuals of the same age and of the same series, collected on a certain day from the excess individuals after the isolations are made for that day. These excess individuals, sister cells of those continued in the isolation cultures, represent protoplasm the history of which is identical with that of the protoplasm of the isolation cultures. They are placed in a Syracuse dish with an abundance of culture medium. Here they multiply, and their progeny, in the course of two or three weeks, will conjugate, provided they are sexually mature. One such pair of conjugating individuals is isolated in fresh culture medium. The two individuals will have separated as ex-conjugants in twenty-four to thirty-six hours, and one of these is isolated to start a filial series. Thirteen such filial series have been started; five of these have completed their full cycles and have died out; eight, in various stages of vitality, are still under observation. The question above has already been answered positively by the results shown in table 4. Every ex-conjugant from the A protoplasm, regardless of the phase of vitality in which the con- jugation occurred, shows an average division rate during the first sixty days of life of 17.1 to 17.9 divisions in ten days. Table 4 also shows that this is the highest average division rate of all sixty-day periods of the life cycle. Environmental conditions of food, treatment, etc., being the same for parental and filial 144 GARY N. CALKINS series, the only possible explanation of the restoration of vitality in filial series lies in the process of conjugation by which such filial series are started. Although table 4 shows that all ex-conjugants have practically the same optimum division rate at corresponding periods of the life cycle, it does not show the actual differences in metabolic activity between parent and ‘filial series in identical calendar periods. If conjugation does not restore vitality, the division rate of a filial series should not differ from the division rate of the parent series from which it has been derived; on the other hand, if it does restore vitality, the extent to which it is restored will be indicated by the difference in the average division rates of parental and filial series in identical calendar periods. These differences in average division rates of parent and off- spring series are clearly shown in table 5. Here two interesting phenomena are evident: first, reading across, the table shows that the discrepancy between parent and offspring increases with the age of the parent series at the time of conjugation; second, reading down, it shows that, while both series are losing vitality, the parent series is losing it more rapidly than the filial series. In regard to the first of these phenomena, it will be noted that the C series came from the A series when the latter was in the 78th generation and that, during the first sixty days of the filial series, the protoplasm of each line had the power to divide 1.53 times more in ten days than that of each line of the parent A series In the same period. The D series came from the A series when the latter was in the 137th generation, that is approxi- mately fifty generations older than the protoplasm which gave rise to the C series. The discrepancy now between the parent series A and the filial series D amounted to 3.03 divisions in ten days during the first sixty days of the D series. That is the — protoplasm of each line of the offspring series had the power to divide 3.03 times more in ten days than that of each line of the parent A series. The filial H series came from the same parent A series when the latter was in the 237th generation, or about 150 generations older than when the C series was formed. The discrepancy between parent A and offspring H series now REJUVENESCENCE IN UROLEPTUS MOBILIS 145 amounted to 4.8 divisions in ten days. Finally, the J series came from the same parent A series when the latter was in the 311th generation, i.e., the protoplasm was about 230 generations by division older than it was at the time when the C series came off. The discrepancy in division rates between the young J series and the old parental A series now amounted to 17.6 di- visions in ten days. This means that, if the protoplasm con- tained in the two weakened cells of the parental A series had not conjugated in the 311th generation, it might have divided during the subsequent sixty days at the rate of twenty-five hundredths of one division in ten days, but having conjugated, it actually divided at the rate of 17.9 times in ten days. These results indicate that the protoplasm of Uroleptus under these cultural conditions has a certain optimum capacity or potential of metabolic activity which is gradually exhausted, but which can be restored by conjugation. The greater the ex- haustion, the more remarkable the restoration. As recharging a storage battery restores its potential of active energy, so con- jugation restores the potential of vital energy. If the battery is recharged before the old charge has been drawn upon, its poten- tial of activity would scarcely be affected. An analogous con- dition is shown by the Uroleptus protoplasm after encystment, and after conjugation occurring in an earlier period of the life cycle (e.g., the A and C series, or the C and F series). If, how- ever, the battery is nearly exhausted before it is recharged, the difference in potential between the newly changed condition and the exhausted condition would be marked. The analogue to this is the J series and the parental A series. The restored potential of vitality in Uroleptus protoplasm is a ‘charge’ of vital energy which is capable of metabolic activities through a period of from 260 to 300 days, or vitality sufficient to produce individuals to the number of 2 to the 300th+ power. In regard to the second of the phenomena shown by table 5, it is interesting to see that, as the protoplasm of a series grows older, the potential of vitality is exhausted at an increasingly rapid rate. Thus, the discrepancy between the division rates of parental and filial series is greater during the second sixty-day CALKINS GARY N. 146 ‘sA@p OT Yove ut Sel1es [BI[G jo 9889 'F PEET'S 9996’ £1 eeLL's eee9'Z eur] 1d asve1ouy SABP QT 10J oul, tod o04e1 GEESE TT|8996 STI PEEE ET/899S ST OOOTE |9990" LT|009S OT/EESE'9T|000%ET|SSE8' ST) WOIStATp oBev19A Vy aeeP = See Dade Soul] AG 18 0981 ors 9999'S |FE86°L |2999'9 |FE8L°L | “PBEP |JUEIe |DOGE'T |ESES'8 |€E8Z'S |999T'S |0009'9 |29T6"Z | -IAIP ATrep Uva sABp OT ove UI SoI1Vs [BI[G Jo PEEO PF 0000 °T GE89 LT 0008 €&E0'€ GEES T aur] od asvorouy sAep QT 10F eur] sod 9981 ‘Suridsyo CEET ET/999T “L1/000Z 91/000" 2T|00S%'0 |ZEk6' LT/ZEES* ZI/ZEEE LT/EEET PI|9VOT LTPEEL “ST/999Z LT] UOISTATP OSBIOAY | yo sep Sa |e ee re Ge al [8 9381 wors 9999°9 |€E8S°8 |OO0T’S [0009'S | S20 |9996'8 |999Z°9 |9999°8 |9990°2 |sEse"s |2998°L |EEE9's | -IAIP ATrep uvay G/6 09 L/L 12/S 4 82/ S1/01 93 91/8 1Z/L 9422/8 L/¢ 04 8/8 8/F 04 L/Z 038d SUOI}VIOUOT EFT SUOI}BIBUIT 9g SUOT}PVIBUET TIE SUOIPBlOUEs 197 SUOIZB1OUIT / ET suoIyeioued gy jt INdUVYd JO ADV saluas SaIUaS|Saluas SA1MaS|Saluas Sa1uas| Saluas SaIUas|Ssaluas Sa1Uas|Saruas saluas a woud I o woud a vy wou ¢ vy wWour H vy wour a vy woud 09 buridsffo pun squasod fo fizypj10 saynsndwuoy $ WIGVL 147 REJUVENESCENCE IN UROLEPTUS MOBILIS €&Z8" OT VEE9'P 8860°2 CSOT F OO9T FP |SEE0 SLZEEO'S |9999 ZT O897 PF |899S TT/O0EL'S8 |ZEE8 ST 0080°% |99TS°2 |99T0'F \€eEE'9 PBop |FUSIed | FES'S |FEBL'S |OS9E'F |9OTF'9 €/T 94 6/11 42/6 94 92/1 8I/TT 93 61/6 ¥/6 99 9/2 9/8 94 1/9 ‘sABp OT YoRe ur SOII9S [BI[G Jo eul| od esvol0uy- SABP QT 105 aul, tod 94¥.1 UOISIAIP O3BIOAY Soul, SAG I[@ 381 wos “TAIp ATrep UBoTN sojeq ‘sXep OT Yovo Ut SoI19S [BI[y Jo aul] rod esva190UT sep OT 104 oul, iod o4vi UOISIAIP 93BIDAY Soul] [[@ 9} RI UOISIAIP ued] soyeq sep 09 YANO sep 09 PAIQL 148 GARY N. CALKINS period of the filial series than during the first sixty days. Indeed the discrepancy between parent and offspring in successive sixty- day periods increases in geometrical proportion. Thus the dif- ferences between the A and C series for the first, second, and third sixty-day periods are indicated by 1.53, 2.63, and 4.10 divisions in ten days. The differences in the A—D series are 3.03, 5.77, and 7.09. In the C-—F series, the differences for four successive sixty-day periods are 1.0, 2.23, 4.63, and 7.26, which is a geomet- rical increase of the disparity in division energy of parent and offspring protoplasm. When we consider that we are dealing here with one protoplasm derived from the protoplasm of the single individual ex-conjugant that was isolated on November 16, 1917, these results offer con- clusive evidence that conjugation rejuvenates or restores vi- tality to an optimum when that vitality is reduced through continued metabolic activity. Many series have died, but I have under cultivation to-day protoplasm of the L, N, P, O, and R series which is directly descended from the original ex-conjugant A and which is living with the same metabolic vigor as that shown by the A series during its most vigorous period. Yet there has been no change in the standardized culture medium with which this protoplasm has been fed, and no variation in the daily treatment. The continued vitality is due solely to the successive conjugations which have taken place between representative bits of this protoplasm. Since one condition, viz., starvation, is found in the conju- gation tests and not in the isolation cultures, the objection might be raised that changes may be set up in the protoplasm due to such starvation or to some other condition of the conjugation tests which would result in a restoration of vitality, thus making conjugation an accessory phenomenon without effect on rejuve- nescence. To test this point twenty individuals which had reached this starvation point as shown by reduced size, and all taken from a conjugation test, but without having conjugated, were isolated and carried on in isolation cultures as though they were ex-conjugants. An ex-conjugant from the same test, and obtained from a pair that were isolated while conjugating, was likewise carried on at the same time in five lines as series U. REJUVENESCENCE IN UROLEPTUS MOBILIS 149 The results (Table 6) show that conditions of the conjugation test hage no stimulating effect on the protoplasmic activities. Indeed there is evidence of a depressing effect as indicated by the average division rates for sixty days. The rates for non-conju- gants are not only lower than that for the ex-conjugant (U Series) but, in all cases, are lower than that of the control parental series. All were descendants of the same series (L series) and all were of the same age. The starved individuals are grouped in four aggregates of five each in table 6, as follows: TABLE 6 NON-CONJUGANTS FROM SAME parent | 2X-CONJU-| cConsJUGATION TEST AS U SERIES. RACE GANT FROM TWENTY INDIVIDUALS IN 4 a iniios ||) 2 Seeses GROUPS OF 5 EACH (6u1nns) | U.SPRIES (5 LINES) Group 1|Group 2|Group 3/Group 4 Average division-rate first 10 cc ESS a eg a ee ead A Be 8.4 9.2 0) | 76, ne oes C29 1 io Eo rhe a oe 5.4 14.0 5.8 | 6.6] 6.6] 6.8 Mind AOWaYS. /. 40260. 2 bao. 10.0 13.4 #4 | 5.6 | 6.8 1524 oumta 0) days... S64... fis gee ete 10.4 14.0 7.0 6.6 8.4 5.4 UO Es 5 en 7.6 13.6 Grout 6:8) | 47 2ulemaee PU OVGA YS sow ch oa oe cs ko eis bes 5.6 13.6 US | TE 7 A Average division-rate 60 days..... 7.9 12.96 6:4|) 6:8 | -7. 1, |) Gus There is some evidence, by no means complete as yet, that vitality of the more recent series lacks the endurance of the earlier series. This is apparent in diagram 1, where the curves of the filial series are progressively shorter from the C series to the J series. Both I and J are now in the last stages of metabolic vigor and no further divisions in either series will occur. The I series is now 235 days old, and the J series 200 days, the former having divided 322 times, the latter 253 times. The shortened cycle appears to be accompanied by greater intensity of division energy than in the earlier filial series, the I series having an average division rate of 13.2 divisions in ten days, the J series, 12.6 (table 3). In a few months the cycles of the L, N, O, P, and R series will be complete and will furnish more adequate data for conclusions on this interesting point. 150 GARY N. CALKINS The Q series, coming from the I series in the 316th generation, has been queer from the start, dividing only twenty-three times: in thirty-three days, and it will soon die out. The exceptional history of this series is probably due to faulty reorganization after conjugation, for the nuclear complex is quite abnormal. If this result is due to the age of the parent series at the time of conjugation, the history of the Q series is an interesting contra- diction to that of the J series where the parent series was rela- tively even older at the time of conjugation. Defective reor- ganization is possible after any conjugation, but the chances of such defective reorganization are probably greater with increasing age of the parent. While the several series described above were all derived from one ancestral protoplasm of the A series, two other series, B and G, came from a different source. The B series was started from an encysted Uroleptus mobilis which had encysted in ‘wild’ stock before the A series was started. It emerged from the cyst. on January 25, 1918, and lived until October 9th, dividing 258 times between January 25th and September Ist, or 11.8 di- visions in ten days on the average. Many epidemics of conju- gation occurred, but only one ex-cenjugant was isolated. This. one formed the G series, the B series being in the 115th generation at the time. The same resultant renewal of vitality and con-- tinued life of the filial series was observed as with the A proto- plasm, the G series starting with an optimum division rate of 18.06 in ten days for the first sixty days, running through 291 generations and dying out January 4, 1919. 3. Does reorganization during encystment (‘endomixis’ or partheno-- genesis) restore waning vitality to full metabolic vigor? Unfortunately, I have insufficient data to draw positive con-. clusions on this subject. Only two series, B and M, were de-. rived from cysts, and of these only the M series is pedigreed. This series was started on November 18th from a cyst that was. formed by an individual of the F series in its 45th generation, on April 27, 1918, and remained encysted for six months. In. REJUVENESCENCE IN UROLEPTUS MOBILIS Et the first period of sixty days after emerging from the cyst, it had a higher division rate than any other representative of the A protoplasm, the protoplasm of each line dividing, on the aver- age, 19.8 times in ten days. If the vitality of the encysted in- dividual was the same in potential as that of the same proto- plasm in the isolation cultures, we would expect the division rate of the isolation series during the sixty days subsequent to the 45th generation to be practically the same as that of the proto- plasm from the cyst. This expectation, however, was not real- ized, for during this period the F series divided only 17.6 times in ten days—a difference in potential of 2.2 divisions in ten days. The difference between the division rate of the F series for its first sixty days and the M series for its first sixty days was 2.6 divisions in ten days (table 4). The difference between the division rates of the parent F series and the offspring M series by parthenogenesis is not actu- ally as large as the figures indicate. The M series was main- tained under the conditions of a constant temperature of 24°C., while the F series was maintained under the conditions of lab- oratory temperature which varied from 19° to 22°C. That the difference in rate is not due solely to these different conditions, however, is shown by a comparison of the M series with the L series (cf. Table 2). The latter came from the I series and the I series from the F series. Both L and M, therefore, had the same ancestry. Furthermore, L and M were started at approxi- mately the same time, L as an ex-conjugant, M from a cyst, and both series were maintained under the same temperature condi- tions. The L series, during the same sixty days as above for the M series, had an average division rate of 18.8 in ten days as against 19.8 for the M series. The unpedigreed B series, which came from a cyst, may be compared with the C series which started as an ex-conjugant at about the same time as the B series. The first sixty days of the B series gave an average division rate of 17.4 divisions in ten days, while that of the C series was 17.2. So far as the evidence thus far obtained is concerned, it ap- pears that the initial vitality after encystment and partheno- THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 29, NO. 2 152 GARY N. CALKINS genesis is as great as, or even greater than, that after conjuga- tion. It remains to be seen whether this high potential has the same capacity of endurance as that obtained from conjugation. The M series is not old enough at the present time to furnish evidence. The B series divided only 258 times, while its con- temporaries, the A and C series, divided 313 and 348 times, respectively. GENERAL It is not my intention to formulate here any theory in explana- tion of the phenomenon of rejuvenescence. The facts con- cerning it may be grouped in two categories: one, physiological, the other, morphological. In the first place, the results presented in this paper show that in Uroleptus mobilis the physiological processes of metabolism are not capable of unlimited activity. The limits vary from the time of conjugation or encystment to between 268 (H series) and 349 (C series) generations by division. Within these limits there is a progressive weakening of metabolic vigor from an opti- mum shown during the first three months after conjugation. This weakening, furthermore, increases by geometrical progres- sion, 1.e., 1t is cumulative, as shown by the geometrical increase of the difference in vitality between filial and parental series in successive sixty-day periods (table 5). Such weakening proto- plasm, if not allowed to conjugate, inevitably dies, as does the somatic protoplasm of metazoa. A second physiological fact is equally well established by these Uroleptus experiments. The same protoplasm is transformed from a condition of metabolic weakness to a condition of opti- mum metabolic vigor by the. process of conjugation. The effect of conjugation is clearly indicated by the extreme case of the J series. Here the protoplasm, at the time of conjugation, had only enough metabolic vigor to divide twice in ten days. The cells that conjugated were both composed of this weak proto- plasm. Ten days later if they had not conjugated, each might have been able to divide at the rate of 0.25 times in ten days, or once in forty days; but, having conjugated, one of them, the REJUVENESCENCE IN UROLEPTUS MOBILIS 153 J series, was able to divide at the rate of 17.9 times in ten days, or 71.6 times in forty days, and there is no reason to doubt that the other ex-conjugant of this pair would have had the same vigor. The experiments also indicate that there is a limit to the ex- tent to which this protoplasm can be rejuvenated. It might be inferred, with reason, that if two weak individuals are trans- formed by conjugation into individuals capable of dividing sev- enteen times in ten days, then conjugation between two indi- viduals from a series in which physiological weakness is not yet perceptible, would result in an ex-conjugant capable of dividing more than seventeen times in ten days. The inference, however, is not supported by the facts. A good illustration is the rela- tion between the F and C series (cf. Table 5). The F series came from C when the latter’s vitality was indicated by 17.2 divisions in ten days. Each of the two individuals conjugating at this time, would have had the ability to divide at the rate of 16.2 times in ten days during the ensuing sixty days, if they had not conju- gated. Having conjugated, they were able to divide at the rate of 17.2 times—a difference, or extent of rejuvenescence, indicated by only one division in ten days. Since all ex-con- jugants, under the conditions of the experiments and regardless of the state of vitality of the parent protoplasm, return to this same optimum of vitality measured by 17+ divisions in ten days (table 4), it is evident that the protoplasm of Uroletpus will hold only a certain charge, so to speak, or potential of metabolic vigor, as a result of conjugation. This optimum, of course, is subject to change by changes in the environmental conditions—heat, for example, increasing it. A third physiological fact is also indicated by these experi- ments, although the numerical support is not as adequate as that in support of rejuvenescence through conjugation. This is the fact of reyjuvenescence following encystment and parthenogenesis, in which no nuclear interchange occurs. The B series and the M series came from cysts. The ancestry of the former is un- known, but during its first sixty days in culture, its metabolic vigor was measured by a division rate of 17.4 times in ten days. 154 GARY N. CALKINS This is the same optimum as that following conjugation. The M series is pedigreed. It came from the F series in its 45th generation, and after a period of six and a half months was re- covered fron the cyst. During the sixty days subsequent to the date of the 45th generation, the F series divided at the rate of 17.6 times in ten days, while the protoplasm of the M series during its first sixty days of culture divided at the rate of 19.8 times in ten days. The difference (2.2 divisions) is undoubtedly greater than it would have been had the temperature conditions remained the same. As explained on page 151, the M series was cultivated under conditions of higher temperature than the laboratory, whereas the F series was cultivated under the lab- oratory temperature. The L series, which came from the F series through conjugations and which was started at approxi- mately the same time as the M series, was cultivated under the same conditions as the M series. Its division rate during this same period of sixty days was 18.8—one full division higher than the usual ex-conjugant. In this case of the M series, therefore, the rejuvenating effect of parthenogenesis was even greater than that of conjugation. Whether the endurance of the partheno- genetic protoplasm differs from that of the protoplasm following conjugation remains to be seen. Parthenogenesis through encystment appears to be an attribute of high vitality, and the ability to encyst is apparently lost at an early date (diagram 1). In the C series it did not occur after the 160th day; in the F series, not after the 110th day, and in the D, I, and J series it did not occur after the 80th, 60th and 20th days, respectively. I do not know what this means, but it is certainly true that no internal reorganization without en- cystment has occurred thus far, for in every series the physio- logical depression is continuous and progressive, and death in- variably follows. Conjugation, with rejuvenescence, however, is possible almost to the end of the cycle. Enceystment, apparently, is not possible near the end of the cycle, but it does occur even in the first ten days after conjugation. Conjugation, on the other hand, does not occur until from thirty to seventy days after the previous conjugation. I am aware of published state- REJUVENESCENCE IN UROLEPTUS MOBILIS 155 ments to the contrary in connection with other ciliates, and it may well be that this condition of sexual immaturity noticeable in Uroleptus is not universal among ciliates. As Maupas found for other hypotrichous ciliates, the condi- tion of physiological depression is accompanied by morpho- logical changes. During the period of active metabolism the cell rapidly grows to full size after division (1404 to 165 yu). The macronuclei are eight in number, with from four to six micro- nuclei. In preparation for division a portion of each macro- nucleus is thrown off and is absorbed in the cytoplasm, while the remaining portions fuse to form a single division nucleus. All but two of the micronuclei are likewise absorbed. In the late individuals of a cycle, the macronuclei lose their characteristic nuclear clefts and in some cases show a tendency to fuse, while in other cases the number is increased from eight to as many as sixteen smaller and irregularly shaped nuclei. The micronuclei do not increase in number, but undergo degeneration by hyper- trophy or by granular degeneration, and finally disappear in the cytoplasm. In the last individuals of a series, the chromatin of each macronucleus collects in a single large, highly refractile, and densely staining granule. The size of the cell is greatly reduced and it is unable to divide. The cytoplasm is probably as much changed as the nuclei, but morphological evidence of such change is difficult to detect. In general there is a tendency to increased vacuolization, while the mitochondria, which form a cortical layer in the normal individual, are rare and irregularly distributed. Some individuals in this final stage of depression live without dividing for thirty-six (C series) and forty (B series) days. In conjugation, apart from the processes of maturation and reduction in number of chromosomes from eight to four, and union of the gametic nuclei, the most significant phenomenon is the granular disintegration of the old macronuclei, and absorp- tion of the relatively large quantity of nuclear substance in the cytoplasm. Not only do the macronuclei thus furnish nucleo- proteins to the cytoplasm, but the micronuclei also contribute no small part. Thus, if an individual goes into conjugation with 156 GARY N. CALKINS four micronuclei, all four of them may undergo the first matura- tion division. Of the eight possible micronuclei thus formed, only two undergo the second maturation division, while six are absorbed. All of the four products of the second maturation division may undergo the third division, and of theeight products of this phase, only two become pronuclei, the other six being absorbed. Finally, in the second division of the amphinucleus, two of the four products form the new micronuclei, one forms the new macronucleus, while one, the sister nucleus of the new macronucleus, is absorbed in the cytoplasm (Calkins, loc. cit., pp. 316-326). The cytological details of encystment are not yet worked out. Anticipating the description of the process, it may be briefly stated here that the macronuclei break up into granules as they do after conjugation, and these granules are absorbed in the cytoplasm. One phenomenon, therefore, common to division, conjugation, and encystment, is the absorption of variable quantities of nuclear substance in the cytoplasm. That the physical and chemical consequences of such absorption are connected with the phenomenon of rejuvenescence seems probable. That the rela- tion between the new amphinucleus after conjugation, or the new nuclear complex after encystment, and this reorganized cytoplasm is likewise connected with the phenomenon of rejuve- neseence is equally probable. The nature of such connections and of such relations is a matter of speculation for which we are not yet prepared. Columbia University March 4, 1919 : i aa cot A 7 ies cas roe meh Oar . jar pH. 3a) ty ie yin? rece) Dita 4 a =< nee ae ia APES v4 4) . raptinht : seee tts £ OPS. Uy ha c o 3 ‘ Bd ccaiyeirt | ” meg . } ey Ob eyes us yes ; : { i i » i i : ; x ‘| Lt : "- - j ar H tla i, RCE TOPE ee ys ee Weg ra 7 + ) ' ’ ‘ Ue ‘ ; bene’ t's Li tip LNG aston ety Hiinn: neil wh of I closinleaee) 1 Wc cus L a I Pa eee OM 7 ae : ae stoke oe VCR SOP SAD Haber ie Va a Lyng a i oh D i Berit ehrices ecg ) a oy ptt ar be “whet Leis! Spry iri t j ( 4 pe Pa) ‘ Pa y's ; ’ nN brea AR ALT MOP Wop) A, OEE PUMA MAH Ta th eames oe ' } A : amet AT : i an Pele WE CY Rat he eapearope ic [okt RURNeL el red See vena ah vi gine es ve en . , ae stk isi Tina ee Muted barns Mitt. 7 yal har by salah ; . ieee sf tiv a i3 so if ore i Re niptreen «hbiasilncing lad Lago 7 (ieee (whoa . see? tie G14 i) ch hie ifelavo .s Shh, meteaael) p antpionles, parity tin Agasiiere: fob ; 7 dai al eee set Handi, wrailatt ‘ro ; f ; ; i i . fe shit: a | ots fv aa ne op) , bess ; Doh. od) eet Fash 1 PARE asd ok Hesumido por los autores, Leslie B. Arey y W. J. Crozier. Escuela Médica de la Universidad del Noroeste y Estaci6n Biologica de Bermuda. Las respuestas sensoriales de Chiton. En el presente trabajo los autores pasan en revista las ca- pacidades sensoriales y modos de reaccién que presenta Chiton tuberculatus Linn., con referencia particular a la diferenciacién de los receptores y a la significacién etiolégica de ciertos tipos de respuestas. Bajo el primer epigrafe sefialan que se puede demostrar un grado considerable de diferenciacién sensorial, y bajo el segundo prestan atencién a la conexién entre las modifi- caciones progresivas del heliotropismo en relacién con la edad, el modo en que se determinan estos cambios y sus consecuencias, que aparecen en la vida de una poblacién de Chiton. La bionédmica de Chiton descubre una serie intrincada de inter- relaciones armOnicas, necesitandose una descripcién exacta de su historia natural. Los autores sefalan el hecho de que el medio ambiente en que vive este molusco ‘‘primitivo,” en el cual la centralizacién nerviosa se presenta en un estado incipi- ente, esta determinado en varias edades por el comportamiento del animal, con ciertas relaciones ventajosas realizadas de una manera automatica. Translation by José F. Nonidez Carnegie Institution of Washington AUTHORS’ ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, august 11 THE SENSORY RESPONSES OF CHITON! LESLIE B. AREY Northwestern University Medical School AND W. J. CROZIER Bermuda Biological Station FOURTEEN FIGURES CONTENTS I. Introduction. . se GSS Il. Natural hielo : : ... 160 1. Habitat sal. Hopeninnce | ie. . 160 Dh Ferrer eaieiclict etic bait In ieanen Mattia tae Ue dade. 4c, eam a ae 169 See DestrUchivierde Cnits: 1.1. vpaysoter ie Meier oe eer ARM ro on yaar cy Ao TS eel 4. Feeding.. prio 74 5. Respiration <3": =e Helis 6. Migrations; neeomieuian Ae PLOUPS! 22 ee ole Raa etn aon es mL The Ba tcdine habits eM eC ew rk Pe eh ae 179 SeBIONOMICECOnne AnlONnSs-. rs wee ace ano cee waco ELST III. Movements and reactions. . ey 1. Local movements. . ahe%: , . 183 2 Bilan CBr eviatlib acsiiaillsacya ual ke, seyret cn eek ae ane 185 SE AGO CONT O1 O My oy yas atch gear SOR ence, oy ecg Nips cine NS as bow na Ra eg OES Ve Vechanical OxXCluauLOMc ns 8. ceria sae ae eee cian ce OO oo OD 1S Parewle-stimiulatlomee? vcd bir taser tee. Cenk. Chiceec cin os Sea eae OD A. Dorsal surface. . ao ilies} B. Ventral parts.. rine: as 194 (Gp Disuibution! dMscnsibinatee SR Aste RES a eet ate LOS iD: rTECUTIGHre Ge GTAe lms RUM MeILe Sy A Sy olin Minas Ub 200 2s Vibtatory. Stine’ 2.7 Pty. SP RR PEO 2 ee 2 SB SV emOTARIS? Soe. Le Rae IL, Joan, ae ee OE AMSRVEOtLOPISIOY 6... fcc 2. Pee Doe eee chess area de ais TR 0G Or Greotroprsni:)f:. Ye. SHE ERS folk GR Py eae, AIR oe OM G2 Summarys Ps PS: Set ee ae ie be ree ss Ea oO 1 Contributions from the Bermuda Biological Station for Research, no. 110, and from the Anatomical Laboratory of the Northwestern University Medical School, no. 66. 157 158 LESLIE B. ARBY AND W. J. CROZIER V. Thermal excitation. . ser acoiua sUaie Nee layouts Babe € A ORLA ARSE A ele eae 1. Behavior at diferent Peperatuneee, sari ae 6.0 Skeas cena ee 2. Local AepleRpOmeMineatnnd cald 270)... 6 221 3. Suing pi ec eee vis os a 5 08 2 G2 ip ciate n Sth.e ea ety eae VIL -Photieexcitiwiioiis weiss 6c bess oo oe eee tlvinlae berotsve a aale oor oo 1. Effects inthlighin: moe ELLE Ye SSE. SEE eee a. Behavior in an Gia eunnied fet «Ae ade ciebin Oa eee b. Results of partial illumination of the body.:.....a:.» . These matters are touched upon at this point for a reason which will become apparent further on in this paper. They are discussed with some completeness in a subsequent paper, by one of the present writers, dealing with the ethology of chiton. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 29, NO. 2 166 LESLIE B. AREY AND W. J. CROZIER kinds and number of these epizoites depend upon the size of the Chiton and the character of its habitat. The most conspicuous of them are barnacles (Tetraclita, fig. 1), Spirorbis, and Serpula. To this group must be added algae, comprising not merely the thin coating upon the valves, but also the Enteromorphas, which (in appropriate habitats) grow plentifully upon and between the SoG K. 128.43 res — a ~ 6.2cms, 129.44 8.8 cms. i 4.loms xl I. 123,/ XZ.33.38 c ) T7cMms. “/ coms. Fig. 3 Outlines of the fourth valves of five chitons of increasing ages (sizes) ; dorsal aspects, anterior up; see text. X %. Fig. 4 Illustrating protective growth of the valve substance resulting from the presence of epizoic barnacles; the outline is that of the valve. X 1. scales of the girdle. They are very important for the production of homochromic effects, because the periostracum of the scales is but little eroded, even in large chitons. The valves are rarely much overgrown with these algae, except among the largest chitons. In the felted matting of algae various young mollusks, nematodes, archiannelida, and protozoa abound. The serpulas THE SENSORY RESPONSES OF CHITON 167 affect only the very largest chitons. Barnacles remain attached to a valve in some cases until they have formed three growth lines (‘year lines’). One effect of the barnacles is important in connection with our preceding remarks regarding the forward growth of the tegmentum as correlated with the erosion of the superimposed umbo. Instances such as that illustrated in figure 4 show how it is possible for the shell to grow in a protecting manner. In studying shell variation in the chitons it must be remembered that the attached barnacles may be removed, before or after their death, and leave no obvious trace, although they may have been responsible for irregular growth of a valve. At sexual maturity the female Chiton tuberculatus is colored in a different way from the male: its tissues are impregnated with a salmon-pink substance concerned in the metabolism of the ovary. If the shell plates are separated, this differential color- ation of the sexes may be detected in dorsal view. Normally, it is quite invisible. This is the first instance of its kind which seems to have been described among mollusks. Its importance has been discussed in another place (Crozier, 719). In the gill channels and under the girdle of chitons obtained on sunlit shores where Enteromorpha and associated plants are growing in a felted covering over the rock, there are nearly always to be found considerable numbers of a commensal isopod. It appears to be the Eusphaeroma (Sphaeroma) crenulatum of Richardson (’02, p. 292; ’05), described by her from specimens collected at Bermuda many years before by Goode, but concern- ing which no information as to habitat or local manner of occur- rence has previously been recorded. The association of this isopod with C. tuberculatus is general throughout the Bermuda area, but the commensalism is of a more or less facultative kind, since the isopod is found sometimes among the algae at some distance from a chiton. Even where the supply of algae is scanty (as in crevices within the walls of caves), the isopods are also sometimes found, but usually not in such abundance, under the girdle of Chiton. As many as twenty or more are to be found under a chiton 8 cm. long. The association is quite independent of the sex and sexual coloration of the chiton. The isopods are 168 LESLIE B. AREY AND W. J. CROZIER small (2 mm. long), and when taken from a chiton at low tide their coloration is quite pale, of a yellowish cast, with minute black markings. The coloration becomes darker in the light, and then reproduces on a small scale something of the greenish-to- black color pattern of the chiton girdle. The sphaeromas fre- quently remain in place under the chiton at high tide, and in a glass aquarium they will reassume a position within the gill channel or under the girdle of medium-sized or large chitons. There they take up stations chiefly along the lateral margins of the girdle, which are slightly raised during the respiration of the chiton. The isopods remain with heads pointed outward, into the incoming respiratory current. Sudden shading causes them to dart back into the ctenidial channel. When in their resting position a small portion of the anterior end may project beyond the edge of the girdle of their host, and under these cir- cumstances their coloration, resembling that of the dorsal tu- bercles upon the girdle, renders them very inconspicuous. Occa- sionally one of these isopods creeps in between the gill filaments, usually at the posterior end of a chiton, and under these condi- tions is forcibly shot out at the posterior extremity of its host by means of the water current; the isopods appear, moreover, to react negatively to currents of this strength, and continue vio- lently to swim away, in a spiral path, from the region of the anal current even after it has ceased to act upon them in a gross mechanical way. At high tide, and when under water in aquaria, the isopods creep freely over the dorsal surface of the chitons; under these circumstances their coloration is to a very high degree homochromic and concealing. This relation between Sphaeroma and Chiton will be made the subject of further study. At present it can be said that there does not appear to be any precise ‘attraction’ (chemical, for example) exerted by the chitons upon the isopods. This commensal isopod is involved in the very complex envir- onmental correlations, which may be clearly analyzed, in the life-history of Chitons. Therefore we mention it here, although detailed work on its behavior and relations must be deferred for the present. THE SENSORY RESPONSES OF CHITON 169 2. Growth and duration of life As with animals in general, the rate of growth of Chiton de- creases with advancing age. For the area considered in this report the growth of Chiton tuberculatus appears to be ade- quately represented in figure 5. A detailed analysis of the material upon which this curve is founded will be given elsewhere. x ce ee oh OS a ee A ge(estimated ). | ad a ¢ 8. oO trys. 3 4 5 Len gth, Fig. 5 Showing the relation between size (total length, in centimeters) and estimated age (years) in Chiton tuberculatus. These est mates are not in- tended to be of final significance. The normal average growth curve differs in shape from that shown. The two lines shown include between them most of the variations found in the Chiton population of Great Sound (April to May, 1918). The maximal duration of life seems normally to be from eight to nine years (fig. 6). A length of existence so great as this appears not to have been suspected previously for the chitons. This species probably comes to reproductive maturity in the second (or third?) year of life (Crozier, 19). The general rate of growth 170 LESLIE B. AREY AND W. J. CROZIER corresponds with that of some other chitons (Heath, ’99, ’05b ’05 c), at least for the early years. It is also true that in certain other chitons the second year marks the incidence of sexual maturity (Heath, ’05¢). The large Cryptochiton stelleri grows more rapidly than Chiton tuberculatus does, but also matures in the second year (Heath, ’05 c). The curve in figure 5 is introduced here for the purpose of correlating some statements to be made subsequently regarding behavior at different ages (cf. Crozier, ’18b). Freguency o8O BG Y i 2°53 4°56. Sie Age, estimated. Fig. 6 The frequency distribution of estimated ages in the chiton popula- tion on the north shore of Long Island, Great Sound (April, 1918). It is neces- sary to consider each local population separately. The one here plotted is fairly representative. 3. Destructive agents The power with which chitons may adhere to the rock surface is well known. When disturbed the girdle is firmly applied to the surface and the shell plates closely approximated. Except by means of a very powerful lateral push, it is impossible to dis- lodge them, once they are ‘set’ (which may happen very quickly); — there is no projecting part of the smooth depressed animal which could offer a ‘hold.’ The dead plates found in the field have almost without exception been those of animals eight to nine years old. THE SENSORY RESPONSES OF CHITON 171 Chitons are eaten as the main constituent of ‘suck-rock soup’ by some of the poorer people of Bermuda. One of us has observed rats quickly seizing chitons and devouring them. From the attacks of many other carnivorous animals of the shore zone, C. tuberculatus is relatively immune. About 7 in 1000 were found with oyster-drill holes in one or more valves. ‘The animals whose shells were so attacked were always still alive. The holes pierced merely the tegmentum, the dense, hard articulamentum being impervious to the oyster-drill’s efforts. Although as many as 125 barnacles have been noted upon one chiton of medium size, it does not appear that they produce a deleterious effect. After the death of the chiton, the barnacles drop off, frequently without leaving any trace; they are never very firmly attached, except in the case of very old chitons with thoroughly eroded valves, and it does not seem as though they can even pierce the periostra- cum. The same applies to Spirorbis. Serpulids grow only on very old chitons; they become incorporated in the substance of the shell, and appear to be in some instances responsible for a local increase in its thickness. Injuries suffered by the girdle can be slowly repaired. Several animals were examined four weeks after they had been marked upon the girdle by having a deep notch cut init. The notch had been partly filled in by new mantle tissue, the new dorsal surface bearing small, irregularly distributed plates. The new plates were at first widely separated and irregular in shape. After six weeks they were still irregular, but had become more closely set together. The power of regenerating the plates (‘scales’) may be related to the fact that the periostracum of these plates appears lustrous and uneroded long after the shell plates have been intensely weathered. Chitons are sometimes found in the field with small groups of the girdle scales renioved, exposing the bare mantle, as well as with notches or ‘bites’ removed from the girdle. The general impression derived from the consideration of destructive agents in relation to Chiton is that these mollusks are very efficiently protected. The length of life which they seem to attain, the variety of habitats which they frequent, and the £72 LESLIE B. AREY AND W. J. CROZIER character of their sensory responses, which determine certain features of their life in these habitats, afford important evidence to this effect. 4. Feeding All the chitons, probably, are vegetable feeders. They rasp the thin coating of algae from the rocks by means of the radula (H. Jordan, ’13), which is long, armed with powerful black teeth, and operated by a complex arrangement of muscles (Plate, ’97). The body musculature is also involved in feeding. The whole body ‘lurches’ back and forth synchronously with the use of the radula, the forward swing coinciding with the retraction of the radula; the foot remains stationary. In rock crevices C. tuber- culatus occurs frequently in groups, piled one animal upon another. Investigation has shown that under these circum- stances they may feed on one another’s backs upon the algae growing there. The radula removes not merely the algae, but some of the rock surface as well. The chitons may be of some slight geological importance in this way, and they may also be in small part responsible for the destruction of the periostraca of their associates and thus for the weathering of their shell plates. Most of their feeding seems to be done at high tide. It is when covered with water that they move about most freely, although in damp places they also move to some extent at low tide. The great majority of the individuals are found well confined within tidal limits. While exposed to the air as the water falls they defecate copiously. The feces are discharged in the form of tiny cream-colored, cigar-shaped masses, varying in length with the length of the animal (fig. 1); in an animal 8 cm. long the fecal masses are 3.1 mm. long and 1 mm. in greatest diameter. The masses consist for the most part of minute granular bits of sand, but contain also undigested plant remains and fatty globules. When treated with acid, bubbles of CO. appear; all but a slight meshwork of algae fragments is dissolved. The mass of plant fibers holds the fecal matter together in a pellet, which persists for as much as twelve hours under water in nature. When it is considered that, along the north shore of Long Island, for THE SENSORY RESPONSES OF CHITON 173 example, more than 700 chitons, averaging 7 cm. length, were found within a strip three-eighths of a mile long, their eroding importance will be admitted to deserve examination. (A study of this matter is being made.) By the time the tide has risen one-quarter, every chiton in an intertidal group is found to have deposited a considerable mass of fecal matter within the anal region of the mantle cavity. At high tide they do not appear to defecate to any great extent. There would seem, therefore, to be some rhythmic sequence of feeding operations roughly codrdinated with tidal events. This might assist in the determination of a metabolic rhythm, which might in turn receive expression in (tidal) rhythms of behavior. 5. Respiration The respiration of most individuals of C. tuberculatus is also subjected to the influences of tidal events. Under water, Chiton obtains oxygen by means of a water current, passing inward laterally along the girdle, through the gills, and escaping at the anal end (fig. 7). Out of water, the gills are more or less con- tracted against the dorsal wall of the ctenidial channel. Some oxygenation may, however, occur out of water, since the gills remain damp, and in nature the girdle is usually lifted from the substrate to some slight extent, unless the creature be disturbed. The girdle is important for respiration, as the region in which it is lifted from the substrate localizes the intake for the water current. When completely submerged, this is commonly at the anterior end. The incoming water then impinges upon the dorsal surface of the proboscis (‘palp’). Water is also taken in at the sides of the body. The latter is exclusively the case when the chiton is but partly submerged (i.e., with the anterior end out of water). The girdle may be locally lifted in the form of channels (fig. 7) or may be completely lifted. The water passes up between the gills, and escapes under an elevation of the girdle at the posterior end. ‘This elevation is of somewhat variable form, although always located between the posterior ends of the right and left gill series. It is formed as a direct result of the water current impinging on the inner ventral margin of the girdle. 174 LESLIE B. AREY AND W. J. CROZIER Figure 8 illustrates this point. When a chiton, partially out of water, on the wall of an aquarium, swings from a vertical posi- tion (fig. 7) to one such as that shown in figure 8, the posterior, elevated part of the girdle travels to one side as a smooth wave. The water current also enables a chiton to sample the surround- ing water. It is of importance for reproduction, since the stim- ulus to egg laying is provided by the diffusion of sperm from near-by males; these sperms are carried past the openings of the oviducts, past the ‘osphradia’ (p. 253), and eggs are liberated Fig. 7 Illustrating the course of the water current n Chiton. Diagram- matic. Fig. 8 The course of the water current in Chiton when the animal is par- tially submerged. Diagrammatic. in their company (Metcalf, ’92; Heath, ’99, ’05 ec). The ne- phridia also discharge their excretions into the respiratory cur- rent. These excretions, together with the water that has been ‘used,’ are usually shot to a considerable distance, because, the anal opening being smaller than the incurrent openings, the velocity of the outgoing current is high; here also, as in Ascidia (Hecht, 718), the ‘used’ water is discharged in such a way that it is not readily employed ‘again for breathing purposes. The ventral surface of the girdle is transversely ribbed, pro- viding minute channels through which water is taken in, even THE SENSORY RESPONSES OF CHITON Ke when the girdle is not detectably lifted; this can be demonstrated with suspended carmine. The girdle can, however, be very tightly applied to a smooth surface. A chiton, if attached to the wall of the dish, will live for two or three days completely submerged in an aquarium containing other dead and decaying chitons. During this time no water is taken into the gill chan- nels. Hence, although chitons appear to frequent regions where, by wave action, the water is well aerated, it does not appear that they are particularly sensitive to want of oxygen.® 6. Migrations; association in groups The larger chitons rarely engage in creeping movements unless they are at least partly under water. Occasionally they are seen to creep about when the wet under surfaces of rocks on which they may be situated are turned over and exposed to the light. They also creep, slightly, on wet rocks covered with algae. In dark pockets within the walls of caves, where compact groups of chitons may be found, they may be seen, if watched carefully, to move slightly upon one another; such places are, however, decidedly damp. When the tide comes in and covers a chiton, it may become active immediately. Conversely, when left by the receding of the tide, a chiton usually stops creep- ing and remains where it happens to be. If, however, water be splashed over it, it will continue creeping for a longer time; if the splashing be stopped, the animal stops creeping immediately. Even when left in the sun to dry, upon the tide’s falling, Chiton is not entirely immovable. In the case of the larger animals, if they be partially covered by a shadow, they will, even in this condition, move forward, or backward, or turn slightly, so as to become more evenly adjusted with reference to the light. The possibility of such movements suggested that an 6 According to Heath (’05 ¢, p. 392), the gills of Trachydermon raymondi, which employs the gill cavities as breeding chambers, may become occluded during the breeding season by the 200 or more young trachydermons therein sheltered (Plate, ’99, Taf. 6, fig. 218); under these circumstances the lateral proboscis lappets become (like the whole proboscis) much distended with blood, and may then be concerned in respiration. 176 LESLIE B. AREY AND W. J. CROZIER ‘anticipatory’ creeping toward the rising water of the incoming tide, based upon some form of hydrotropism, might be dis- covered in Chiton, and was accordingly looked for. None was found. Chiton is in this regard analogous to the actinians (Parker, ’17 b); there is no ‘memory’ of recurring tidal events. Having in mind the possible metabolic basis of tidal rhythms in behavior, discussed in connection with feeding and respiration, the behavior of Chiton has been studied for the occurrence of other tidal rhythms—inactivity as to creeping, movement out of water, and the like. Nothing of this kind seems to occur in C. tuberculatus. Chiton, unlike a limpet, does not settle down into a de- pression closely conforming in outline to the impression of its shell. Neither does it, like a limpet, leave evidence upon the rock surface of wanderings and returnings to a ‘home station’ (Orton, 714). Inasmuch as a number of chitons seemed always to be present in certain depressions, or ‘pockets,’ which were ex- amined at low tides, and since observation of the behavior of other chitons showed that they usually began to move about as soon as the rising tide had wetted them, data were sought to answer the question as to whether chiton exhibits in one form or another ‘homing habits’ of the type which have been described for Patella and its allies (Kafka, 714). ' An experiment of this sort is here recorded: June 15, 1914. Observations were restricted to a definite area of smooth rocks below the boat house on Agar’s Island. The chitons were marked for subsequent recognition by means of a deep notch cut in the girdle on either side of the body. (As noted elsewhere, about four weeks were required for such notches to be even par- tially obliterated through regeneration.) In the area of shore con- cerned in this record there were several deep crevices and niches into which chitons crept. The observations begun at this date were con- tinued until July 14 (see table 1). This table shows plainly that for a period of twenty-six days no material additions were made to the chiton population of this particular section of the shore, although it did appear that there were occasional new arrivals. In all, twenty-four chitons were marked, and at the end of the experiment, twenty-four days after THE SENSORY RESPONSES OF CHITON 177 the last one had been marked, eight of these still inhabited the restricted region which was examined. ‘The occasional arrival of a new chiton in this area is consistent with the gradual and fluctuating disappearance of the marked individuals. Perhaps the handling and stimulation due to cutting for marking pur- poses caused an initially increased wandering of the marked TABLE 1 Concerning the migration and ‘homing habits’ of Chiton TOTAL | PREVI- DATE | HOUR ee sae MAR REMARKS SEEN ED 6/15 | 11.30} 6 0 4 (1 lost; 1 injured and rejected) 10 found in 6/15 | 6.00) 16 3 10 deep crack; these overlooked before? 6/16 | 10.00} 12 11 if 6/16 5.30) 18 12 6 | A new niche found, overlooked before; both marked and unmarked animals were in it. 6/17 | 11.00} 10 9 i 6/17 6.00) 13 12 0 | One individual in a crevice, could not be seen well. A new one? 6/18 | 11.30} 14 13 1 6/18 6.15) 7 a 0 | Tide not completely down; hard to see clearly. 6/19 | 9.30} 9 9 0 6/19 6.30) 10 9 1 6/20 | 10.00} 7 7 0 6/21 | 11.00} 8 a 0 | One hidden; impossible to distinguish whether marked or not. 6/29 | 11.00) 12 9 0 | Three others seen but inaccessible; marked? Ue 6.00} 10 8 0 | Two others seen but inaccessible; marked? 7/14 | 6.00) 9 8 0 specimens. The general result is clear, however: Chiton is not stationary, it does move about to some extent, but adult ani- mals, such as those used in this experiment, do not move fre- quently from place to place. Further observations showed that there is probably some correlation with age in the matter of migration. The youngest 178 LESLIE B. AREY AND W. J. CROZIER chitons move about more rapidly; if a small area of a shore which they inhabit be cleared of them, it will in some cases be found in several days’ time to be again inhabited by chitons of about the same age. A number of chitons were marked and placed on marked spots in the locality from which they had been taken, but were all found to have moved to new places within a single tidal period. Older chitons, however, eight to nine years of age, seem more restricted in their wanderings. Thus, upon the ver- tical face of a concrete wharf at Dyer Island, in a relatively protected spot, one solitary chiton of this size was watched every day for a period of nine months. During this time it remained within an area about 3 feet by 2 feet, sometimes moving up or down with the tide, sometimes being relatively stationary for periods of several days. At the time of writing it is still in place. In the case of younger and medium-aged chitons inhabiting caves, it does not appear likely that they migrate from these caves to any appreciable degree, until they become very old. In this way, and particularly through their habits in relation to light, more or less temporary groups or associations of chitons may be formed in pockets or cracks in the rock or under boulders. Migration from island to island would seem to be quite rare, if indeed it takes place at all, during adult life. The occurrence of chitons in groups is the general rule, save in the case of the very oldest individuals. A boulder, a crack, or crevice, even a very shallow depression, serves for the collec- tion of a group. Usually a group of this kind consists mainly of specimens of about the same size (age). 7. Breeding habits Metcalf (’92, ’93) observed in ‘C. squamosus’ and in ‘C. mar- moratus’ (possibly including the form we have studied) that fertilization was external, sperms being shed first, then eggs. Heath (’99, ’05 c) found that in various chitons the diffusion of sperms from nearby males provided the stimulus for egg laying. There is one point connected with this matter which deserves passing attention. The groups of chitons to which we have THE SENSORY RESPONSES OF CHITON 179 referred contain individuals of both sexes. This happens by chance, since the groups usually contain a considerable number of specimens and the habits of the two sexes are identical. Soli- tary individuals are of either sex. Some examples may be given: April 4, 1918. Hawkins Island, about 1 foot below high water mark, in the zone of Modiolus and barnacles, in a horizontal crack were found nine chitons. NUMBER LENGTH SEX cms, VI:111. 5 4.5 rot 9 6.5 rot 4 6.8 of 1 Gal Q 7 7.6 of 8 lel oh 6 7.9 & 3 8.6 of 2 8.8 2 No other specimens within a radius of 50 feet. The preponderance of males here was exceptional. April 10, 1918. Long Island. On the north face of a rock distant 10 feet from the shore, three chitons. ‘ NUMBER LENGTH SEX cms. V1:129. 46 7.8 2 44 8.8 . 45 9.4 of The occurrence of the larger individuals more or less in groups by themselves, imperfectly illustrated by these two cases, is a very real condition of the distribution of Chiton, although difficult to describe in any detail. Now, it would seem that the liberation of sperm occurs with the submergence of a male by the rising of the tide. Spawning occurs mostly in June and July, although sperm may be liberated in May, especially when the animals are transferred to aquaria; at this time of year relative calm pre- vails. Hence it seems possible that a male should (at least at times) fertilize mainly the near-by females. Since the larger 180 LESLIE B. AREY AND W. J. CROZIER animals tend to occur in isolated groups, and further from high- water level than the younger individuals, there results a type of segregation which is favorable to the occurrence of some degree of homogamy (assortive fertilization). This would result in the economical utilization of sperms, and might possibly have ad- ditional effects of an ‘adaptive’ kind. Some further correlation between habitat and breeding habits in other chitons have been noted by Heath (05 ¢, 07). a tok a ae y SEs we a : ’ z ee er 4 = naan wait waist SOS ir H- Sa shore Fig. 9 A group of Chitons in a shallow depression. X 4. 8. Bionomic correlations Chiton tuberculatus is strictly intertidal in habitat. It, there- fore, becomes possible to examine the details of its natural history rather extensively. The complexity of the catenary systems of relations revealed by such examination renders orderly description difficult. Some of these relations we have referred to in the preceding sections. Numerous others remain THE SENSORY RESPONSES OF CHITON 181 to be considered. They concern phenomena of coloration, re- production, determination of ‘choice’ of habitat, and similar features, comprising some of the things which involve explanation in terms of the animal’s sensory physiology. In the matter of coloration, for example, the chitons in general exhibit homo- chromic (‘concealing’) characteristics which are commonly of some precision (e.g., in Cryptochiton, Heath, 05 b, p. 213, and in other genera which we have observed; cf. also Plate, 1901 a, p. 376). In C. tuberculatus this homochromie correla- tion is decidedly evident—in most cases it is unmistakable. It _ involves several pigments, their mode of distribution, the over- growth of the shell by algae, barnacles, etc., a shifting of the chiton during growth to stations further below high-water level, the erosion of the valves, and a further shifting to more exposed habitats, with corresponding changes in the appearnace of the creature. There is little reason to doubt that in the later growth of the chitons (three to four years old) conditions of food supply directly determine through the course of metabolism the char- acter of the pigmentation displayed in the periostracum. The most fundamental factor concerned in the changing habits and appearance of chiton with advancing years, however, is its move- ment into more illuminated areas. The whole problem of its bionomie correlations becomes, from this standpoint, somewhat more directly open to attack. In general, it is not: How are the bodily processes kept going by the aid of movements? and, How does it happen that the movements are of such a character as to keep the processes going? (Jennings, ’07, p. 57), but rather: What is the relation between the sensory capacities which deter- mine and direct the bodily movements, on the one hand, and on the other hand the way in which the bodily processes are found actually to be going? Naturalists have long been content to assign a given ‘reaction’ to some one or another of the categories of adaptation, and to rest satisfied that progress had thus been made in explaining it. No progress can be made in this way. Neither are we greatly helped by placing the responsibility for the adaptation in a general way upon the environment. The situations requiring THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 29, No. 2 182 LESLIE B. AREY AND W. J. CROZIER analysis are too specific. The real problem is to trace to their sources of origin some of the harmonious correlations—involving habits, coloration, and the like—which specific organisms display. For studies of this kind Chiton tuberculatus affords eminently advantageous material. The analysis of its sensory character- istics, forming the body of this report, is taken therefore as the starting point for a series of quantitative investigations in ethology. III. MOVEMENTS AND REACTIONS ‘For the analysis of the sensory capacities of Chiton we depend upon its motor reactions under various forms of activation. It is therefore necessary to outline the different modes of response exhibited by these animals. The movements of chitons present, in fact, a certain degree of diversity, somewhat at variance with the traditional epithet ‘sluggish,’ so frequently applied to them. Slow, as a rule, the movements undoubtedly are, and for that reason particularly favorable for examination, as the responses can be studied with precision. The motor reactions of Chiton comprise movements of local parts of the body, bendings and twistings of the animal as a whole, and pedal locomotion. This classification of movements is largely artificial, but convenient. Each of these classes may be dealt with in further detail. 1. Local movements Local responses may be obtained from almost every part of Chiton. Since the muscular organs concerned in these move- ments are described in Plate’s monograph (’97, ’99, ’01 a), they will not be considered here. The girdle (fig. 10) reacts locally by puckerings and by bending movements. The individual shell plates may be pushed apart from one another, elevated, depressed, and closely approximated. These local responses are involved also in the general movements of the whole animal. _The local movements of the ventral parts are less directly in- volved in responses of the animal as a whole. At the anterior end (fig. 11), the mouth, a transverse slit, is situated upon a THE SENSORY RESPONSES OF CHITON 183 proboscis (fig. 14), clearly marked off from the foot. The pe- riphery of the proboscis is thin and very mobile, and reacts by local contractions and bending movements when irritated. During feeding the mouth opens in rhythmic fashion to permit the extrusion of the subradular organ and lingual ribbon, as Heath (’03) observed in Cryptochiton. The whole proboscis may be retracted and temporarily covered by the forward ex- tension of the anterior part of the foot; Heath (’99, p. 579; sep., p. 4) noted that the proboscis of Ischnochiton magdallensis was completely exposed in animals up to 4 mm. in length, but that with further growth it became normally covered by the pig- mented anterior part of the foot. The surface of the foot itself is locally reactive, as shown by puckerings upon its surface and along its margins due to contraction. The substance of the foot may be swung, either as a whole or in any local part, to one side or the other, and may also be considerably extended as well as tightly contracted. The gills respond singly to local stimulation by contracting in such a way as to be pulled dorsally toward the wall of the gill channel. They may also, under certain circumstances, exhibit synchronous movements. The anal papilla is capable of movements of extension, retraction, and sidewise bending. The neuromuscular mechanism of these movements is to a large extent locally contained. If the head segment or the tail segment be cut off, the tissues in the piece removed (head, sole of foot) are reactive to touch; stimulation of the foot or of the mantle causes the foot to be drawn toward the source of irri- tation. The part of the chiton remaining after the amputation also gives the customary responses, although the gill reactions are usually weak. The ‘shock’ effect of such an operation is of course severe, and is probably even greater when an animal is bisected transversely. In this case both halves are reactive, but the amplitude of the responses is much decreased, the gill responses to touch being absent. The local ‘reflex’ character of the regional movements is fully substantiated by experi- ments to be described on subsequent pages. This condition is reflected in the nervous architecture of Chiton, the central nerv- 184 LESLIE B. AREY AND W. J. CROZIER ous apparatus being relatively unconcentrated and containing ganglion cells along the length of both pedal and pallial nerve strands. The large size of C. tuberculatus and the fact that it Fig. 10 Outline of a medium-sized Chiton tuberculatus, dorsal aspect. The detail drawing of a part of the girdle is magnified diameters. > 1. Fig. 11 The same, ventral aspect. X 1. will remain in an active condition out of water for a long time make it possible to study the local reactions in this species with considerable detail. THE SENSORY RESPONSES OF CHITON 185 2. Movements of the animal as a whole Probably the most striking general reactions of chiton are the suction process, whereby the animal adheres tightly to the rock, and the curling up, armadillo-fashion, which it exhibits when detached. Locomotion has less frequently excited remark (Cooke, ’95, p. 400; Heath, ’99, p. 579; sep., p. 4). The suction power of chiton is well known to collectors (Dall, 07, p. 23). When the animal is disturbed, the girdle is applied to the substrate over its whole length, the shell plates are closely approximated, and suction is also exerted by the foot. The girdle is, however, the most important organ concerned in this protective response. Its efficiency is in part conditioned by its flexibility and by the fine riblets upon its ventral surface, but especially by the fact that it is morphologically differentiated into two concentric rings. This differentiation is exhibited in the coloration of the girdle, a narrow pale line being frequently located immediately inside the peripheral half of the girdle breadth. When firmly attached, a depression appears along this line, the more peripheral zone of the girdle being applied to the rock, and the inner zone being then sharply arched (figs. 10 and 11). Ona smooth glass surface a chiton may readily be pushed about from side to side; in this case the foot is not exerting any suction, although the animal seems to be as firmly attached as ever. The girdle is important not only for protection as a ‘holdfast,’ but also because it prevents the entrance of rain-water and of sand into the gill channels. Rain-water is quite toxic for Chiton, killing it in about four hours when the animal is placed upon its dorsal surface in a liter of such water. However, chitons will live for twelve hours or longer completely submerged in rain- water, provided the foot and girdle are free to come completely into contact with a solid surface; we have already noticed the completeness with which a foul solution can be excluded in this way. Chiton rarely frequents situations where it might be cov- ered with sand, but occasionally it is left by the receding tide with the girdle and more or less of the back so covered. The 186 LESLIE B. AREY AND W. J. CROZIER girdle is then closely applied to the substratum, and although, while under water, a respiratory current may be demonstrated, no sand grains gain admittance to the gills. The riblets and tiny channels normal to the girdle margin are important in this con- nection. If during creeping a lightly sanded spot is encountered, the girdle acts immediately as a plough, causing the sand to be pushed to one side. Although the tactile response of the girdle is thus very delicately adjusted, the commensal isopods (p. 167) are able to insinuate themselves beneath it without (usually) inducing any response. The ‘rolling up’ of the body is the activity of chiton most frequently mentioned in descriptions. The animal when de- tached from the rock, even in the case of the smallest specimens, usually bends the head end sharply ventralward, the curvature of the posterior end following, so that the body becomes ultimately rolled together, the anterior edge of the girdle sometimes being beneath the posterior extremity, at other times the two ends being simply in close contact. This response might conceivably be of significance in the life of chiton. When placed upon its dorsum on a smooth surface, it is impossible for C. tuberculatus to right itself. When rolled together, however, it could easily be moved by wave action to a location more favorable for righting. Moreover, the dorsal sur- face of the valves being sharply arched in the mid-line, the ani- mal automatically rolls over to one side. This results in righting behavior somewhat similar to that evidenced by Holothuria (Crozier, ’15b). That it is ever actually resorted to in nature is quite improbable. It seems merely that the ‘curling’ is an un- natural result of the tendency to maintain the foot in contact with the substratum, its protective appearance and functional value in righting being illusory. The flexibility of the body shown in ‘curling up’ is also evi- dent in other movements. Although the plates are closely ar- ticulated, some sidewise bending is nevertheless possible. The animal may also become arched dorsally to a considerable extent, as well as bent sharply in the ventral direction, at any level. This flexibility is rarely shown in the natural habitat of chiton THE SENSORY RESPONSES OF CHITON 187 except when it is creeping over a sharply curvesurface. There seems to be a pronounced tendency for the avoidance of un- even tensions among the muscles. The normal place of resi- dence is upon a flat surface; the somewhat unexpected flexibility of the body is nevertheless important, since it enables the girdle and foot to remain, during creeping, in close contact with the substratum, even though the latter be quite irregular. The flexibility of the body is more evident in a form such as Ischnochiton purpurascens. This animal is long and narrow. It creeps with unexpected freedom, drops from one rock surface to another, when stimulated by light, and rights itself easily. - 3. Locomotion The locomotor activities of Chiton demand a few words at this point, and we are able to add slightly to previous descriptions of its pedal movements. C. tuberculatus characteristically pro- gresses in an anterior direction. This is accomplished by means of pedal waves which are of a retrograde character, coursing from anterior to posterior as the chiton advances (Parker, ’11, 714); in this respect it resembles another placophoran studied by Vlés (07). As Parker observed, however, C. tuberculatus can also make backward movements of limited extent. Olmsted showed (17 a) that in these backward movements of chiton the retro- grade direction of the pedal wave is retained, as is also true in the Fisurella which Olmsted forced to creep posteriorly for a short distance; this we can confirm both for the pedal wave in chitons constrained to creep posteriorly, as in Olmsted’s experiment, by having merely a small part of the posterior region of the foot attached to a substratum, and also for the occasional backward movements which occur when the whole foot is attached. Lat- eral waves, or at any rate one lateral wave-like movement at a time, are produced on the foot when the animal is intensely stimulated on one side (Parker, ’14); in this case we find that the pedal wave courses from the unstimulated to the stimulated side (1.e., it is retrograde), but it not noticeably lifted from the sub- stratum in wave form. Similar movements appear at the ante- 188 LESLIE B. AREY AND W. J. CROZIER rior end of the foot during active creeping. The lateral wave produces an appreciable sideways shifting of the whole animal. Parker (714) also noted that “‘by swinging the anterior portion of the foot to one side and the posterior portion to the other; the animal can rotate its body with the middle of its foot as a pivot.” We have observed that in the case of these turning movements, which are frequently employed by chiton, the anterior end of the animal is the one primarily and principally concerned; the ante- rior end of the foot is, as a whole, pushed over to one side and diagonal retrograde waves bring the rest of the foot into the new position. The posterior end of the foot is pushed, as a whole, toward the side opposite the anterior one, but relatively not so far. During the turning maneuver the shell of the animal and the girdle are usually left behind, but after one or two pedal waves have passed, the foot (now straight) is held stationary, while the whole body of the chiton is swung slowly into the new position. We have spoken of diagonal waves upon the foot during turning; these waves are diagonal so far as the anterior end of the foot is concerned, but they usually become almost perfectly transverse before they reach the middle of the animal’s length. From the foregoing account it will be seen that there is, in comparison with most gastropods, a considerable degree of flexi- bility as to the use of the foot as a whole and as to the nature of the muscular coérdinations producing pedal waves upon its sur- face, although this flexibility does not by any means involve such complexity of movements as appears in the foot of the gastropod Cyprea (Olmsted, ’17 a). In the main Chiton tuber- culatus progresses anteriorly by means of retrograde pedal waves; these waves in their characteristic form run almost entirely lengthwise on the foot from the anterior end backward and are not free to course in all directions across the foot as they are in the pedal dise of sea anemones (Parker, 717 b). Undoubtedly, this difference in the character of the pedal waves in the two cases is determined by the nature of the nervous arrangements within the pedal organ, and in fact the disposition of the nervous system of chiton allows us to analyze the relation experimentally. THE SENSORY RESPONSES OF CHITON 189 During ordinary locomotion one or two waves appear upon the foot of Chiton; usually two waves are present when the animal is engaged in turning, but even in the absence of pivoting move- ments, one wave may appear at the anterior end before its prede- cessor has reached the posterior extremity of the foot. These waves are 5 to 7 mm. in anteroposterior extent, and involve the full breadth of the foot. They require fifteen to thirty seconds, usually about twenty-five seconds, to pass from one end of the foot to the other. The speed of progression of the pedal waves is less at lower temperature; it is identical in either sex, pro- vided the chitons are of the same size. At 27°C, the speed of propagation of the wave is usually about 12-15 em. per min- ute, being therefore faster than the rate of movement of the pedal wave in actinians (Parker, ’17 b, 1 to 3 cm. per minute). The pedal wave is a region, occupying about one-tenth to two- tenths the area of the foot, which is temporarily lifted from the substratum (Olmsted, ’17 a) and locally moved forward by mus- cular contraction. In backward locomotion, which may readily be induced by partial illumination of the shell, the retrograde character of the pedal wave is retained. This is especially evi- dent in Ischnochiton purpurascens, which creeps freely back- ward if stimulated by horizontal light striking the anterior end of the shell. In chiton there can sometimes be seen a distinct longitudinal depression running the full length of the foot in the midline, as if the foot were about to be folded together lengthwise. This is more easily seen in Ischnochiton. No trace of this activity is apparent in the pedal waves, however. Nevertheless, it can be shown that the foot is controlled in a bilateral manner. If an incision be made into the foot sufficiently deep to divide the con- nectives which join the pedal nerve strands, the lateral halves of the foot exhibit independent wave movements. If such an incision is made at the posterior end, a normal pedal wave may bifureate when it reaches the anterior end of the incision, one half of it becoming obliterated while the other half may continue. ‘Stationary waves,’ sometimes opposite, sometimes unilateral, appear on a foot completely divided in this way; four or five such waves may be present at once, 6 to 7 mm. apart. 190 LESLIE B. AREY AND W. J. CROZIER That the essential nervous mechanism of progression is locally contained within the foot is shown by the fact that when com- pletely excised the foot will exhibit spontaneous wave motions; usually such a foot (which will live for two to three days in sea- water) does not actually creep for more than a centimeter. The isolated foot reacts locally at its margin and on its ventral sur- face to touch, in the latter case giving well-defined suction re- sponses. The pedal waves formed by the isolated foot are normal as to their speed of transmission; moreover, they appear one at a time, in succession, as in ordinary creeping; usually two or three waves exhaust the foot for half an hour. The foot of placophorans, as of gastropods, serves also as a holdfast (Parker, ’11), either by means of slimy secretions or through the action of the foot as a sucker. Parker (’14) pointed out that in Chiton tuberculatus the foot sucks locally, so that. “af to the foot of an inverted chiton a rigid body with an area 5 mm. square is applied, the animal can attach itself to this area with sufficient strength to allow its weight to be lifted.” As we shall point out subsequently in this paper, this 5x 5 mm. area is about the minimal surface to which the Chiton foot will react by attachment and suction, so the full physical efficiency of its suction cannot, perhaps, be measured in this way. A chiton of 8 cm. length weighs approximately 50 grams, so in Parker’s experiment just cited the foot was probably exerting a suction pressure of not more than 2 grams per sq. mm.—considerably less than the almost perfect suction efficiency of the tubercles upon the column of Cribrina (Parker, ’17 e). These observations indicate that, although the chiton foot is employed as a hold-fast, the foot itself is not sufficient to account for the full suction power of these animals. The tenacity with which they adhere to a rock surface is sufficiently remarkable to have gained for them the local name ‘suck-rocks,’ and in a pre- ceding section we have shown how the girdle is of prime impor- tance in this connection. An individual from which the girdle has been completely removed may with relative ease be sepa- rated from a stone over which it has been creeping. This is also true if a chiton is caused to become attached to a glass plate in THE SENSORY RESPONSES OF CHITON 191 which there is a hole, provided the hole be situated beneath the gill channel. During its normal existence, however, the foot is of course the organ whereby the chiton maintains its position. The whole girdle is often, especially when under water, completely removed from contact with the substratum; the support of the animal depends, in fact, almost entirely.upon the foot. From the ease with which the chitons preserve their position in places where wave action is considerable, and upon the under surface of rocks, or (as has been noted through continuous observation) upon the relatively smooth vertical wall of a concrete wharf for periods of more than five months, it will be seen that the working power of the foot is, after all, adequate for the creature’s needs. Chiton gets the most possible out of this suction power of its foot by keeping its whole area closely pressed against the substratum. Since it commonly inhabits smooth rock surfaces, the foot usu- ally exhibits no great unevenness when the animals are freshly examined; but they are occasionally obtained creeping over bits of stone or groups of small Modiolus or barnacles, and if these individuals are inspected it is to be noted that the whole surface of the foot has been thrown into blebs and deep depressions cor- responding closely to the unevenness of the substratum. Some of the blebs produced under these circumstances clearly demon- strate the basic mechanical principle upon which the foot works, for they appear as thin-walled vesicles filled with fluid (in the females, orange in color like the coelomic juice). Sunilarly, if a Chiton be caused to creep over a small hole (4 to 5 mm. diameter) in a glass plate, the substance of the foot will be perceptibly pressed into the hole. Apparently the foot of chiton can exert suction only in a very local fashion, for if a portion of glass tubing of 8 mm. internal diameter, 10 mm. external diameter, corked at one end so as to provide a cylindrical chamber 4 mm. high, be used to test the sucking power of the foot, it is found that in most cases the chitons cannot become attached to the circular rim of the tube with sufficient force to bear their own weight in air. In this experiment, it should be noted, the total area available for direct contact with the foot is about 28 192 LESLIE B. AREY AND W. J. CROZIER sq. mm., agreeing with the minimal area for attachment as found by Parker and in our own tests. The local character of the suc- tion mechanism of the foot is further suggested by the minute depressions, usually long and narrow and more or less communi- cating with one another, which are to be found on the foot of a non-creeping Chiton attached for some hours to a glass plate in air. These local suctions are probably assisted by slime secre- tion, which, although small in actual amount, enables a chiton to remain rather firmly attached to a smooth surface (e.g., of a glass plate) after the animal has been allowed to die in air or after it has been killed by heat (44°C.) in water. They do re- main so fixed, even when the girdle is not in contact with the substratum, and the slime may therefore be important during the use of the foot in life, including early postlarval stages (Heath, 99, p. 640; separate, p. 65). IV. MECHANICAL EXCITATION 1. Tactile stimulation In testing the local sensitivity of Chiton to tactile excitation, use was made of a blunt-pointed dissecting needle, a glass rod, or a blunt pencil-lead. In some instances, also, minute air bubbles (formed at the end of a pipette) and several other means of stimulation were employed. The responses observed when different regions of the dorsal and ventral surface of Chiton were lightly touched with one or the other of these objects are de- scribed in the following summary. Attention was given to the possibility that the responses of Chiton might vary depending on whether the animal was submerged in water when tested or was out of water. There were discovered no differences in behavior which require consideration at this point when the reactions of chitons in these two situations were compared. For the study of the responses obtainable from the ventral surface, we have mostly employed animals in air, placed upon their dorsal sur- face. The inability of chiton to right itself, coupled with the relative insensitivity of the shell surface, allowed us to work in this way without introducing serious complicating disturbances. THE SENSORY RESPONSES OF CHITON 193 In the more critical experiments we made use of a method of graphic registration, subsequently described. A. Dorsal surface. a. The shell plates appear not to be sensi- tive to touch. No responses were obtained when the surface of the tegmenta was lightly touched. (This is further considered on a subsequent page.) b. The manile between the tegmenta, 1. e., the tissue covering the * insertion plates, may be somewhat exposed, when Chiton is at- tached and ‘at rest,’ by the separation of the shell plates through the extension of the body. When the mantle was touched in this region the plates immediately adjacent to the site of stimulation were quickly approximated, covering the mantle area which had been touched. c. The girdle. When Chiton is attached, the lateral extension of the mantle, known as the ‘girdle,’ which is flexible, is locally lifted from the substrate unless the animal be disturbed. Under water the girdle may be completely removed from contact with the rock or other surface, but in air this elevation is usually local and commonly takes the form of slight puckerings of, at most, a centimeter or so in length. To asingle touch the girdle, where elevated, responds by local lowering to the substrate at the point of excitation. A more vigorous touch causes a greater extent of the elevated girdle to be lowered. Four or five moder- ate touches in succession affect a still greater length of the girdle, as much as one-quarter to one-third of the circumference, and the time elapsing before recovery to the original elevation is longer than that following a single touch. Even when the girdle has not perceptibly removed from contact with the sub- stratum, it responds by a detectable ‘tightening,’ or flattening. Several successive touches upon a ‘flattened’ region of the girdle induce near-by elevated parts to return to the substratum. Un- less the excitation is continued for nearly one minute, however, or is in the first place very vigorous, the response to touch is strictly homolateral. The anterior end of the girdle is more reactive than the middle or posterior parts, and its peripheral border is more sensitive than the rest of its dorsal surface. A chiton quietly creeping in water, with the girdle lifted, re- 194 LESLIE B. AREY AND W. J. CROZIER sponds instantly to a touch upon the anterior or posterior girdle by ceasing locomotion and adhering firmly to the substratum. B. Ventral parts. d. The ventral surface of the girdle. The serial arrangement of eight dorsal shell plates affords a con- venient means of dividing the surface of the animal into defi- nitely delimited areas for reference. We shall consider the ven- tral surface of the girdle in terms of four ‘quarters,’—an ‘ante- rior quarter,’ delimited by the posterior margin of the second shell plate, two ‘middle quarters,’ and a ‘posterior quarter’ correspondingly marked off by the transverse borders of each succeeding two shell plates. The end quarters of the ventral mantle are more reactive to tactile excitation than are the middle quarters. a. The end quarters. To the single stimulation of an end quarter, the response is a curling of the animal in that region, as if it were beginning to roll up; the bending process elevates the stimulated end about 2 to 5 mm., after which the animal straight- ens out again. When stimulated on the posterior quarter, the foot may be pushed caudad at the time the bending response occurs. When stimulated on the anterior quarter, the head may retract somewhat, and the buccal region may be slightly in- verted, the head and ‘palp’ tending to close over the mouth. Light touches, when several times repeated, elicit a much stronger response. The stimulated end reacts first, by bending, and then the opposite end bends also, though to a less extent. This is true whether the anterior or the posterior end is the one stimulated. The muscles of the midportion of the animal are not specially contracted, however, and the closure of the shell is incomplete. 6. The middle half. The reactions listed under a are pro- duced most clearly when the most anterior or the most posterior region of the ventral girdle surface is stimulated. The reponses obtained from the ‘middle quarters’ of this surface are qualita-. tively identical over the whole anteroposterior extent of the ctenidia. In other words, the convenient descriptive division of the animal into ‘quarters’ does not afford a basis for the organic classification of responses, inasmuch as the ctenidia extend an- THE SENSORY RESPONSES OF CHITON 195 teriorly and posteriorly beyond the limits of our ‘middle half.’ ‘This artificial subdivision into quarters is retained in our de- scription, however, since the responses we are considering are most characteristically displayed in the respective ‘quarters’ of the chiton’s surface, although there is some ‘overlapping’ and, as already stated, the subdivision is by no means an organic one. A single touch applied to the midventral surface of the girdle is followed by a local puckering of the girdle toward the source of irritation. The foot, in the region immediately adjacent to the level stimulated, is pushed laterad and dorsad, toward the mantle, tending thus to assist the girdle in covering the gills. This reaction of the foot is not evident when the dorsal surface of the girdle is lightly stimulated. Unless the tactile stimulation is severe or several times repeated the homolateral side only of the foot is involved in this response. Simultaneously with these movements of the girdle and foot, a contraction of the gill ele- ments occurs opposite the singly stimulated area. This involves five or six ctenidia anterior, and as many more posterior, to the point of excitation. In this reaction the ctenidia are elevated dorsally, the tips are drawn toward their bases (thus throwing each element into a more convex arch), and at the same time they are drawn somewhat anteriorly. The whole response involves a movement something like the fairly rapid closure of the fingers of one hand. The response spreads in both directions from the level of stimulation, although at ordinary temperatures the propagation wave is difficult to observe because of its rapidity. Successive stimulations of the girdle lead to a greater puckering in toward the source of excitation, and to a more pronounced rolling up of the whole body. The foot is locally brought slightly laterad toward the girdle and is drawn dorsad to a considerable extent. This response of the foot is at first confined to the stimulated side, but subsequently spreads to the other side, finally involving the whole substance of the foot at the level of stimulation. Successive touches, 1 to 1.5 seconds apart, lead to a tetanic contracture of the foot and ctenidia; during this phase the animal tends to roll up. A ‘refractory period’ succeeds the application of repeated light touches, until relaxa- 196 LESLIE B. AREY AND W. J. CROZIER tion is fairly complete, during which touching locally does not evoke any response. The duration of this ‘refractory period’ de- pends upon the intensity of the original stimulation. The foot recovers very quickly and is not easily fatigued. As in the stimulation of the dorsal surface of the girdle, the periphery of the ventral aspect is more sensitive than its more medial regions. The most striking feature of these reactions is the strictly homolateral character of the response on the part of the ctenidia. In no ease did the reaction spread to the gills of the unstimulated side. Homolaterality in response is not so clearly shown in the behavior of the foot. . e. Tactile stimulation of the mantle lining of the shell in the region of the ctenidia induces movements of the foot and gill filaments similar to those which follow touching the ventral surface of the girdle at a corresponding level. The responses obtained from the stimulation of the mantle lining of the ante- rior and posterior extremities are also similar to those obtained from the ventral surface of the girdle in the same regions, but. the surface of the girdle is decidedly less sensitive. As with the girdle, no movements of the ctenidia are produced by tactile excitation of the mantle except along the anteroposterior extent. of the ctenidia themselves. f. The region of the anus is not excitable in any special way. To a light touch it may appear quite insensitive; to a stronger touch the anal papilla responds by local contraction on the side touched, and it may be somewhat retracted. When the mantle in this region is activated the behavior of the girdle and of the foot is essentially as already noted for other regions. g. The ctenidia, when directly touched, respond as has already been described in the cases where the girdle and ventral mantle in the gill region were touched. Touching either dorsal or ven- tral surface of a filament results in the same response, although in one case it is directed toward, in the other case away from, the source of stimulation. Contraction in the manner pre- viously described is the single mode of response of the gills. The base of each filament is more sensitive than the free tip. The THE SENSORY RESPONSES OF CHITON 197 central ‘rib’ of each filament is by far the most sensitive portion; it is possible to pass a fine needle between the gill filaments with- out getting any appreciable response, although the filaments could be seen to move. By employing a finely twisted bit of tissue-paper or a sande air bubble, it was possible to obtain the contraction of a single filament when delicately touched, although generally at least the two immediately adjacent ones were involved. This result, taken together with the low tactile irritability of the lateral borders of the gill filament, shows that the antero- posterior spreading of the gill response when the mantle is touched is a nervous matter, and is not merely the result of one contract- ing filament mechanically involving its neighbors. This is also confirmed by tests upon animals having the pallial nerve strands sectioned at various levels. The operation produces no serious disturbance. The response of the homolateral gill series to fairly severe tactile irritation at one spot does not spread past the level of the cut ‘nerve,’ but ceases abruptly at this point (al- though the gills themselves have been quite undamaged). The tactile sensitivity of the ctenidia is important for the efficiency of respiration. Foreign particles (e.g., sand) drawn into the gill channel by the respiratory current strike against the ventral surface of the gills, inducing a sudden local depression of the girdle, which squirts water out from under the girdle, remov- ing the foreign object. h. The surface of the head, the ‘palp,’ and the region of the mouth, are very sensitive to touch, the reactions produced being, however, merely local contractions; to more vigorous excitation, the animal responds by rolling up. The free margin of the head region is especially sensitive. i. The edge of the foot, when touched in the region of the gills, induces responses such as those already described for the mantle lining of the ctenidial chamber. It is very sensitive. The pos- terior end of the foot yields responses like those of this region of the mantle. j. The sole of the foot reacts by puckering away from a source of tactile irritation, such as a blunt glass needle. To larger areas it THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 29, NO. 2 198 LESLIE B. AREY AND W. J. CROZIER responds positively by local attachment. The negative response of the foot to the touch of small surfaces is entirely local for a single stimulation. When repeatedly stimulated in the same spot, the primary local puckering spreads slowly across the sur- face of the foot, producing a dorsally arched depression 4 to 8 mm. broad, which is similar to the pedal locomotor wave, but generally deeper. Repeated stimulation also tends to produce the rolling-up response. The anterior and posterior quarters of the foot are more reac- tive than the middle area. In these regions a single touch ini- tiates a tendency to roll up, which can be produced on the mid portion of the foot only through repeated applications. C. Distribution of sensitivity. The responses of Chiton when different parts of its surface are touched enable us to outline the distribution of tactile sensitivity.over its body. In a general way the anterior and posterior ends—as is almost, if not quite, universal among animals—are more sensitive than the middle portions, and the peripheral parts than those more medially situated. Employing as criteria the relative effect in causing the animal to roll up, and, on the sides, the relative effect upon the gills, the following orders of sensitivity have been distinguished : ; a. The most anterior and the most posterior regions of the mantle are about equally sensitive. It is difficult, if not impossible, to detect any constant difference in their reactivity to touch. b. At the head end: Surface of the head and palp = inner sur- face of mantle > ventral surface of girdle > dorsal surface of girdle. The extreme outer margin of the girdle is about as sensitive as its ventral surface. c. At the caudal end: End of foot = inner surface of mantle > ventral surface of girdle > dorsal surface of girdle. d. On the sides: Inside of manile = ctenidia > edge of foot > ventral surface of girdle > sole of foot > dorsal surface of girdle > dorsal mantle between shell plates, the last judged by its effect upon the approximation of the shell plates. e. The shell plates are insensitive to touch. It will be seen that in a broad sense the capacity of response to THE SENSORY RESPONSES OF CHITON 199 tactile irritation is distributed upon the body of Chiton in a manner appropriately correlated with its structure and habits. The various responses obtained from contact with a small sur- face are such as would have a protective influence. Reference should be made at this point to the local closing together of the dorsal valves when the intertegmental mantle is irritated, and especially to the ventralward movement of the girdle, associated with local retraction of the ctenidia and a corresponding local movement of the foot, which follows a touch upon the dorsal or ventral surface of the girdle. The preservation of the ctenidia from injury, and more particularly the effective use of the girdle for the exclusion of foreign objects and as a hold-fast, are de- pendent upon responses such as we find the girdle to exhibit. Further detailed correlations of this character might be pointed out, but enough has been said to indicate the useful nature of the responses. The ‘rolling-up’ reaction has, when carried to com- pletion, a clearly ‘purposeful’ aspect, as already intimated. Yet the natural history of Chiton yields no evidence that this re- sponse is ever used. We consider that it is the inevitable out- come of maximal possible contraction in the chiton’s effort to produce suction, and that it is neither of specific protective sig- nificance nor of the nature of a ‘reflex.’ Confirmation of this view is found in the fact that sometimes a ‘rolled-up’ chiton will re- main for hours tightly curled, although placed in position pur- posely made favorable for reattachment should it unroll. On the other hand, after a short time upon its back, a chiton may spontaneously uncoil itself and remain fully exposed for a long time, if unstimulated. Moreover, isolated parts of the animal give (or attempt to give, so far as their deficiencies permit) the ‘rolling-up’ response when they are activated. D. The tactile receptors. The superficial layer of the shell plates of chitons is traversed by numerous canals, occupied by specialized organs having the histological appearance of sensory receptors. These canals are more or less nearly perpendicular to the surface (at least peripherally), and the organs they con- tain, piercing the tegmentum, are described as projecting slightly beyond its general surface. The remarkable character of these 200 LESLIE B. AREY AND W. J. CROZIER structures is well known. Definite evidence as to their functional significance has been completely lacking. In addition to the ‘eyes’ (Moseley, ’85; Plate, ’99; Nowikoff, ’09); micra- and megalaesthete organs of varied form are present, and some of them seem so constructed as to be (?) serviceable as tactile re- ceptors. A function of this sort has in fact been somewhat doubtfully suggested for them (Kafka, ’14, p. 100). As already stated, however, the shell plates of adult chitons seemed insensitive to touch. foot, palp, ctenidia, for heat; foot, palp > ctenidia, for cold. To tactile excitation, the order of reactiveness is slightly dif- ferent (see p. 199), but there exists no adequate criterion for the comparison of the relative sensitivity of the several parts to heat. Only in the case of the dorsal mantle, between the plates, does it appear that thermal sensitivity is relatively enhanced as compared with tactile, since the high-temperature threshold (37°) seems to be lower than for other regions (40°) which are superior to the inter- tegmental mantle in tactile reactivity. The amplitude and vigor of the responses from this region are comparatively slight, how- ever, and little emphasis can be put upon this result. On the basis of their relative distribution, thermal and tactile receptivity cannot be clearly separated. By differential exhaustion an apparent separation of this kind can, nevertheless, be effected. When water at 10°C. was poured, in | ce. portions, from a pipette several times in succession over the anterior ctenidia of a chiton in water at 24°, the animal ceased to respond after the fourth treatment; six applications of cool water were made at intervals of three minutes. Very weak tactile responses were then obtainable from the affected ctenidia, al- though they still did not respond to water at 10°C. In attempt- ing to differentiate between ‘heat’ and tactile responses, this method of attack fails completely, since, as we have described, when chiton is placed in water at 38°, its general tactile reactivity was perceptibly increased, and much more conspicuously so im- mediately after immersion in water at 40°, although tactile re- activity gradually decreases after a few minutes’ exposure to this latter temperature. As a consequence of this condition, we are not warranted in speaking either of the separateness or the sen- sory identity of ‘heat’ and tactile effects, even though the surface of the foot and the ctenidia did almost cease to respond to touch after they had been exposed to four local treatments, in air, with 1 to 2 cc. of water at 40° at intervals of two minutes. THE SENSORY RESPONSES OF CHITON peB 8 3. Summary The evidence we have presented relative to the existence of a. temperature sense in Chiton shows that if specific thermoceptors. of some sort do indeed occur upon the surface of the animal, they are of a very poorly developed kind. Responses to high tem- perature, under the various conditions of these tests, cannot be adequately distinguished from tactile effects or even from direct influences upon muscle. The minimal temperature (37° to 40°) eliciting a ‘heat response’ is very close to the maximal temperature which the chitons successfully withstand, and is even higher than that which induces a distinct effect upon the muscular ‘sphincter’ about the oscula of Stylotella (Parker, ’10), where no receptor organs are involved. Although this temperature is identical with that producing heat responses in Amphioxus (Parker, ’08), it cannot be clearly shown by exhaustion tests—as apparently it can in Amphioxus—that ‘heat’ and tactile receptivity are in any way organically distinct. Only in the case of the intertegmental mantle is there a suggestion of special thermal sensitivity, and here the response elicited is not of a character favorable for analysis. With low temperatures, as with high, the limiting temperature producing perceptible responses in Chiton is prob- ably just outside the range of its normal thermal experience. The ‘cold’ responses, however, elicited at 12° to 15°, are of a definite character and may apparently be separated, through differential exhaustion, from purely tactile responses; that they are mediated by special sensory structures remains uncertain, but is possible. This matter of sensory differentiation is an exceedingly com- plex one. The fact that isolated cells of the metazoan body (e.g., chromatophores) are capable of excitation by heat, as well as by chemical agents, local pressure, and light (Spaeth, ’13), has of course no decisive weight as an argument for ‘generalized receptors’; yet the degree of heat (high temperature) effective as. a stimulus is in such cases of an order of magnitude comparable to that found effective in the sensory activation of many inver- tebrates. In comparing the relative sensitivity of different 224. LESLIE B. AREY AND W. J. CROZIER species, account must be taken of the toughness of the tissue concerned (the delicacy of the respective cell surfaces). This may explain why the delicate, internal, protected surface of the oral siphon of Ascidia (Hecht, ’18) exhibits a sensitivity to heat and to cold superior to that known for many other animals. VI. PHOTIC EXCITATION 1. Effects of light a. Behavior in an illuminated field. a. Preliminary experiments. Chitons collected more or less at random and without much at- tention to size were tested in a qualitative way with reference to their photic behavior. At the bottom of one end of a wooden box, 29 em. long by 23 em. wide by 30 cm. deep, there was cut a horizontal slot about 12 cm. long and 1 em. high. This box was coated on the inside with lampblack suspended in turpentine, giving an approximately dead-black finish. A rectangular glass jar containing sea-water to a depth of several centimeters was placed inside the black box, within which it fitted closely. Chitons were put in the glass jar, the box covered, light admitted (or directed) through the slot, and the subsequent movements of the animals determined. With diffuse sunlight twenty-one experiments upon twenty individual chitons gave this result: 6 did not move at all during the course of the test (lasting about one hour); 1 oriented a few degrees away from the light, while 13 animals made definite progress toward the light, irrespective of their original orien- tation in relation to it. These chitons were probably all of average size or larger. In some instances they were allowed to become fixed to the bottom of the aquarium with their anterior ends toward the light, in other cases they were placed with long axis perpendicular to the light, in still others deliberately headed away from it or quite at random. The nature of the result in these experiments will be evident from the following record: THE SENSORY RESPONSES OF CHITON 225- Experiment 1. 11:15 a.m. A chiton placed with long axis parallel to the slot admitting light, distant 13 cm. from it. 11:19 a.m. Oriented toward the light. 11:20 a.m. Began moving forward. At first moved in a diagonal di- rection, until the girdle touched the side wall of the container; it then turned further and moved directly toward the light. 11:24 a.m. Stopped, half-way toward the slot. 11:30 a.m. Began moving again. 11:34 a.m. Reached light end and began climbing end wall. 11:39 a.m. All except posterior quarter attached to end wall of con- tainer. Stopped. 11:43 a.m. Began again and moved until all of body was on vertical end wall. Turned until body axis was parallel to water line, where, just submerged, it lay directly over the slot. Experiment 2 11:45 a.m. Same chiton as in experiment 1, placed transversely to the light, but with other side illuminated, and 26 cm. from the light slot. 11:48 a.m. Began turning away from the light. 11:50 a.m. Had rotated away from the light,’ then back toward it, through an angle of more than 270°. 11:54 a.m. End had come in contact with side of container. Animal now began to climb. No forward progress toward the light. Experiments 3 and 4 In two further chitons tested in this way, orientation was in one case direct, beginning almost at once; in the other it required 29 min. (involving a preliminary turn- ing through 45° away from the light). Both animals made definite progress toward the light. These tests indicated in a general way the presence of a definite, though sluggish, positive phototropism, with reference to dif- fuse light. . With direct sunlight, reflected horizontally from a mirror, three individuals oriented promptly and move directly toward the light, two oriented toward the light and then away from it, four individuals immediately oriented more or less away from the light, and two did not move at all. This result obviously required further analysis; it might have been the outcome of a general illumination of the whole aquarium or might have ref- 8’ Note this apparent persistence of a turning tendency once established. 226 LESLIE B. AREY AND W. J. CROZIER erence more specifically to some definite peculiarity of behavior. These tests were made with chitons of relatively large size (6 to 8 em.) in which the shell valves were probably more or less eroded, although at the time no special note was made of their condition. 8. Analyses of responses to general illumination. The fore- going section indicates the somewhat obscure relations, with respect to phototropism, discovered in random samples of the chiton population. The younger individuals, especially those less than 2 em. long, live in dark situations. When stones bearing them are turned over, the chitons creep rapidly to the under, dark side. Not until a length of 7 to 9 cm. is attained does Chiton occur with any frequency upon illuminated rocks. If chitons from habitats representing these two divergent ex- tremes are compared, it is found that in ordinary sunlight the larger ones are photopositive, the younger ones photonegative. Their orientation is precise, definite, and without ‘trial move- ments.’ There are, however, certain complications in the mode of orientation which will be fully considered on a later page. The foliowing test is typical:° LENGTH OF INDI- HABITAT PHOTIC BEHAVIOR VIDUAL cm, 1.0 | Under stone on a sandy beach, | Consistently photonegative to the south side Darrell Island. (No. weakest daylight used. VI, 119.B)® | Same locality. (VI. 119.C) Photonegative to direct sunlight; photopositive to weak diffuse light, and to twilight. 3.5 | In a pocket at the mouth of a | Photonegative to direct sunlight; bo _ cave, north shore Long Island. photopositive to light from a (VI. 122.1) north window 10 ft. away. 5.0 | Same locality (VI.128.9) Same 7.2. | On an approximately horizontal | Photopositive to diffise daylight; rock, exposed in the sun, north photopositive to direct sunlight shore Marshall Island. (VI. from a cloudless sky 140.2) 8.3 | Same locality. (VI. 140.4) Same. ° The specific animals bear definite numbers given to them in the field note- book. In a subsequent report on the ethology of C. tuberculatus the necessity for this will be made apparent. THE SENSORY RESPONSES OF CHITON Dok These six specimens illustrate a correlation which is universal in our experience: the youngest chitons are found in dark- situations, the older ones in the light; these two groups are, respectively, photonegative and photopositive to ordinary sun- light; animals of intermediate size are photopositive to weak light, photonegative to stronger light, and the character of their normal photic environment (typically, in horizontal crevices near the mouth of caves, and in other non-brilliantly illuminated spots) is completely correlated with this behavior. There is usually a region of intensities of ordinary daylight within which an animal of intermediate size may be either photopositive or photonegative in different tests—a region of seeming indifference to light. The largest chitons are usually quite indifferent to weak, diffused light. The actual distribution of the chitons of different sizes in the field shows in a most convincing manner that this differentiation in photic responses is not a matter of adaptation to environmental circumstances, but is on the con- trary based upon structural changes determined with advancing age. Note, for example, the following instances in which an individual photonegative to sunlight (as found by test) occurred on a shore where no deep caves were available, nor any large stones under which it might hide. VI. 111. (Apl. 4, 1918). North shore of Hawkins Island; a more or less horizontal shelf of rock, 1 foot beneath high water mark; nine chitons in a closely compacted group, in the zone of Modiolus and barnacles. Eight of the chitons with eroded valves, forming a fairly close match with the color of the exposed rock; in sunlight; these chitons 6.5 to 8.8 em. in length. One chiton, however, was a o’, 4.5 em. in length, the valves greenish, very slightly eroded; it was located under another individual (o’, 6.8 em. long), which completely con- cealed and sheltered it from the light. VI. 140 (Apl. 22, 1918). -Northwest shore of Marshall Island. In a small pocket in the rock four chitons of 7.2 to 8.7 em. were found; of these the shells were eroded and bleached. Under one of them a fifth specimen, 2.8 em. long, blue-green in color, valves uneroded. b. Results of partial illumination of the body. «a. One side of the body illuminated. Chitons were adjusted in the apparatus shown diagrammatically in figure 13; they were so situated, directly under the vertical partition, that either the right or left 228 LESLIE B. AREY AND W. J. CROZIER side was in comparative darkness, the other side in the light. Diffused sunlight, employed with chitons of medium to large size, induced responses of a variable kind. Experiment 1. Two chitons introduced. One moved partly toward the dark side, and subsequently, in the course of 45 min., moved back into the center of the light compartment. The other one rotated through 180°, its anterior end passing through the dark side and then orienting into the light. Experiment 6. Two chitons used. One began to move at once; oriented 90° into the dark, and moved so that almost the whole of ’ its body was in the dark compartment; after 5 min., it turned through 180° and moved straight out into the light. The other one within a Fig. 13 Sectional view of apparatus for the partial illumination of a Chiton: a shallow pan fitting the bottom of a box with blackened walls; one half of box open to receive light, as indicated by the arrows, the other half covered and separated from first half by a blackened vertical partition extending into the sea-water, that nearly fills the pan; horizontal overhang from the lip of the pan exposed to light reduces reflection. few minutes oriented weakly into the light, halted 1 min., and then oriented anterior end back into the dark, so that half of its body was on either side of the partition. In 15 individual tests of this nature, 4 chitons turned and moved directly into the light, where they remained. 0 chiton turned and moved directly into the dark. 4 chitons turned and moved into the dark, then into the light. 0 chiton turned and moved into the light, then into the dark. 4 chitons turned directly into the light, but did not creep. 0 chiton turned directly into the dark, but did not creep. 1 chiton oriented into the light, then into the dark. 2 chitons gave no response during the time allowed for the ex- periment. THE SENSORY RESPONSES OF CHITON 229 g. One end of the body illuminated. Similar tests were made in which either the anterior or the posterior end of the animal was illuminated. Numerous experiments of this sort were also made in the field. In brief, chitons upon sunlit rocks were found to move into the light when either the anterior or the posterior end had been shaded. The locomotion subsequent to illumina- tion of the posterior half only, by means of bright sunlight, was in a posterior direction; usually, before the chiton had moved completely into the light, it executed a turning movement. The readiness with which backward creeping, for distances of several centimeters, may be resorted to, is worthy of remark. y. Analysis. The predominating movement in these tests is photopositive. The peculiar variations observed remain to be explained. This can be done, as in the case of orientation by general illumination, through consideration of the size and habitat of the individuals and of the character of their periostracum and tegmenta. This analysis agrees in its results with that pre- viously given. The small chitons, less than 2 cm. in length, move into the shadow when their surface is half illuminated. They creep backward with greater readiness than do the large ones. — These experiments show that with light approximately ver- tical in direction the effect of partial illumination of the body is such as to parallel completely the result in orientation to hori- zontal light. 2, Differential sensitivity a. Shading. An attached chiton, undisturbed and at. rest, tends to lift the girdle from the substratum, either along its whole circumference or else in one or more local areas. If a raised portion of the girdle be shaded, there results a relatively quick and smooth lowering of the girdle to the rock or other surface. This response occupies about two seconds; its vigor depends upon the original distance of the girdle from the sub- stratum. After a little time (about ten seconds) the part con- tracted is relaxed to its.original condition. The speed of re- action, and the time for recovery vary considerably in different animals. In some cases a single stimulation resulted in the THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 29, No. 2 230 LESLIE B. AREY AND W. J. CROZIER girdle being closely applied to the substratum for a long time (save at the anterior and posterior ends, involved in respiration). In other instances five to nine shadings and responses were obtained, after which the response grew weaker until it finally disappeared. When the response became very weak, or imper- ceptible to a single shading, two or more shadings in fairly rapid succession were still effective in producing a reaction from the girdle. Associated with depression of the girdle is a pronounced contraction of the gills in animals shaded dorsally. Chitons resting upon their dorsal surface and shaded ventrally gave also a pronounced response. Some individuals were very sensitive, coming from a fully extended condition to complete rolled-up closure as the result of a slight decrease in the illumi- nation; if already partially rolled up, a general shading induced still further, but still incomplete, curling of the body, after which the chiton returned to its original condition. The response became fatigued after seven to sixteen trials. In good reactions of this type, five seconds were required for the curling up of the shell, fifteen seconds for its recovery. To successive shadings the curling-up process becomes more complete; the time required for this effect varies directly with the size of the animal, as seen in the following notes (table 3); some animals are decidedly less reactive in this way than others, however. In Holothuria (Crozier, 714) conditions of similar import are known. The reaction to shading of the ventral surface comprises longitudinal bending, contractive movements of foot and head, inward curving of the edge of the girdle, and contractions of the gills. The time occupied by the response of the ctenidia (‘reac- tion time’) to a single shading at the posterior end of the animal is also proportional to the length of the Chiton (table 3). The response to shading is due to a decrease in the intensity of light in the visible region of wave lengths. The interposition of a glass plate between a chiton and the source of light led to no reaction, and the shading response was easily elicited through a considerable thickness of glass. Glass ray filters transmitting restricted portions of the spectrum were placed between chitons and a source of sunlight (which induced shading responses), THE SENSORY RESPONSES OF CHITON 231 and then the effect of cutting off the colored light was deter- mined. The transmitting regions of the spectrum for the color filters used are given in table 4. These values are of course not very precise. The red filter transmitted light that seemed about twice as bright as that coming through the blue. Nevertheless, the re- actions obtained upon shading chitons through these filters TABLE 3 Sa, TIME REQUIRED TO EFFECT ROLLING UP OF ‘REACTION-TIME’ OF THE SHELL BY REPEATED] POSTERIOR CTENIDIA TO SHADINGS OF THE AN- SINGLE SHADING Number Length TERIOR END cm. seconds seconds 1 3.0 5 2 5.5 10 S 9.0 15 4 ono 1-2 5 6.0 2-3 6 8.9 4-5 TABLE 4 Regions of spectrum transmitted by certain color filters COLOR Te PES Tete rT RANGE hi TR ASy | a Ee & a RE «fc Ra 690-634 56 SELON pe ea i a CR Cie ia 690-605 85 CSneCHe. te ee Sse are ee eee: 589-508 81 ete 3. 6:28 adie 84, ont ks as Se as srs 523-450 + 73 were sufficiently clear cut to be of some use. Responses pro- duced by cutting off the light coming through the green or the blue screen were equal in amplitude to those obtained in sunlight; whereas the red and yellow lights, when occluded, led to quite obviously weaker responses. This is also the case in the shading reactions of the shore barnacle (Crozier, ’15 b, p. 273; the same light filters were used). Chitons of all sizes (ages) and from every type of habitat give 232 LESLIE B. AREY AND W. J. CROZIER precise responses of the character described when the light in- tensity is suddenly reduced. b. Increased light intensity. Chitons from medium to large size, under water in aquaria placed in direct sunlight, give also a response to increased illumination. This response is not evident except in bright sunlight. It consists in a depression of the girdle’ similar to that induced by shading. The response is, however, never of such quickness, vigor, or completeness as that to shad- ing. In smaller chitons, or with larger ones in diffuse light, the reaction to increased illumination (if present at all) is so slight as to escape detection. At its maximal development, it comprises a local depression of the girdle to the substratum, and does not involve a ‘suction reflex’ of the whole animal, such as is induced by a very slight decrease in illumination. The thresh- old of sensitivity for increased illumination is, indeed, very much higher than that for decreased. It should, however, be noted that if a large chiton is shaded dorsally the girdle may not be much elevated again for some time, provided the state of lowered intensity is allowed to continue. On removing the shadow, the girdle is again elevated. If the intensity be slightly decreased the girdle of a large chiton may be lowered rather slowly, so that after two seconds the shadow may be removed before the girdle has been completely depressed; the movement of the girdle may then cease abruptly with the incidence of the more intense light. A similar response is evidenced upon the ventral surface of the animal. If the light be suddenly increased (to direct sunlight) about two seconds subsequent to the beginning of the ‘curling-up’ response induced by a shading, this response sometimes ceases abruptly; the chiton may or may not proceed then to straighten out. The effects of increased illumination are more conspicuous in lamplight at night. Clear indications have been obtained of a specifically higher photic sensitivity at night as compared with daylight hours. A discussion of this matter awaits further investigation. THE SENSORY RESPONSES OF CHITON 233 3. Distribution and nature of photoreceptors a. For illumination. We have described thus far the responses obtained from chitons submitted to photic excitation involving the edge of the girdle, a large portion of the surface, or the whole surface of the animal. ‘There remain to be considered the dis- tribution and variety of the photoreceptors. A chiton from which the girdle has been completely removed is still oriented by light. This fact, taken together with the known histological structure of the sensory organs in the shell plates, suggested that the surface of the shell might contain photosensitive organs. Chitons were rigidly clamped in the way already described in discussing tactile stimulation (p. 200), and allowed to write their contractions kymographically. Tested in air, the shell was found to be sensitive to moderate faradization. When so stimulated, the animal undergoes a pronounced general ‘curling up,’ which ceases when the current is interrupted. KNO,’ > KBr, = et was chosen as the most satisfactory. c. Successive dilutions of several representative substances were employed to stimulate various regions of chiton’s surface, with the object of establishing the respective limiting dilutions effective in excitation. Sea-water was largely used as the sol- vent in these tests, since it is the normal fluid medium for chiton; moreover, as we shall see presently, the surface of the animal is reactive locally to osmotic conditions differing from the normal. KCl: To KCl solutions more concentrated than N/16, made up in sea-water, all of the ventral portions of Chiton are reactive. N/16 All portions respond except the girdle. N/32 No response from the sole of the foot. N/48 Good reactions from the mouth region; the gills some- times fail to react. N/64 Gills fail to respond. N/80 Mouth response good; edge of the foot weak. THE SENSORY RESPONSES OF CHITON 243 N/112 Edge of the foot fails; the palp is sensitive on its edge and bends first toward, then away from, the stimulus. N/160 Faint responses from the mouth region. From one animal, considerably more sensitive than the rest, faint ‘mouth-region responses were also obtained at dilutions N/224 and N/256; this was very exceptional. KOH: The dilutions tabulated are here only approximate, owing to the precipitation of calcium and magnesium hydrates at concentra- tions above 0.012 N + KOH (Haas, ’16). N/10 Reactions produced on all parts except upon the ven- tral surface of the girdle (mantle edge). N/50 to N/125 Same. N/250 to N/375 The edge of the foot is but slightly sensitive. N/500 No responses from the foot; reaction at the mouth just perceptible; gill responses of fair intensity. HCl: N/10 No reaction from the ventral surface of the mantle edge. N/50 to N/425—responses from all parts, save the mantle edge. N/500 ~=No reactions from the foot; very faint responses from the mouth; gills respond more weakly still, but Gil fair sensitivity. Several: organic acids were also tested. At M/10 concentration, in rain water, tannic, malic, lactic and acetic acids produced reactions from all parts, including the mantle edge. M/100 solutions of the same acids made up in sea-water also induced responses, but not from the mantle edge; these acids were compared by noting the ‘recovery time’ of the gills when stimulated with them (as in the case of salts); the measurements obtained were: Tannic, 35 sec. Malic, 20 sec. Lactic, 12 sec. Acetic, 8 sec. Picric acid: made up in sea-water, produced very powerful reactions everywhere, at concentrations above M/150. M/250 ~—_—‘Responses of the same character, but weaker. M/500 to M/750 The edge of the mantle fails. M/1000 The foot still sensitive, especially along its periphery. M/1200 Gills fail; foot and mouth give good reactions. M/1500 Foot fails; mouth alone responds, but smoothly and evenly. 244 LESLIE B. AREY AND W. J. CROZIER Cocaine hydrochloride, atropine sulphate, and nicotine at M/100 concentration in sea-water stimulated very well the mouth, gills, and the sole and edge of the foot. Chloretone M/200 in sea-water also stimulated these parts; but Urethane under the same conditions did not. Curare in saturated sea-water solution, gave rise to good, but not pronounced responses. Urea. A M/10 solution, in rain water, did not lead to any clear response on the part of the foot, and gave but weak responses from mouth and gills. Ethyl and methyl alcohol solutions 10M, in rain-water, evoked reac- tions from all parts save the mantle edge. With the methyl alcohol the responses were slightly stronger, the gill retraction enduring about twice as long (about 20 sec., as compared with 8 to 10 sec. for the ethyl); this may have been due to impurities; the ethyl alcohol was especially pure. Full strength amyl alcohol induced responses as with the two previous types, but considerably stronger movements resulted, It has been of interest to inquire if the surface of Chiton is generally sensitive to sugars or to ‘sweet’ substances as a class. In view of the usually high value of the limiting stimulating concentration for sugars, it is necessary to consider the follow- ing observations in the light of the subsequent experiments upon osmotic excitation: Maltose: solutions M/3 or M/10, in sea-water, produced no clear: responses from any part. A 1M solution in rain-water gave responses from mouth and gills only. Sucrose: 1M, in rain-water, led to no responses from the mouth region, but gave good reactions from the ctenidia. M/3 in sea-water, gave general reactions from all parts. M/2, made by adding 1 volume of 1M in rain-water to 1 volume of sea-water, led to fair responses. M/3, made by adding 1 volume of 1M in rain-water to 2 volumes of sea-water, induced no reactions at all. Lactose; M/2 in rain-water gave strong responses everywhere except at the mantle edge. d. To a variety of more obvious ‘irritants,’ the whole soft ventral surface of Chiton reacts powerfully. Thus, H2Oz2 (3 per cent) induces strong movements in all regions. Ether, chloro- form, carbon bisulphide, aniline oil, and oils of cassia, juniper, cloves, pennyroyal, thyme, bergamot, and origanum, when ap- plied as a drop of the raw substance or in the form of a satu- THE SENSORY RESPONSES OF CHITON 245 rated extract in sea-water, induced good reactions from all parts, including the girdle, which reacted by bending and twisting away. The order of sensitivity of the several parts of the surface was made out to be: gills > head > edge and posterior end of foot > sole of foot > ventral girdle surface > dorsal girdle surface. e. Osmotic excitation of the general ventral surface of Chiton was investigated by means of dilutions of sea-water. _ Three parts sea-water + 1 part rain-water: gave responses only upon the lips. Two parts sea-water-+ 1 part rain-water: no response from the mantle edge, but other parts react weakly. One part sea-water + 1 part rain-water: no response from the mantle edge, good responses from other parts. Rain-water: sole of the foot puckers away strongly; gill response very active; even the girdle reacts, and the shell tends to roll up. Sea-water concentrated by evaporation to one half its original volume gave good reactions on all regions. The same upper limit of osmotic stimulation appears in the effects of glycerin solutions: Glycerin: in sea-water solution, 5M_ Strong reactions everywhere, even from the mantle edge. 3M Mantle edge fails; pitting of the foot away from the fluid is deep and local. 2M Same, but weaker responses. 1M Ctenidia give a faint response, but only occasionally; mouth region still sensitive; very doubtful responses from the foot. M/2_ Faint response from the lips; ctenidia doubtful. 2. The mode of excitation by solutions The observations above detailed require analysis from several points of view. We shall deal first, briefly, with the evidence they contain relative to the method of activation by solutions. a. It is clear that ‘osmotic sensitivity’ is possessed by the soft superficial tissues of chiton. Whether this depends upon the activation of ‘tactile’ or other end organs or upon the stimu- lation of chemoreceptors proper we cannot at first entirely THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 29, NO. 2 246 LESLIE B. AREY AND W. J. CROZIER decide. The comparative distribution of tactile and of osmotie sensitivity is nevertheless suggestive in this connection. We have seen that for stimulation by gentle contact the reactivity of the several regions of chiton’s surface was as follows: head = ctenidia > edge of foot > girdle > sole of foot; while for irritants, such as essential oils, the order of reactiveness for the same parts was: ctenidia > head > edge and posterior end of foot > sole of foot > girdle. To local osmotie disturbances the sensitivity of these areas ap--. peared, on the basis of the experiments with dilutions of sea-water, to be related in the following sequence: head > ctenidia = foot > ventral edge of girdle. So far as these responses go, they indicate that the receptors concerned in osmotic excitation are distinct from those concerned in tactile reactions, from those concerned with responses to irritants, and (as seen in a following section) from those impli- cated in chemical excitation, but the evidence is not conclusive. The sensitivity of the proboscis (‘head’), especially of its peripheral edge, is probably concerned in determining the rela- tive immobility of chitons in exposed places at the period of low tide; the same, to a lesser degree, is perhaps true of the edge of the foot. In active creeping the anterior edge of the proboscis is kept in close contact with the substratum, as shown in figure 14; this organ undergoes ‘spontaneous’ local contractive move- ments, depending for their execution on the pressure of fluids contained in its interior spaces (Heath, ’05 b). b. The osmotic reactivity of Chiton’s soft surfaces is im- portant in connection with the question as to whether sugars are successful as activating agents for this animal. Sea-water of 36.5 per mille salinity (5/8 M) has at 27°C. an osmotic pressure of something more than 25 atmospheres, corresponding to a sucrose solution about 0.8 M. The limits within which various concentrations of sea-water do not stimulate were found to be 4/8 to 8/8 M (for the lips, the other regions being less sensitive) ; M/2 sucrose in 5/16 M sea-water gave fair responses from all a THE SENSORY RESPONSES OF CHITON 247 parts, although 1 M sucrose in rain-water did not, except from the ctenidia. This total concentration (< 6/8 M) is well within the ‘osmotic limit,’ and indicates that sucrose may be mildly efficient in stimulation. The behavior of maltose also shows faint indications (at 1 M in rain-water) of some stimulating capacity. According to Brooks (’16), sucrose penetrates (plant) protoplasm quite readily, and affects permeability after the manner of a monovalent kation, although not with special ra- pidity. The fact that in Holothuria (Crozier, 715 b) no evi- dence was obtained that sucrose could stimulate the integument —although maltose and glycerin apparently did, while in the experiments of Olmsted (17 b) and of Hecht (’18) no sucrose effect was detected apart from that exerted through the osmotic pressure of its solutions—shows that some such factor as the ‘ Fig. 14 Showing the manner in which the proboscis is kept in contact with the substratum during creeping. Anterior end, seen from the side: a, proboscis margin (‘palp’); 8, foot; y, girdle. Arrow shows direction of creeping. time of exposure to the sugar solution (or the concentration of accompanying salts?) may be important in determining whether or not activation occurs. Sensitivity to sugars, even if present, is however, undoubtedly very low on the general soft surface of chiton. With Chromodoris zebra we have found no stimulation induced by maltose or sucrose 1 M in rain-water, applied to various parts of the animal’s surface in small volumes from a pipette, the animal being in sea-water. Similar tests with Balanoglossus (Crozier—unpublished experiments) resulted in no detection of activation by 1 M solutions of sucrose or of maltose. c. Independently of their osmotic sensitivity, the soft surfaces of Chiton are also open to activation by dissolved electrolytes, which are much more powerful as excitants than are non-elec- 248 LESLIE B. AREY AND W. J. CROZIER trolytes of the sugar type. The order of kation stimulating efficiency for the alkali chlorides, . Kee NE: > Li > Na, is that found in the sensory activation of Balanoglossus (Cro- zier, “15 a), Ascidia (Hecht, ’18), Chromodoris (Arey), and other marine animals, and is in accord with the general order of action of these ions upon various protoplasmic processes. The anion order, MY SING: br oi does not agree with that determined by Hecht (’18) for Ascidia; but the methods of experimentation were in the two cases quite different, in the tests with Ascidia the method of limiting effec- tive dilutions being employed and the salts being dissolved in sea-water. For Synaptula (Olmsted, ’17b), the order Cl < Br < I, as in Ascidia, was found, by the same general method. Each of these anion series has its counterpart in other salt actions (Hober, ’14), but why they should be reciprocal is not altogether clear, particularly since the kation orders obtained by these respectively different methods are in agreement. In Chiton, general chemical excitation is primarily an affair of the kation. The minimal concentrations of different electrolytes which are effective in the excitation of different regions of Chiton appear to be as follows (in sea-water solutions) : The three organic acids we employed stimulated according to the following order of efficiency: Malic > lactic > acetic (M/100 in sea-water). This order indicates that for Chiton lactic acid is less efficient as a sensory excitant than malic, which is not the sequence shown (Crozier, 16a) by the earthworm’s reactions, nor in the penetration of tissues by these acids, but does correspond with the respective magnitudes of acid strength. Since the solutions were made up in sea-water, much emphasis cannot be placed on this point. The very general nature of Chiton’s sensitivity, involving excitation bysa great variety of, materials in solution, adheres — nevertheless to! the rules already available in the activation of THE SENSORY RESPONSES OF CHITON 249 numerous protoplasmic processes. The effects of salts, acids, alkalies, anaesthetics, and other substances in evoking reactions are best understood upon the assumption that cells of the external epithelium are acted upon by these substances after the manner of cells in general. Thus, with anaesthetics: chloretone (M/200) gave well-defined responses, whereas urethane (M/200) did not; the anaesthetic effect of these materials follows the same order (Crozier, ’16a). Some form of union with the surface of the epithelial cells is undoubtedly involved in the process of excita- tion, but these experiments cannot be understood on the assump- tion that excitation is determined by an increase in cell perme- ability. The decreased permeability produced by bivalent cations (e.g., Ca) is a specific function of the cation; nevertheless, TABLE 5 The minimal concentrations of various substances effective in the stimulation of Chiton REGION KCl HCl KOH Picric acid RRR ics tals cep Seas es Mone N/160 N/10) (>N/10) M/600+ while CaCl. apparently does not serve as a sensory activator under the conditions of these tests, good reactions are elicited by Ca(NOs;)2 (compare also the case in Balanoglossus, Crozier,’15a). There is no specific parallelism between efficiency in sensory activation and permeability-increasing properties. (Compare, for acids and alkalies, Crozier, ’18 a.) 3. The chemoreceptors a. The ventral parts of Chiton exhibit a chemical sensitivity which is essentially similar to that found for the general integu- ment of other marine invertebrates (table 6). The limiting dilutions of various substances effective in activation of various parts of the animal (table 6) follow an order which is, for each 250 LESLIE B. AREY AND W. J. CROZIER tissue, consistent among different substances. This suggests that—-since the ‘thickness,’ ‘toughness,’ or density of these parts is in the same general order (lips > edge of foot > ctenidia > sole of foot > girdle), with the exception, perhaps, of the ctenidia—we are in reality dealing with a generalized form of sensitivity, the effectiveness in arousing reactions depending upon, 1) the ease with which the surface of the cells locally con- cerned may be actively penetrated by the excitant, and, 2) upon the relative richness with which these parts are respectively TABLE 6 Minimal concentrations of various substances effective in the sensory activation of various animals ANIMAL HCl pees NaCl KCl QUININE PICRIC ACID AUTHORITY Visite vee ee M/1000| M/400|M/50 M/25,000 Parker (’12) Amelurus..... N/20 N/100| N /50 M/150 Parker (712) Amphioxus...| N/500 M/1,250 | Parker (’08) Ascidia?...... N/625 | N/100 N/4_ | M/2,500 Hecht (’18) Balanog- lossus......| N/500 | N/400 N/200) M/1,000 Crozier (un- published) Chiton: 20:27. N/500 | N/500 N/160 M/1,500 Chromodoris.| N/700 | N/200 N/10 M/10,000 | Crozier and Arey (un- published) Synaptula....| N/600 | N/200} N/4; N/40 | M/10,000 Olmsted (17 b) Holothuria...| N/500 | N/500 N/500 Crozier (15 b) 1 “‘Uncorrected’”’ concentrations; Hecht’s paper (’18). innervated. In a previous chapter we have shown that the tactile reactivity of these regions of chiton’s surface follows approximately the order: head (palp, lips), ectenidia > edge of foot > girdle > sole of foot. There is, thus, a distinct inconsistency in the relative sensitivity of these parts to touch and to chemical excitation, which shows that differences in the method of excitation undoubtedly exist THE SENSORY RESPONSES OF CHITON PATS | and that probably the relative richness of general sensory inner- vation is not the sole factor determining the chemical sensitivity of any one region. Further evidence for the distinctness of the chemoreceptors will be considered in the following section. Through the comparison of the effectiveness in stimulation for different materials in Chiton and in other animals (table 6), it will be seen that the integumentary sensitivity of invertebrates corresponds in its general features (limiting effective dilutions) with that of taste in man rather than with the common chemical sense; there are, however, noteworthy differences from the physiology of taste excitation (Parker and Metcalf, ’07; Crozier, 15 b). The fact that the isolated substances considered are, as such, foreign to the daily experience of Chiton, has nothing to do with the information they give concerning the process of excita- tion. That we are not concerned with general ‘pain’ reactions in the Chiton experiments can be shown in this way: The ventral. surface of the girdle of Chiton, although relatively the most insensitive region to chemical excitation, is nevertheless decidedly reactive to touch. This region is excitable by pure anaesthetics, saturated sea-water solutions of essential oils, rain-water, by 10/8 M sea-water, and by 5 M glycerin, but is inexcitable (un- der the conditions of our experiments) by alkaline chlorides (other than KC] at 5/8 M concentration in rain-water), by KCl (in sea-water) more dilute than M/16, by HCl or KOH more dilute than M/10, by picric acid more dilute than M/700, or by ethyl aleohol even in 10 M concentration. Hence it would appear that only under conditions of an excessively heterologous quality may the ventral surface of the girdle be excited by these chemical agents, under such conditions in fact that ‘pain,’ tae- tile, or any other form of sense organ might be activated. This result adds to the conviction that the general chemical sensitivity of Chiton’s soft tissues is distinct from any form of tactile irritability, and isnot consistent with the view that here —as there may be in Balanoglossus (Crozier, 715 a), or in Synap- tula (Olmsted, ’17b)—there are ‘generalized sense organs’ (Nagel). Some indications are afforded that a distinct general chemical sense is adequately represented in Chiton. This con- yAsys LESLIE B. AREY AND W. J. CROZIER clusion may be tested through the attempted physiological isola- tion of chemical and tactile irritability. Such tests are, of course, open to several sources of serious error; the most critical results should be given by cases in which sensory exhaustion to chemical stimulation did not interfere with tactile irritability (Parker, ’08, p. 440). A result of this kind is free from the objec- tion that sensory fatigue may result in heightening the threshold to the more delicately acting forms of activation. Such a result is readily obtainable with Chiton: the ctenidia cease to respond to chemical activation by 5/8 N NaCl after about ten trials at brief intervals; they continue, however, to respond to touch. This finding strengthens the opinion already derived from the distributional study of tactile and chemical activation in chiton. The fact that the reactions induced by these modes of activation are in some cases qualitatively identical, involving similar muscular contractions, is no obstacle to this view. b. In the buccal cavity of chitons cup-shaped organs, with a suggested ‘gustatory function,’ have been described, as well as numerous nerve terminals in the subradular organ. We have nothing on this subject to add to Heath’s (’03) observations, which we can confirm; these observations showed that positive food-taking responses are initiated by the excitation of the ex- ternal surface of the mouth region with the materials upon which chitons feed. We have seen that mouth movements are initiated by chemical activation. It will be of interest to dis- cover to what class of substances chitons react by food-taking responses. Some writers speak of an osphradium situated at the base of each ctenidium in .the chitons (Burne, ’96). Pelseneer (99, p. 13) has described and figured the two erectile ridges on the inner side of the posterior inner ventral border of the girdle (the ‘Jateral lappets of the mantle fold” of Plate, ’97, pl. 2, fig. 15, llp.), immediately caudad of the gills (figs. 7 and 8); he con- siders these structures to be homologous with the osphradia of other mollusks. In some species these protuberances assume the shape of well-defined papillae. They are situated on the distal face of the pallial nerve cords, which supply them with a rich THE SENSORY RESPONSES OF CHITON 253 innervation, and constitute, in Pelseneer’s opinion, special sen- sory regions, each of them being protected by the ventral face of the papilla, which is sometimes spiculose. We have described how the respiratory water current impinges on the dorsal face of each papilla. The conditions are therefore favorable for the sit- uation of an organ ‘‘testing the quality of the water.” The papilla may be significant for egg-laying responses (i.e., in the reception of a stimulus provided by the spermatic fluids), as it seems more prominent in the females. This remains to be tested. Copeland (18) has brought forward good evidence that the osphradium of carnivorous snails is concerned in the recep- tion of chemical excitation by dilute solutions of materials ema- nating from food; his further contention, that this organ is therefore an ‘olfactory organ,’ because the exciting agent is very dilute, is am unnecessary metaphor (cf. Arey, 718 b). VIII. THE NERVOUS SYSTEM AND SENSE ORGANS OF CHITON The main features of the foregoing account may now be briefly summarized. This report makes no claim to completeness; it does lay a solid foundation for further investigation in at least two directions: 1) the phenomena of seeming adaptation in the ethology of chiton, and, 2) the physiology of certain types of irritability. The sensory conditions are here unexpectedly com- plex. The major pathways of nervous transmission are, by contrast, unusually clear and well defined. The manner in which sensory capacities and modes of reaction are involved in the complex determination of natural behavior can be followed in great detail. a. Tactile receptors are absent from the shell surfaces. The “seales’ and ‘hairs’ upon the girdle are important tactile organs. The ctenidia are also sensitive to touch, as are the proboscis, the foot, and the ventral surface of the girdle. The foot is posi- tively thigmotactic to large surfaces, but retracts locally when stimulated by a small surface. ; The tegmental aesthetes are photosensitive; they are activated by light of constant intensity and by sudden decrease in light 254 LESLIE B. AREY AND W. J. CROZIER intensity, not by an increase. The dorsal surface of the girdle (scales) is also sensitive to ight—characteristically to a decrease of light intensity, also to the constant intensity of light, and to a sudden increase in light intensity, provided this intensity be great. The soft ventral surfaces are sensitive to light. The periphery of the girdle is the ventral part most sensitive to shading. The superficial soft tissues of Chiton are open to chemical activation, to stimulation by abnormal osmotic pressures, and by ‘irritants.’ Evidence has been secured, through the study especially of the topographic distribution of the various types of excitability, that tactile receptors, photic receptors, and chemoreceptors are physiologically distinct. ‘There is no clear evidence of sensitivity to heat; that to cold is less doubtful. There is a pronounced tendency for the animal to come to rest in positions avoiding uneven tensions in the musculature. This is responsible for the precise negative geotropism exhibited by Chiton. This mollusk is not sensitive to vibratory mechanical disturbances. b. This brings us to the consideration of one of the most un- settled problems in sensory physiology: In what manner is differ- ential irritability determined? The immediate receptors of ex- citation in metazoans above the sponges are cells which function primarily as detectors and transmitters of disturbances in the energies of the environment. In a general way it is true that all forms of protoplasm are capable of being changed (activated) by light, heat, cold, pressure, chemical agents, and so forth. Con- siderations of this order, which hold also in certain cases upon the quantitative side (as in the action of chemical agents), have been responsible for the view that ‘irritability’ is a generalized property of living matter, best studied in uni- (or non-) cellular organisms. It does not seem probable that this conception can at present be of any great help; in spirit it is deductive, whereas the manifestations of irritability (e.g., in the diversified taste receptors of the human tongue) are manifold, specific, and must be investigated in a more purely inductive manner. Neverthe- THE SENSORY RESPONSES OF CHITON 255 less, it is true that in those instances where photoreceptors, for example, may be isolated and distinguished, we frequently find that the sensory elements are removed from the external surface of the animal, protected from the action of environmental chemi- cal disturbances and deforming contacts. From this standpoint one of the factors determining differential receptivity is to be found in the degree of anatomical isolation of the receptor; another, in the morphology of the sensory cell or organ, as in the development of distal projections.. These factors of form and position undoubtedly facilitate the respective operation of dif- ferent qualities of stimulation, and to that extent determine the functioning of differential irritability. But the problem is not in this way wholly solved. The evidence for ‘generalized sense organs’ is in some cases good, though perhaps not final. Even on the sole of the foot of Chiton physiological evidence of sen- sory separateness for photo-, tacto-, and chemo-reception is available. The skin of Amphioxus is fully as sensitive as that of other marine animals to touch and to chemical influences {Parker, ’08), but is insensitive to light. The additional factor is probably found in the possession by certain receptor cells of special substances which enter into excitation reactions. Even if in some cases it could be shown that the epithelial cells of an animal were open to sensory activation by a variety of means, it would not lead to the view that a ‘universal’ type of sense organ (Nagel) was that first developed in evolution, re- ceptors of special kinds being by some obscure metaphysical proc- ess subsequently derived from it. In coelenterates, so far as we know (Parker, ’17 a), tactile receptors and chemoreceptors are organically distinct. c. The reactions of Chiton to local stimulation are of a character consistent with the known distribution of the central nervous system. At the sides of the body, those parts innervated by the pallial strands are conspicuously homolateral in their responses. The coérdination of the pedal musculature for the production of locomotor waves depends upon a mechanism locally contained, and apparently upon the integrity of the pedal cross-connectives. The coérdination of the gill movements on one side of the body 256 LESLIE B. AREY AND W. J. CROZIER depends upon the intact condition of the pallial nerve strand on that side. The responses of isolated portions of an animal sec- tioned transversely are such as to show the absence of any strong centralization. This is in agreement with the known occurrence of ganglion cells throughout the whole length of the nerve strands. In Chiton nervous centralization is relatively at an incipient stage. d. Alteration in the behavior of Chiton toward light with ad- vancing age of the animal is the primary variable determining the exhibition of a very complex series of environmental inter- relations. The young Chiton is photonegative, the old Chiton photopositive, to sunlight. Chitons of intermediate age are positive to weak light, negative to strong. Photic orientation is direct, and is determined by the constant intensity of light, not by change of intensity. The progressive alteration in the sense of phototropism is determined by the erosive destruction of the photosensitive aesthetes, conditioning in older Chitons a lower specific stimulating power of the light. The erosion of-the shell is in turn produced, in part, by, 1) normal growth effects; 2) the activity of organisms settling upon the shell plates. ¢ The homochromic coloration of Chiton is determined by the nature of its algal food and by the organisms living upon its dorsal surface. The older chitons are relatively stationary; therefore specific local environmental influences have oppor- tunity to affect the appearance of the chitons. The animals associate in groups, commonly of a certain average size and con- taining numbers of both sexes. Certain seemingly ‘adaptive’ consequences may reasonably be attributed to this mode of occurrence. A homochromically colored isopod is characteristi- cally associated with Chiton tuberculatus. These and other harmonious correlations, of which mention is made in the body of this report, follow automatically in the wake of the changing phototropism of Chiton. The animal’s habits determine the environment in which it dwells. The pre- cise and intricate bionomic correlations here briefly mentioned are an automatic consequence of its modes of reaction. Dyer Island, Bermuda, May, 1918 THE SENSORY RESPONSES OF CHITON 25¢ LITERATURE CITED Argry, L. B. 1918a The sensory behavior of Chiton (abstr.). Anat. Rec., vol. 14, p. 95. 1918 b The multiple sensory activities of the so-called rhinophore of nudibranchs. Amer. Jour. Physiol., vol. 46, pp. 526-532. Baunackge, W. 1914 Studien zur Frage nach der Statocystenfunktion. II. Noch einmal die Geotaxis der Mollusken. Biol, Centralbl., Bd. 34, S. 371-385; 479-523. BuumricyH, J. 1891 Das Integument der Chitonen. Zeit. f. wiss. Zool., Bd. 52, S. 404-476, Taf. 23-30. Cited in Schneider’s Lehrbuch, 1902. Brooks, 8. C. 1916 Hi} + Mit, A DEMO o 5 Fb Wotan at ” ibe” ane ME 6 oils OF Se Lshetsean eb itatat, Nhe Opieicgicto: iWin Jey coe oft MY, pte Ohi seeg l ger ant: AND: ply ; Le pir ry Bs) 18 4A ) big, SB Pr todas “nei Job cas f 140) Raeeen see a Resumen por los autores, W. J. Crozier y Leslie B. Arey. Estacion Biol6gica de Bermuda y Escuela Médica de la Universidad Del Noroeste, Chicago. Reacciones sensoriales de Chromodoris zebra. Los mecanismos receptores diferenciados que originan reac- ciones bajo la influencia de la estimulacién tactil, quimica y luminosa, y bajo la accién de la intensidad constante de la luz y tal vez del calor, inducen respuestas locales por medio de las redes nerviosas periféricas no sindpticas, que, en las branquias plumosas y tal vez en otras partes, exhiben una decidida polari- zacion. Las reacciones de las partes distantes del sitio de acti- vacién local suponen la transmisién central, ganglionar y sin- Aptica. Los nudibranquios son positivamente fototrépicos, prob- ablemente por intermedio de los ojos. El collar branquial es también sensitivo a la luz produciendo la extensién. de las branquias plumosas. Estas ultimas reaccionan de un modo variable a los cambios de intensidad luminosa. Los individuos maduros sexualmente son geotrépicos negativos. Una tem- peratura de 31° a 32°C. induce reacciones negativas. Los ‘“‘rindforos’”’ son 6rganos directivos del reotropismo negativo en corrientes fuertes. Los animales no responden a las vibraciones transmitidas por el agua. Las respuestas quimiotrépicas son importantes para la conjugacién. La locomocién es principal- mente muscular y se verifica por los bordes laterales del pié, que actia localmente como ventosa. El estereotropismo posi- tivo del extremo anterior del pié es la causa de la vuelta a la posicién normal. Translation by José F. Nonidez Carnegie Institution of Washington AUTHOR’S ABSTRACT OF THIS PAPER ISSURD BY THE BIBLIOGRAPHIC SERVICE, AuGusT ll SENSORY REACTIONS OF CHROMODORIS ZEBRA! W. J. CROZIER Bermuda Biological Station LESLIE B. AREY Northwestern University Medical School EIGHT FIGURES CONTENTS RU PUD is oo Note icc’ eae 8 cles eb 2 PASS AMIS en shoe o 9 bye gan 261 MimeViechanicolvexcitwolOma ns. soto corss cele accu oe nti d we cc doce cles sisal etme IMePactile stimulation. (ees. oe hee eth ee bs. A 267 Peoiiehonig peotroepiemy.: .i97).).. i4)..J2ikiics). G0). i iivs aod eS 2a Sah a AGEN CEPERISEEN a 8 cos TAT 9 ot KP 8S eRe yh O bo Sicia's,s la oie sxe tl epee Bag bee Ame N CTV ISMEE LUIS Merce ayia achat save aire idee oie ake Gis 0 Rass di o.0le cage ag ae MEM SEN MTERUIEIGH vo ek ete te hee ok a eee weeds ce oti te Soe ne Pee Eger hbchalsnis)- (sd piv dese Pett te MQ ERT E- bap hoe ess tgs) gigas 292 STRIVE 4 fol A ON co hci cot Wee leis oie, ~ sin vadin ow us ke eee MP ernal Gxemation.......0. WE te ccc r ters cocoon ee de ete ascsacde os gah Od he heiens exene tions i). 0s ares OU, A Oe, SW. Alls OD). A Min Summary)... lb. cts bss cues mae ack ieiee. aie? lo. Slacks eal AERA LTIPG ACICEEY = Tet ROW he hoe Ace S cecin 3 om g200'a dyes fl cus pZape’> piadaya s Siehcgpuaene 308 I. INTRODUCTORY With the objects of presenting data valuable for the compara- tive study of ‘animal behavior’ and of laying a foundation for the intelligent pursuit of certain inquiries in sensory physiology, this article aims to record and briefly to discuss some aspects of the sensory responses of the nudibranch Chromodoris zebra Heilprin.2. In 1884 Bergh wrote that almost nothing was known 1 Contributions from the Bermuda Biological Station for Research, no. 111, and from the Anatomical Laboratory of the Northwestern University Medical School, no. 68. 2 Data upon which this paper is based were in part secured by L. B. A. dur- ing the summer of 1914. These observations have been extended and the paper written by W. J. C. 261 262 W. J. CROZIER AND LESLIE B. AREY about the activities of the large exotic nudibranchs, and very little has since then been added to the subject. We shall deal with questions of natural history and behavior only in an inci- dental way, paying attention specifically to the manner in which Chromodoris responds when activated by various stimulating agents. It was desirable to undertake a study of this kind be- cause C. zebra has already provided material, of a very exceptional character, for the treatment of some questions in which sensory phenomena are implicated (Crozier, ’15a, ’16a, ’16b, 717d, 18d). A knowledge of the sensory capacities and modes of re- sponse in Chromodoris affords, also, some information as to the comparative physiology of the nervous system in mollusks, about which, particularly in nudibranchs, very little is known. Chromodoris zebra (fig. 1) is a large species very common at Bermuda,? with the form typical of the genus. An account of its morphology will be found in the papers of Smallwood (’10) and of Smallwood and Clark (’12). The body is elongated, es- pecially in creeping, and measures up to 18 em. in length. The animal as a whole is very soft and contractile, and becomes read- ily bent or twisted under appropriate conditions. Throughout the year individuals of a variety of sizes are to be had by dredg- ing in depths down to 10 fathoms. From September to June great shoals of them, numbering thousands in all, crowd up at intervals into shallow water (Crozier, ’°17b). They become notably concentrated in certain shallow mangrove creeks con- nected with Great Sound. The cycle of events which determines the shoreward flocking has not yet been fully established. Its coloration makes this nudibranch easy to distinguish upon the bottom (Crozier, 716 b), and the migratory movements of the species, owing to its lack of concealing behavior, may to some extent be followed in the field. The animal moves with a smooth, even, gliding motion over rock surfaces or on the muddy bottom, the entire surface of the 5 As with many other marine forms found at Bermuda, it is probable that the range of C. zebra is quite extensive, although so far it has been reported only from Bermuda (Heilprin, ’89; Smallwood, ’10). I am informed by Prof. W. H. Longley that a few individuals were obtained by his collectors, in seining on grass-flats at Tortugas. W. J. C. SENSORY REACTIONS OF CHROMODORIS ZEBRA 263 foot being applied to the substratum. The whole body can, however, be supported by the use of a small part only of the foot; thus C. zebra has sometimes been observed to creep over the edge of a submerged rock, the body of the animal projecting horizontally beyond the edge, or its anterior part being even sharply elevated, until only a centimeter or so of the posterior region of the foot served as a hold-fast. Chromodoris can also swim attached to the surface film of quiet water, but has not - been observed to do so in nature. (This behavior has been ob- served in another nudibranch, Facelina goslingi, at the time of its reproductive activity.) Fig. 1 Outline figure of Chromodoris zebra; dorsolateral view from the right side (after Crozier, ’18 d). Especially in the mangrove creeks already spoken of, C. zebra is: found in localities where eel-grass grows in great profusion. It creeps up the blades of eel-grass and is often found at the very tip. In connection with this habit it is important that the foot, although perfectly flat and without a trace of median division, when in contact with a flat surface, exhibits a most pronounced tendency to fold together, so that the lateral halves of the whole surface of the foot are in contact. This occurs whenever the whole or a part of the foot is removed from a substratum, and makes it possible for the nudibranch to climb the flat-bladed eel-grass. In C. roseapicta no such behavior of the foot is seen, and this species lives under stones along the shore, never coming into contact with eel-grass. 264 W. J. CROZIER AND LESLIE B. AREY According to Vlés (’07), no pedal waves are observable upon the foot of AXolis, at least in the form of color differences, al- though the direction of the pedal wave was said by him to be direct, on the basis of the rippling movement seen at the edge of the foot during creeping. In Doris, also, the pedal wave was said by Vlés to be direct. In Tectibranchs the direction of the pedal wave, well defined in this case, is retrograde (Parker, ’11, p. 156). The locomotion of Chromodoris is accomplished in the form of a smooth, planarian-like glide. No macroscopic waves have ever been seen on the foot, even when the animal was swimming attached to the surface film and therefore in a very favorable position for observation. Some unevenness of move- ment may usually be detected along the lateral margins of the foot, and, when fixed to the surface film, where locomotion is very slow, gentle ‘billowing’ movements can usually be detected. Hence it is possible, as Parker (’11) has suggested, that in this case the muscular activity of locomotion is not codrdinated in wave form, but is arhythmic. The lateral surfaces of the body and the surface of the foot are, however, richly ciliated. Much slime is laid down by the foot in creeping. The possibility, then, cannot be entirely excluded that progression is in part at least ciliary, especially since the cilia beat in the appropriate direction, namely from anterior to posterior. Small pieces cut from the mantle or from the foot continue to move for several days in a fixed direction about the bottom of a dish, owing to the beat of the cilia which they bear. According to Olmsted (17), cilia are the means of locomotion in Marginella, Haminea, and Bulla. Local muscular activity, however, shares in loco- motion, for the extreme lateral margins of the foot are the re- gions exerting suction when the foot serves as a hold-fast. This may be seen in the cup-like puckerings along the edge of the foot when ‘swimming’ along the surface film. These regions are not sticky, as if with slime, yet to a sufficiently large surface the out- line of the foot, save at its anterior end, becomes so firmly at- tached that it is with difficulty pushed loose at any given point, although it can be loosened instantly when the animal begins — to creep. Local movements of this muscular rim of the pedal SENSORY REACTIONS OF CHROMODORIS ZEBRA 265 surface are visible during creeping. It is possible that the margin of the foot is alone concerned in active progression,’ at least in so far as this is independent of cilia. With an indi- vidual attached to the vertical wall of a glass aquarium, the axis of the animal being horizontal, the lower edge of the foot some- times becomes freed from contact with the glass, the creature then being suspended by the attachment of the upper margin of the foot; the median furrow is then plainly visible on the foot surface. On such a free lower pedal margin two or three dis- tinct waves may be made out at one time. These waves, retro- grade in direction (as in Tectibranchs), are confined to the outer margin of the foot. Similarly, in specimens resting in the angle made by the vertical wall of the aquarium with its bottom, the foot may be free anteriorly, being attached only at the posterior end. In these cases two or three waves were observed on each side of the foot at one time, the waves on the two sides having neither definitely ‘opposite’ nor ‘alternate’ relations, but ap- pearing to be quite independent of each other. The fact that copious slime secretion occurs along the margin of the pedal surface in animals anaesthetized with MgSO, or with chlore- tone enables a test to be made of the adhesive properties of this slime; it is not sufficiently sticky to cause the attachment of the foot to a glass rod, although in the unanaesthetized nudibranch such attachment is readily demonstrated. These several lines of evidence agree in pointing to the essentially muscular, non- ciliary, nature of the act of creeping in Chromodoris. It should be noted that the marginal pedal waves are retrograde, not direct, as Vlés states for Doris. The direction of progression is always anterior. Onasmooth - surface the rate of creeping in active, large animals (at 27°C.) is about 1 cm. in five to seven seconds; when swimming on the surface film, about half this rate. Although we are not concerned to give an exhaustive account of the effector systems of Chromodoris, mention may be made of 4 In an unidentified species of Leptoplana I have observed locomotion essen- tially of this kind, obviously muscular, in which the outer edge of the body was the only part in contact with the substratum. W. J. C. 266 W. J. CROZIER AND LESLIE B. AREY ‘the fact that slime glands are important for the production of a slippery condition of the whole surface of the animal; and that repugnatorial glands, in part at least under nervous control (Crozier, 16 b; 17 a), are also involved in the creature’s effector equipment. In addition to ciliary activity and several types of gland secretion, muscular movements of some variety are evi- denced. Chromodoris has no hard supporting skeleton; its movements depend conspicuously upon the distribution, under muscular pressure, of the body fluids, and comprise: bending movements, twistings, contraction and extension of the whole body, and of its projecting outgrowths—tentacles, ‘rhinophores,’ and gills; protrusion and retraction of the proboscis, of the genital papilla, and of the oviduct; rhythmic contractions of the extended oviduct during egg laying, as well as local contractions of practically every part of the animal’s surface. C. zebra is functionally hermaphroditic, and reproduces at all times of the year (Smallwood, 710; Crozier, °17b). The ani- mals employed in this work were for the most part collected in Fairyland Creek, near the laboratory of the Bermuda Biological Station, where a practically unlimited supply of material was available during spring and early summer, when these experi- ments were chiefly performed. This nudibranch is easily main- tained in aquaria (Crozier, ’18 d), contrary to Smallwood’s (710) belief, but freshly collected individuals were almost always used. The largest specimens collected in late spring are the least viable; at a length of 16 to 18 em. Chromodoris zebra undergoes natural death. It would be of interest to determine the growth rate of the animal, but this cannot as yet be attempted. A more de- tailed account of the natural history of C. zebra will be found in reports dealing with the phenomena of its breeding habits (Crozier, 717d, 718d) and of the coloration of the species (in course of publication; Crozier, ’16 b). SENSORY REACTIONS OF CHROMODORIS ZEBRA 267 II. MECHANICAL EXCITATION 1. Tactile stimulation The oral tentacles are very sensitive to touch, especially at the tip. When the tip alone is very lightly touched with a fine glass hair, it is contracted and slightly introverted. To slightly more intense stimulation, however, and always when touched at the side or at the base, the tentacle is introverted at the base after the fashion of a glove-finger. To unilateral stimulation of one tentacle, even to sharp and repeated touches, that tentacle alone responds. But after a tentacle has been completely, ex- cessively, contracted, strong continued local mechanical stimula- tion of it (while remaining retracted) causes the opposite tentacle to be retracted. In this case the whole head region is more or less contracted, and it may be that the skin at the base of the retracted tentacle must be stimulated in order to result in a spreading of the response to the other one. The ‘rhinophore’ on the same side with a stimulated tentacle usually contracts slightly, by a twitch of the muscles at its base, synchronously with the activated tentacle itself. If the stimu- lation is originally strong or if it is repeated, the opposite ‘rhino- phore’ also responds, but usually to a less degree. Stimulation of a tentacle also involves response from the head region gener- ally, causing it to retract; at the same time the buccal veil is drawn down so as to cover the whole mouth region, including the anterior edge of the foot. The anterior end of the body is under these circumstances contracted more strongly on the stimu- lated side, and after reextension the whole body is usually caused to bend in the opposite direction, away from the side originally stimulated. If the anterior part of the foot should not be in contact with the substratum, it also contracts, on the homolateral side, when a tentacle is touched. This general form of reaction is the common response when the nudibranch is stimulated any- where with sufficient severity. Further evidence for the neuro- muscular unity of the head region will be found in what follows. To a single light touch upon a tentacle, the general head re- sponse is very slight, but is nevertheless evident. The full head 268 W. J. CROZIER AND LESLIE B. AREY reaction involves a deep insinking of the dorsum at the level of the ‘eye spots.’ This form of ‘reflex’ is seen also in other Dorids. (C. roseapicta, Lamellidoris, etc.). The tentacles do not easily become exhausted. After ten to fifteen successive applications of a glass rod, a tentacle is still reactive to light touch, although the resulting contraction is not. so complete. The ‘rhinophore’ of Chromodoris (Arey, ’17, 718) is a some- what complex structure. Its extreme distal tip is usually pale blue or white, the rest of the organ deep blue or purple. On either side of an anterior, median line, which is plain and smooth, the ‘rhinophore’ bears a series of twenty-eight projecting leaves. To a light touch at the extreme tip, a ‘rhinophore’ responds by partial retraction; the anterior, unmodified, median line is less. sensitive; the posterior and lateral surfaces are the most sensitive. Even moderate intensities of activation cause a ‘rhinophore’ to be retracted within its collar, suddenly and completely, then reextended, more slowly. In animals of average size (8 to 12 em. long) the ‘explosive’ type of response is the result of even light tactile stimulation. To a very delicate touch on the lat- eral or posterior face the retraction is only partial. The ‘rhino- phore’ is itself contractile, longitudinal contraction occurring locally along its length when lightly touched, and it is pulled within its collar by the operation of basilar muscles within the ‘rhinophoral’ pocket. The retraction of a ‘rhinophore’ involves. the subsequent sphincter-like closure of its basal collar. A ‘rhinophore’ is not easily exhausted. When approximately the same spot on the side of the organ is touched fifty times at. ten-second intervals, the amplitude of contraction decreases, but the ‘rhinophore’ is still reactive. Stimulation of a ‘rhinophore,’ even repeated stimulation, does: not influence the homolateral tentacle. The ‘rhinophore’ reac- tion is itself characteristically homolateral, as was seen particu- larly in the case of abnormal variates in which the ‘rhinophores’ were found naturally fused in varying degrees (Crozier, ’17 e). A sharp tap administered to one ‘rhinophore’ results in the partial, less complete contraction of its mate, and also of the SENSORY REACTIONS OF CHROMODORIS ZEBRA 269 gill crown. Less vigorous stimulation has no effect on gill contraction. The gill plumes are individually sensitive, and react sepa- rately to slight stimulation. More violent activation (e.g., a sharp tap or gentle pinching) of a single plume spreads through the other plumes, according to its intensity. A single plume presents a smooth, narrow, distally tapering outer edge, a simi- lar inner edge, and running between them two broad blade-like faces from which jut out the thin gill plates. Tactile excitation of the outer or of the inner faces leads to similar reactions of about equal magnitudes. The gill-bearing faces of the plume are less sensitive; frequently, an individual gill may be bent back and forth without eliciting a response. Presumably this occurs naturally in tidal currents, and during the movements of the gill crown as a whole. To a light touch, the common form of response is a local constriction of the plume, usually not equal on the two sides, accompanied by local longitudinal contraction, so that a slight swaying movement of the plume results. The plume as a whole may or may not be pulled down at its base. To stronger stimu- lation the characteristic response involves the following events: local constriction, spreading distally from the point of activation, leading to the collapse and ‘shriveling’ of the plume distal to the point of activation; this is succeeded by the retraction of the plume through the traction of muscles not intrinsic to the plume itself, but situated in the basal tissue of the gill crown. Still stronger activation leads to longtitudinal contraction of the gill plume, both distally and proximally to the site of touching. The reaction of the plume distal to the point of activation is nicely -demonstrated by plumes which have acquired a branching or dichotomously divided form (Crozier, *17e). If one of the branches of such a plume be touched on the side, this branch alone, and only distal to the stimulated point, contracts, unless the stimulation be too strong. The type of polarity evidenced in the reactions of a gill plume is curiously akin to that seen in the responses of an actinian tentacle under similar conditions of local activation (Rand, ’09, 970 W. J. CROZIER AND LESLIE B. AREY 15; Parker, ’17)—with this important difference: in the actinian tentacle it is the part proximal to the point of activation which contracts. The polarization of the gill plume, which is a neuro- muscular matter since it disappears under magnesium sulphate anaesthesia, is further seen in the fact that the distal tip of a plume, when touched, gives rise to only a slight longitudinal contraction in the immediate region of the tip, although the basal contraction may lead to the retraction of the plume as a whole. Bionomically, the significance of the difference between the neuromuscular polarizations within the actinian tentacle, on the one hand, and the gill plume of Chromodoris on the other, lies in the fact that the actinian tentacle carries food to the animal’s lips, hence the part between the disk and the point of excitation is shortened; whereas the mode of retraction of the gill plume probably saves it somewhat from being bitten by fishes. The gill plumes are bitten at by fishes, and there is evi- dence to show that some of the structural variations which they ° present (Smallwood, ’10; Crozier, ’17 e) originate as the result of injury. The basal contraction of a plume spreads to other plumes in proportion to the intensity of the stimulus and to the nearness of its application to the base of the gill-plume; this is the reaction which is responsible for the retraction of the whole gill crown. Contraction at the base of a gill induces collapse of the whole plume. Any desired degree of contraction of a single plume or of the whole set may be induced by grading the intensity of the tactile stimulus. Also by stimulating single plumes weakly and one at a time, as many plumes as may be desired can be caused to contract; e.g., all but one may be made to contract. Under’ slightly stronger activation, especially in the case of the more anterodorsal gill plumes, it can be demonstrated that the suc- cessive stimulation of two adjacent gill plumes is much more effective for the production of retraction of the whole gill crown than is the equivalent stimulation of any single plume. Thus, if two adjacent plumes are touched in quick succession or simul- taneously the whole crown is retracted more or less completely; SENSORY REACTIONS OF CHROMODORIS ZEBRA 271 whereas if either one of them is itself touched twice in this way, even if on the opposite faces, it alone reacts, though more vigor- ously than to a single touch. It is best to use large animals, with gill plumes widely extended, in testing this point, as other- wise the plumes may stimulate each other through mutual con- tact, a single slight stimulation then sometimes inducing rela- tively complete gill retraction. A single gill plume will react as many as twenty-five times in succession, when repeatedly touched at its tip, without mechanically involving another plume and without leading to contraction of the gill crown. It would seem that the contraction of the whole gill crown when the gills are touched is a secondary phenomenon, depending upon the extent of the disturbance produced in the basal tissue as the result of the individual gill plume activation. Two gill plumes separated by two or three intervening members of the series do not, when touched in succession, lead to retraction of the whole gill crown. The base of the branchial apparatus is also sensitive to touch. Stimulation of the brim of the anus, within the circlet of plumes, causes retraction of the gill plumes immediately adjacent to the stimulated site. To stronger stimulation of the anal brim, more and more of the plumes become involved in contraction. Acti- vation of the anal brim is more efficacious in causing retraction of the gill plumes than is stimulation of the plumes themselves. The surface of the branchial organ outside the base of the gill plumes is very sensitive to touch; a slight stimulation induces complete retraction of the plumes. It is difficult to fatigue the gill reaction. If the plumes are forced, through adequate stimulation, repeatedly to contract completely within the branchial collar, and the time required for subsequent expansion is noted, it is found that the time first shortens, then lengthens, as in this example (times in seconds): 85, 65, 45, 30, 23, 17, 15, 17, 23, 22, 23, 26, 30, 36, 34, 50, 55. The relative rate of exhaustion of the phases of the gill response is seen in the following experiment: 272 W. J. CROZIER AND LESLIE B. AREY The anal brim was stimulated by touching it with a glass rod.. After twenty-four applications, at successive intervals, the gill plumes still responded by contraction; although after the first twelve responses it demanded longer and harder stimulation to accomplish complete retraction of the gill crown. In another animal the same form of stimulation was used, but the stimuli were supplied in groups of three successive touches. The indi- vidual touches were throughout of about the same force. lst application of three touches—complete retraction within collar. 2nd application of three touches—incomplete retraction; not drawn within collar. 3rd application of three touches—about half the ‘normal’ response. Ath application of three touches—slight contraction only of the gills. 5th application of three touches—gills moved, but did not contract longitudinally. 6th application of three touches—gills moved slightly. The integument of the head region is very reactive to touch. The anterior edge of the mantle fold (buccal veil) seems to act as the chief or immediately receptive part when the nudibranch during creeping meets obstacles raised above the general level of the substratum (fig. 2). A single stimulation of this part has a slight effect on the tentacles; repeated light touches cause first the homolateral tentacle, then the opposite one, to be fully retracted—or, if the median region of the mantle edge be touched, both simultaneously. Thé reaction of the ‘rhinophores’ when the buccal veil is touched is very marked. It tends to be homolateral, as in the case of the tentacles, but is much more pronounced. The ‘rhinophores’ react as fully and as quickly as when they them- selves are directly touched, although in the latter case the re- sponse is less easily fatigued. In some individuals this response of the ‘rhinophores’ is elicited by tactile activation of the dorsal integument as far back as the anterior level of the ‘eye-spots;’ in others, only as far back as the level of the ‘rhinophores’ them- selves. In the region behind this level the effect on the ‘rhino- phore’ becomes suddenly much weaker, and a response from the gill crown comes in, increasing in amplitude as places nearer the branchial collar are touched. The dorsum of Chromodoris is soft, flexible, and very easily stretched and distorted. Therefore the delimitation of the re- SENSORY REACTIONS OF CHROMODORIS ZEBRA We ceptive field for the ‘rhinophore’ reaction is best established with the aid of mild faradic stimulation. The electrodes can be placed in position without inducing local response and without leading to the reaction of distant parts, owing to mechanical deformation of the body wall. The receptive field of the ‘rhino- phore’ reaction, made out in this way, agrees precisely with that already given. It is much more clearly defined than in the case of the gill response, as we have indicated above. Stimulation of a ‘rhinophore’ does not ordinarily lead to a reaction from the gill plumes, unless the ‘rhinophore’ be sharply struck or pinched. The ‘rhinophoral’ collar reacts by constriction, sphincter-wise, when touched lightly, but first of all the ‘rhinophore’ is re- tracted. To a very light touch the stimulated part of the edge Fig. 2 Outline lateral view of C. zebra (anterior end) during active creep- ing, showing the manner in which the lips, tentacles, anterior part of the foot, -and the buccal veil are related to the substratum. of the collar contracts locally without inducing the sphincter-like constriction and without leading to movement of the ‘rhino- phore.’ When the ‘rhinophore’ is retracted as the result of being itself stimulated, the collar contracts over it. The branchial collar behaves in a precisely similar way. The projecting mantle-margin, if touched dorsally or ventrally, is locally depressed ventralward. This is also true of the caudal veil which carries on its ventral side the conspicuous mantle glands (Crozier, 717 a). When the animal is generally disturbed by being handled, or is from any other cause much contracted, the margin of the mantle along-the sides of the body is thrown into prominence laterally, owing to the forcing of fluid into its internal spaces; whereas the caudal veil continues to be bent ventralward, unless the point of application of the stimulating 274 W. J. CROZIER AND LESLIE B. AREY agent be on the caudal extremity of the foot or on the dorsal surface of the caudal veil itself (Crozier, 17a). If strongly stimulated at one point on its dorsal surface, the projecting mantle margin is turned sharply dorsalward. Stimulation of the caudal veil leads to pronounced local contraction of the body musculature at that level, accompanied by contraction of the projecting ‘tail’ of the foot (analogous to the ‘head response’ at the anterior end of the animal). The peculiarities in the be- havior of the caudal veil are related to the functioning of the large mantle glands, of which mention has already been made. The peripheral edge of the foot is not very reactive to touch, excepting at its truncated anterior end. There, especially at the faintly projecting corners, corresponding to the place of origin of the ‘foot tentacles’ in other nudibranchs, it is very sensitive, a single touch causing head and tentacles to retract. Here again the response tends strongly to be homolateral. Stimulation of the lateral edge of the foot causes the foot to fold together lengthwise. This is also true of the sole of the foot (e.g., of an individual swimming at the surface film) ; a very distinct line, about 1 mm. broad, which owes its appearance to the vertical, dorsalward contraction of the foot muscles, makes itself apparent even when the foot may not fold together. Touching the margin of the ‘tail’ of the foot induces deep local puckering, and that part of the foot itself is pulled forward and upward. The pharynx, when extended, is found very sensitive to touch. The lips of the fully protruded proboscis are relatively insensitive, but the very faintest touch upon its lateral wall leads to violent retraction of the whole head region; the pharynx is itself also momentarily introverted. Here, again, the reac- tion tends to be homolateral, as in all the responses of the head region. The oral area in animals with retracted proboscis is not so sensitive to touch as are the tentacles. In normal creep- ing it usually happens (fig. 2) that the proboscis is partially everted, so that a portion of its surface, exceedingly sensitive to touch, is brought into immediate relations with the surface over which the animal crawls. One-sided stimulation of the pharynx causes the homolateral ‘rhinophore’ to retract; stimulation of a SENSORY REACTIONS OF CHROMODORIS ZEBRA 275 ‘rhinophore’ does not affect the pharynx, unless it be repeated several times. The genital papilla and the mouth of the oviduct, when everted, react locally to light touch, always contracting away from the point touched. They induce no general reactions of the whole body. Chromodoris is relatively insensitive to vibrational stimuli. Continued tapping of the wall of a thin glass dish containing the nudibranchs may cause near-by resting individuals to begin to move, owing apparently to tactile irritation of the foot. No effect whatever is produced on creeping individuals, and no reactions are given under any circumstances by either ‘rhino- phores’ or gill plumes. In nature the gills, ‘rhinophores,’ and mantle edge are moved about by tidal currents, and the body is by the same means caused to sway from side to side, with- out leading to noteworthy response, save in the case of the ‘rhinophores’ 2. Righting: geotropism When the foot is removed from contact with a substrate, Chromodoris contracts to one-third or one-half its normal length, then subsequently becomes extended. The foot folds together longitudinally. The head end, after the preliminary re-extension of the body, is twisted on the long axis until the anterior part of the foot can be attached. The body is quite flexible, and in righting it may be twisted 180° or even 270° about its long axis. The foot is attached progressively, beginning at the anterior end; nevertheless, as already described, when once attached the poste- rior end is so well fixed to the substrate that the animal may be fully supported by this end alone. The process of righting occupies about one minute, ten to twenty seconds of this time being taken up with the twisting of the body in the effort to secure contact by means of the anterior part of the foot. The anterior edge of the foot is the only part which becomes spontaneously attached in this way. Observa- tions on many animals in the field, as well as in the laboratory, have shown that there is no pronounced tendency for C. zebra THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 29, NO. 2 276 W. J. CROZIER AND LESLIE B. AREY to maintain an upright position of the body, with the foot ven- tral. The righting behavior is probably due merely to the stereoptropism of the foot, especially at its anterior extremity. A Chromodoris placed on its back will become attached to a glass plate appropriately held in contact with the foot, even though the body remain upside down. ‘This is best tested in C. roseapicta, where the foot does not tend to become folded to- gether. The surface of the foot must be in contact with some- - thing. When removed from a substratum the foot folds to- gether longitudinally so that the lateral halves of its surface are in mutual contact. The-origin of the twisting movements is probably found in the mechanical excitations of the skin induced by placing the nudibranch on its side or back; the anterior edge of the foot also exhibits writhing movements when the animal is so stimulated, but upon getting into contact with a solid surface it reacts positively, by attachment and slime secretion, and righting is begun. Repeated tests have been made to discover good evidence of geotropic orientation in Chromodoris, but without a decisive result being always obtained. Many individuals, in the light or in the dark, creep upward to the water’s edge in an aquarium; but they also move downward, horizontally, or in any inter- mediate direction with perfect freedom. When situated on a glass plate which was tilted in various directions, they continued creeping ‘as they were,’ and could not be made to alter at the experimenter’s will the direction of their creeping. These ex- periments were made at temperatures of 17° to 27°C. There seemed a somewhat more pronounced tendency to upward move- ment at 17° than at 25° to 27°, but the difference was not clear- cut and is perhaps fictitious. If C. zebra possesses statolyths (otoconia) similar to those known in other nudibranchs, and perhaps of general occurrence in the group, they are not conspicuously involved in determin- ing the direction of the animal’s movements in the laboratory, nor the posture of the body in nature. It is of course conceiv- able that a vaguer type of geotropism is really functional, which might be difficult to detect in laboratory experiments. The SENSORY REACTIONS OF CHROMODORIS ZEBRA 277 movements of the nudibranchs in nature are suggestive of this possibility (Crozier, ’17 b, and section III of the present paper). For reasons subsequently discussed (section III), it seemed ad- visable to test the possibility of a relation between geotropism and the temperature. The experiments already referred to were made at different seasons of the year, and the possibility was not lost sight of that the reproductive phase of a given individual might, through in- ternal secretions or otherwise, be instrumental in determining, or in helping to determine, geotropic behavior. The natural movements of C. zebra are of very considerable complexity, and the following statements cannot be applied to the total analysis ‘of these movements. These statements are based upon experi- ence with many hundreds of C. zebra during the last five years. During the winter months, at an average laboratory tempera- ture (in the aquaria) of about 17°, C. zebra is notably geotropie, orienting upward and tending to remain at the water surface, especially when about to deposit eggs. This behavior is also notable in the field. After the egg mass is laid, the animal may wander downward again. At 27°, in summer, the same be- havior is manifest, but less pronouncedly. Hence it is unlikely that the decided upward creeping in the first case is the result of oxygen-want. The effect of oxygen-want, or of some associated condition, may be tested in two ways: 1) by observing the behavior of ani- mals from which the branchial plumes have been removed or in which these organs are prevented from functioning, and, 2) by observing the behavior of C. zebra on a vertical surface in a jar closed above, containing no air space, but communicating with oxygenated water at its lower end. That the gills are respira- tory organs is suggested not only by their blood circulation, but also by the fact that in sea-water of decreased alkalinity (p, = 7.95-8.00) the gill plumes become widely extended, the base of the gill crown being then inflated and protruded beyond the pro- tecting branchial collar. The result of such tests was as follows: Sexually ripe indi- viduals tend to move upward, even though this be away from the 278 W. J. CROZIER AND LESLIE B. AREY oxygen supply. ‘Spent’ individuals do not. When the branchial collar was sewed together so that the gill plumes could not be ex- truded, non-geotropic individuals did not tend to creep upward, but remained on the bottom. The correlation of egg-deposition with negative orientation was very marked. For example, a group of fifteen nudibranchs had been in the laboratory for four months; during the last three and a half months of this time they deposited no eggs and re- mained for the most part at the bottom of their aquarium; sud- denly, on the same morning, six pairs were formed, the animals crept up to the water edge, and deposited eggs, after which they wandered aimlessly. If the temperature be gradually increased in a vessel in which Chromodoris is creeping upward under diffuse light, the ante- rior part of the foot becomes detached from the substratum when the temperature reaches 29° to 30°C., and if this temperature is maintained the animal creeps or falls to the bottom. This is probably an indirect effect of temperature upon geotropism. It is possible that the reproductive mass, enlarged when ripe, acts as a statolyth; if this is correct, geotropic orientation may re- sult: from, 1) a general increased irritability accompanying sexual ripeness, plus, 2) the mechanical stimulation of the loosely anchored internal organs; on a vertical surface, the animal would then turn away from the side against which these organs pressed. This would result in negative geotropism, as found. Together with positive phototropism (vide infra) and a negative reaction to high temperature, geotropic behavior might, then, be im- portant for the determination of the vertical migrations of the species into shallow water at periods of breeding; it would never- theless be incorrect to say that the animal ‘‘moves into shallow water for the purpose of breeding.”’ 3. Rheotropism In some of the situations where C. zebra abounds, as, for example, in Fairyland Creek, the nudibranchs are well exposed to the possibly directive influence of tidal currents of considerable SENSORY REACTIONS OF CHROMODORIS ZEBRA 279 volume and velocity. The habitat of this animal is preemi- nently within the semienclosed lagoons or sounds at Bermuda, where tidal currents must frequently be encountered; it does not occur upon the reefs. It was important to discover the nature of the animal’s rheotropism, if it should be found to be oriented by water currents. Since the usefulness of information upon this point lay in its application to the natural movements of the ani- mal, the experimental work was done in the field. Laboratory tests, moreover, were found unsatisfactory because water cur- rents of sufficient volume could not be employed conveniently. N KILL X B Fig. 3 Chart of a portion of Fairyland Creek (F. C.), showing the situation (cross within a circle) for testing the behavior of C. zebra in tidal currents (see text). A is drawn to a scale of 6” tolmi. B is an enlarged sketch of the region within the rectangular area surrounded by dashes in A. An appropriate situation was found at the western end, or mouth, of Fairyland Creek/fig. 3), in a locality where hundreds of the animals were living at the time the experiments were made, and involving, therefore, water currents normally encountered by the nudibranch. At the period of the falling tide the currents in the location selected were as shown in figure 3. The observations were at first confined to sunny days. When low water or high water occurred at about midday—no water then (for a short period) flowed across the channel indicated—the sun was sufficiently 280 W. J. CROZIER AND LESLIE B. AREY far south (in December) to cause the nudibranchs, not singly, but by the dozen, to migrate southward in this channel. During falling tide, with the current flowing northward, the nudibranch, moved with the current; but once out of the channel they tended to turn around so as to face the sun as soon as they were out of the current. In this way groups of fifty to sixty individuals were caused to collect just beyond the northern end of the little channel. For the study of rheotropic orientation a flat slab of rock was placed horizontally in this channel, and nudibranchs were placed up on it in various positions with reference to the current. In some cases the Chromodoris was allowed to become attached to the rock while surrounded by an inverted glass jar which temporarily protected it from the action of the current. So long as the current was of fair velocity,.orientation was always precisely negative, the nudibranch moving with the current. Under these conditions, the whole body is swayed to one side or the other by the force of the current, the gill plumes are moved by it, the ‘rhinophores’ are bent backward or to one side, and the buccal veil of the mantle is irregularly distorted. The gill plumes and ‘rhinophores,’ in particular, are forcibly moved by a cur- rent too weak to noticeably affect the body as a whole in a grossly mechanical way, yet leading to precise orientation in the current. It was considered that some ‘or all structures mechanically distorted by the water current might be responsible for the negative rheotropic orientation. Experiments were begun with the ‘rhinophores.’ When exposed to mild water currents of sufficient volume, as in the natural channel already considered, the ‘rhinophores’ are forced backward (fig. 5). When the current is stronger, the position assumed is as shown in figure 4. The ‘rhinophores’ are easily removed by seizing with forceps and cutting close to the collar. In one experiment six nudibranchs from which both ‘rhino- phores’ had been removed the previous day (they crept about in an entirely normal fashion, for removal of the ‘rhinophores’ has no seriously adverse effects) were found not to be oriented by a current in the natural channel, although a dozen or more individuals with intact ‘rhinophores’ oriented precisely. The SENSORY REACTIONS OF CHROMODORIS ZEBRA 281 rhinophoreless individuals assumed a position like that in figure 4, and the gills and buccal veil were forcibly distorted to a maximal extent by the current, but no orientation took place. A group of fifteen nudibranchs was then prepared, from nine of which the right ‘rhinophore,’ and from six the left, was removed. When the current was allowed to impinge on the anterior end of the nudibranch, parallel to the long axis of the animal, in almost every case orientation was prompt, and the bending of the body took place in such a way that the side contracted was the one carrying the intact ‘rhinophore.’ Experiments were also made with more localized currents. A stream of sea-water flowing through a tube of 6 mm. bore at a speed of 4 to 5 cm. per second was allowed to impinge horizontally Fig. 4 Showing the posture of the body in a Chromodoris exposed to a tidal current (see text). upon the anterior end of C. zebra. To this current normal indi- viduals became promptly oriented, the process occupying three to five minutes (at 17°C.). Animals without ‘rhinophores’ were conspicuously slow and unsuccessful in orienting away from this current, although the buccal veil and the gills were moved by the current to an equal extent in both cases. The negative orientation to the current did occur in some cases, but only after half an hour or longer. The ‘rhinophores’ are easily distorted by currents and do not retract when moved in this way. A current of small cross section, affecting only the ‘rhinophores’ (fig. 5), causes the animal to bend toward the unstimulated side. These results leave no doubt that to currents of adequate velocity the nudibranchs are negatively rheotropic and that the ‘rhinophores’ are the prime receptive organs for this kind of reaction. 282 W. J. CROZIER AND LESLIE B. AREY 4. Nervous relations 1. There are several very characteristic features about the sensory responses of Chromodoris; these are of considerable general significance. Yet the apparent variability of these responses has made it necessary to study them very carefully and in many individuals. To local excitation, not too intense, the response is local merely; to more vigorous stimulation, the response obtained involves more distant structures—at the anterior end, the general head contraction; at the posterior end, the caudal contraction; A B Fig. 5 Behavior of the ‘rhinophores’ in a water current; A showing the ‘positive’ reaction of the ‘rhinophore’ itself; B indicating the method of reaction to local current affecting directly only the ‘rhinophores.’ likewise at the anterior end the ‘reflex’ involvement of other structures than the one stimulated proceeds upon a conspicuously homolateral plan. In the case of the tentacles and ‘rhinophores,’ moreover, stimulation of a tentacle easily induces contraction of the homolateral ‘rhinophore,’ whereas the reverse operation is exceedingly diff cult. The manner in which, at the head end, additional structures become concerned in reaction to the activation of distant parts, corresponds precisely to the distribution of the main anterior nerve trunks; and the character of the responses, particularly in the apparently non-reciprocal nature of conduction between tentacle or pharynx and ‘rhinophore,’ is strongly suggestive of true reflex action. SENSORY REACTIONS OF CHROMODORIS ZEBRA 283 2. On the other hand, in the case of each of the projections from the body (tentacles, ‘rhinophores,’ particularly the branch- ial plumes, and perhaps the pharynx) the local reaction of each stimulated part has certain definite peculiarities, best studied in the gill plumes, but seemingly identical in all the parts enumerated. These peculiarities are: localized longitu- dinal contraction at the immediate side of activation; circular constriction and longitudinal contraction beyond (distal to) the level of activation; contraction at the base when the activation is sufficiently intense, in this case involving a spreading of the response to neighboring parts; a lesser reactivity when the tip of the organ is activated than when it is touched near its base; in the gill plumes, neuromuscular polarization such that the activation spreads distally from the point of excitation; slight fatiguability of the local reactions, whereas the heterolateral responses (e.g., in case of the ‘rhinophores’ and tentacles) are much more readily exhausted by repeated activation. The local responses are exhibited in pieces of the mantle removed from the body. When the central nervous ganglia, supra- and subcesophageal, have been completely extirpated, stimulation of the head region near a ‘rhinophoral’ collar causes that ‘rhinophore’ to be retracted, the collar closing over it, as normally stimulation of a tentacle leads to its reaction, but does not involve retraction of the homolateral ‘rhinophore.’ Tactile excitation of a ‘rhinophore’ in a Chromodoris with the gangha excised -causes the ‘rhinophore’ to retract, after which it is slowly re-extended. The phenomena of local response to faradie stimulation in the excised gill plumes are also substantially similar to those of the individual plume in the intact nudibranch. Within fifteen minutes after amputation a gill plume becomes relaxed, though, like the excised tentacle of an actinian (Parker, ’17), it is not so fully extended as it may be when attached to the animal, because no fluid i being forced into it. The relaxed, isolated gill plume is fully as sensitive to touch as when forming part of the normal nudibranch, the peculiarities of its reactions are identi- 284 W. J. CROZIER AND LESLIE B. AREY eal, although it is somewhat more quickly exhausted and only rarely responds at all to shading. The responses disappear under chloretone anaesthesia, but returnagaininsea-water. The neuromuscular polarization of the gill plume is therefore a local matter, conditioned by a self-contained nervous apparatus which conducts impulses more easily distalward than proximally. These facts speak unmistakably for the presence of local peripheral conducting paths, having the characteristics of true nerve nets. Similar nerve nets have already been identified in Octopus (Hofmann, ’07), and in Aplysia (Bethe, ’03). The body of Chromodoris may be laid open by a dorsal or a ventral incision, and the animal will live for a long time in sea- water. The nerves which originate from the ‘cerebral’ and sub- cesophageal ganglia and traverse the body cavity are readily employed for faradic stimulation experiments. The results of such tests confirm Bethe’s (’03) description of the effects of nerve-trunk stimulation in Aplysia. Local responses, of no great magnitude, are induced; much more general effects are obtained, with the same stimulus intensity, when the integument is acti- vated directly. These experiments incidentally afforded infor- mation relative to the old controversy as to whether the pro- jecting marginal ridge is an epipodium (Herdman, ’90; Herdman and Clubb, ’92) or a mantle structure proper (Pelseneer, ’94, p. 70). Pelseneer was undoubtedly correct, at least so far as our species is concerned, for the motor nerves to this =e are pallial, not pedal. 3. The general result of these experiments is to suggest the probability that peripherally a true nerve net is concerned in local sensory responses, but that a-reflex system involving cen- tral conducting paths is called into play by more intense acti- vation. We are able to offer in addition physiological proof of a different kind that the peripheral conducting systems are nerve nets, and that the central paths of nervous transmission are part of a synaptic system, to which the term ‘reflex’ may properly be applied. This proof is based upon the assumption that the effect of strychnine affords a good test of synaptic transmission. SENSORY REACTIONS OF CHROMODORIS ZEBRA 285 The following notes are derived from observations with eleven Chromodoris of medium size (10 to 14 cm. in length) into which 1 ec. of half-saturated strychnine hydrosulphate in sea-water had been injected. This quantity was found by other tests not to be fatal and to be the optimal concentration for our purpose. The injection was made into the region of the heart, on the dorsal surface. The behavior of each animal was studied individually, before injection, during the action of the strychnine, and after its effects had worn off. As a control, each individual was studied in comparison with an animal into which 1 cc. of sea- water had been injected. The latter operation had no detectable consequences of any kind. Tactile activation was mostly used. The results herein summarized are to be compared with those given in the first section (p. 267). Following strychnine injection, the body remains for some minutes much contracted, its surface being ‘wrinkled’ and thrown into edematous blebs; the genital papilla is protruded, and the posterior mantle glands are made prominent, owing to the forcing of fluid into the spaces surrounding them. These effects appear under any conditions leading to pronounced general contraction of the body muscles. The gill collar, however, is strongly con- tracted in a peculiar way, its edge being rolled outward. The foot is folded together lengthwise and does not attach to the sub- stratum. The gill plumes remain half contracted within the branchial collar. The plumes tend to exhibit more or less rhyth- mic contractions, followed by rapid but incomplete expansion; perhaps this is in some way mechanically induced by the beat of the heart, which distorts the neighboring dorsal integument. The reaction of the gill plumes to shading is not apparent. After the lapse of half an hour to an hour in different indi- viduals, the body is less strongly contracted, the gill plumes more fully extended. The reactions of the plumes to touch are curious and important at this point: to a single touch, a plume reacts. precisely as in non-strychninized individuals; but when two successive touches are administered to adjacent plumes, the reaction is of unexpected violence (fig. 6). A reaction of this amplitude is obtainable in normal animals only by six or seven 286 W. J. CROZIER AND LESLIE B. AREY repeated proddings of the gill crown, but relatively slight taps of adjacent plumes will produce this effect under strychnine. The ‘rhinophores’ are not retracted under these circumstances; whereas, if the ‘rhinophores’ themselves are touched, the gill plumes do contract. In some individuals the ‘rhinophores’ were found to retract noticeably, but not completely, when a bit of graphite or the end of a glass rod or of an aluminum wire was brought near them (within 2 to 3 mm., but not touching). Presumably this repre- sents a heightened tactile irritability such as that seen in some teleosts (Crozier, ’18 ¢) after the removal of the eyes (i.e., when the central reflex interference of optic impulses has been re- Fig. 6 Outline of Chromodoris to show gill-crown reaction under strychnine. moved). This type of irritability is not apparent in the non- strychninized animal. L After one hour, the dorsum is still wrinkled, but the animal attempts to creep, usually falling over to one side after such an attempt has endured for two to three minutes. The main body is no longer forcibly contracted, but usually assumes a gentle spiral form about the long axis, the head pointing downward on one side, the ‘tail’ of the foot pointing upward and to the opposite side; the surface of the foot is for the most part longitudinally folded. Touching a ‘rhinophore’ causes both it and its mate to retract; sometimes the opposite ‘rhinophore’ contracts before the stimulated one, and usually the gill plumes contract also. The lightest touch applied to a tentacle causes the homolateral ‘rhinophore’ to be fully retracted. When a ‘rhinophoral’ collar is touched, it contracts, sphincter-wise, so quickly and so forcibly, SENSORY REACTIONS OF CHROMODORIS ZEBRA 287 that the ‘rhinophore’ has difficulty in being itself retracted within its pocket; this behavior is never seen normally. If the anterior edge of the buccal veil is touched at one side, there results, as in the normal nudibranch, a homolateral ‘rhinophore’ retraction, and also a retraction of the gill crown, which is rarely seen except under strychnine. After one and one-half to two hours, the pharynx invariably becomes extended; if touched at the side, an exceedingly violent homolateral head response is the result. If the lips be touched, however lightly, both ‘rhinophores,’ as well as the proboscis, are violently retracted. On reextension, the ‘rhinophores’ both retract when either one is touched, but the pharynx (extended) does not contract at all. Nor at any other time does ‘rhinophore’ activation induce retraction of the pharynx. In all of these reactions, for example, when both ‘rhinophores’ retract as the result of one of them being touched, it is very difficult, if not impossible, to secure the double response for five to ten minutes subsequent to the reaction, although each one responds to local activation readily enough after a 1- or 2-minute refractory period; this is true of the double response independ- ently of whether the ‘rhinophore’ first activated or the opposite one is the one subsequently stimulated. NH, Li>Na. The stimulation here is primarily a matter of the cation. KCl, KB, KI, and KNO; lead to equally strong reactions from all SENSORY REACTIONS OF CHROMODORIS ZEBRA 303 parts. CaCl, and MgSO, (0.625 N) did not activate; MgCl. gave fairly good responses over the whole surface of the nudibranch. The limiting dilutions of different substances effective for acti- vation are likewise characteristic. For any one substance the distribution of limiting effective concentrations for the different parts of the body affords a measure of their respective recep- tivities. In the following summary it will be seen that the several concentrations, especially for the oral region, are of the orders of magnitude found in the stimulation of other animals: Picric Acid, dissolved in sea-water. M/150 Good, strong reactions everywhere. M/200 Gill plumes respond weakly, or merely flatten out against the body wall. ney Good responses everywhere, except from the gills. M/4,000 Oral tentacles still very sensitive ‘Rhinophores’ less sensi- tive: they may merely bend before the stream of acid, and not distinctly contract. The dorsal skin is also slightly sensitive M/8,000 Same as M/4,000, but weaker. M/10,000 The mantle edge fails to respond; the edge of the foot, espe- cially its anterior part, is still sensitive. Response from the oral tentacles is more constant than from the ‘rhinophores.’ This solution is distinctly bitter to human taste. KCl, in rain-water. . M/700 Gill-plume reaction weak. M/1,000 No response from the gills. No responses distinguishable from those to an equal volume of rain-water were obtainable with weaker concentrations. In sea-water solution the reactions from the different regions of the body, at M/409 concentration, indicated the following regional order of decreasing irrita- bility (Crozier, 16 a, p. 272): anterior tentacles, ‘rhinophores’ > base of the gill > crown > buccal mantle > posterior mantle veil > lateral mantle edge (ventral surface) . > edge of foot (at the sides) > dorsal integument. KOH, in rain-water. M/300 Gill plumes commonly fail to react. KCl, in rain-water. M/3 Responses from all parts. M/16 No responses from gill plumes. M/50 Doubtful if the responses are distinct from those to rain-water in sea-water. M/15 Gill plume reactions fail. M/20 No reactions from any part. 304 W. J. CROZIER AND LESLIE B. AREY c. From the comparative reactivities of different parts of the body and from the relative limiting concentrations of each sub- stance required for the activation of these parts, the distribution of general chemical sensitivity can be made out over the body surface of Chromodoris. These two criteria lead to mutually concordant results, as an inspection of the preceding paragraphs will disclose. The gill plumes are distinctly the least sensitive of the out- growths from the body; the oral tentacles probably the most sensitive; the ‘rhinophores’ almost as sensitive as the oral ten- tacles, but occupying apparently an intermediate position. On the ground of distribution, it would appear that chemoreception is served by distinct receptors, for the gill plumes are reactive to shading, touch, ete., as already described, in a way which indicates their possession of delicate receptive mechanisms for these sources of activation, yet their chemical reactivity is slight. Evidence of similar import is afforded by comparing the responses of the oral tentacles and of the ‘rhinophores.’ At elevated temperatures (38°C.), tactile responsiveness is quickly destroyed on all parts, but sensitivity to KCl solution (0.625 M) is pre- served. After complete exhaustion to shading, the gills are fully responsive to KCl, M/1,000 picrie acid, ete. The genus Chromodoris is characterized by the fact that many, or most, of its members tend to develop a blue or purple pigmentation of the skin. This pigment is a delicate indicator of acidity (Crozier, ’14, ’716.a), turning pink with acids. This color change is not indicative of an alkaline reaction in the cell interior (Simroth, ’14, p. 484), because, although the cell contents are more acid than sea-water, the pigment is still blue under faintly acid conditions (Crozier, ’16a). This natural indicator offers an exceedingly favorable opportunity for studies on the penetrability of cells for acids, leading to the possibility of in- vestigating the nature of the reaction between acid and tissue in the process of stimulation (Crozier 718 a). In the case of neutral salts, we must also suppose that stimulation is due to some chemical influence of the salt upon the surface of the receptive elements, possibly owing to the fact that the applied SENSORY REACTIONS OF CHROMODORIS ZEBRA 305 salt influences the ionization of protein salts located at these surfaces. The very beautiful experiments of Loeb (’18 a, b) open a way to precise interpretation of this matter. The relative effectiveness of various ions of neutral salts follows an order familiar in many cases of physiological action, frequently regarded as evidence of action upon the colloidal, as distinct from simply chemical, properties of tissue proteins. These effects cannot be interpreted in terms of ‘permeability,’ since, according to Oster- hout’s (’16) exact experiments, the influence of neutral salts upon permeability does not follow this plan. Neither for salts nor for acids can stimulation be regarded as due to increased per- meability of the cell surface. d. Those reactions of Chromodoris which concern its ‘behavior’ in the larger sense have to do with feeding and with copulation. The generally accepted idea that the ‘rhinophores’ are specialized chemoreceptive organs concerned with olfaction has been already disproved (Arey, ’17,’18). C. zebra does, however, give evidence of being activated by low concentrations of materials secreted ° by its companions. These reactions are chemopositive, they are of several kinds, and they are important for conjugation. It is also probable that chemoreception enters into food taking, for it is only when creeping upon algae that the radula is brought into operation. When several sexually ripe nudibranchs are placed in a dish, they very soon protrude the genital papilla, and move toward one another. In fresh sea-water, as, e.g., in an aquarium with running - water, conjugation is quickly effected.® These nudibranchs produce constantly, when undisturbed, more concentratedly if irritated (Crozier, ’16 b), a curiously pene- trating ‘spicy’ odor. This odor is evident in sea-water with which they have been in contact. If the water is unchanged it assumes a blue color owing to the secretion of pigment. Instagnant water a Chromodoris may, but usually does not, protrude the genital - § According to Pelseneer (’11), and as we also have observed, many smaller nudibranchs deposit several egg masses following each insemination. In Chro- modoris this is not true, a single egg mass being the consequence of each insemi- nation (Crozier, 718°), 306 W. J. CROZIER AND LESLIE B. AREY papilla when isolated by itself. In stagnant (i.e., non-circulating) water copulation is interfered with (Crozier, ’18 d), although the nudibranchs may be active and healthy. These facts indicate a chemopositive response to low concentrations of some secretion, which is inhibited by higher concentrations. Mating behavior remains the same after the ‘rhinophores’ have been amputated. Therefore the ‘rhinophores’ are not of special importance for this response. Not only is the genital papilla protruded, but the pharynx as well is everted, even before two conjugating individuals come into contact. It is very difficult to experiment with the sensi- tivity of the pharynx, because it is seldom extruded in a position Fig. 8 The ‘mouthing’ behavior of Chromodoris preparatory to copulation favorable for observations, but it is undoubtedly very sensitive both to touch and to chemical activation. To KCl, M/20 in sea-water, the lips of the fully protruded pharynx were found more sensitive than its outer surface, whereas the walls of the organ ~ (section I) were more reactive than the lips to touch. It usually happens that before becoming mutually adjusted for copulation . (Crozier, 718 d), the nudibranchs pass the end of the pharynx over each other’s surface, moving closer together the while (fig. 8). It seems hardly doubtful that the lips are the chief chemore- ceptive regions in this case, because they may be the only parts in contact with the other animal. This form of chemical attrac- tion is curious, because ‘sexual’ secretions can hardly be involved in a functioning hermaphrodite when reciprocal fertilization SENSORY REACTIONS OF CHROMODORIS ZEBRA 307 takes place (Crozier, 18 d); still, it is possible, for one of the indi- viduals is commonly more active than the other in the maneuvers preliminary to copulation. VI. SUMMARY 1. Physiological evidence is adduced for the existence in Chro- modoris zebra of differentiated receptive mechanisms mediating reactions to tactile, chemical, and shading stimulation, to the constant intensity of light, and perhaps to heat. 2. Locally, responses of the general integument and all of the outgrowths of the body, gill plumes, ‘rhinophores,’ tentacles, pharynx, depend upon locally contained, peripheral, non-synaptic networks. In the gill plumes, and probably in the other pro- jecting parts, these nerve nets are polarized. 3. Reactions involving parts distant from the site of activation depend upon central, ganglionic, transmission. The _ peculi- arities of heterolateral response; of irreciprocal conduction be- tween the several homolateral parts; and of behavior following strychnine injection, show this central nervous system tobe essentially synaptic. 4. The nudibranch is positively phototropic, the chief recep- tive organs probably being the eyes. The branchial collar is also sensitive to light, which causes the gill plumes to be ex- panded. The gill plumes react by contraction when they are shaded; this response is very variable. When sexually ripe, Chromodoris is negatively geotropic. It is negatively rheotropic to strong water currents, the directive organs being the ‘rhino- phores.’ Vibrations transmitted through the water are not responded to. Temperatures of 31° to 32°C. induce negative reactions. Chemotropic reactions to body secretions of other individuals lead to conjugation; ‘olfactory’ stimulation (which does not essentially involve the ‘rhinophores’) as well as ‘gusta- tory’ stimulation (of the lips) are concerned in this behavior. 5. The locomotion of Chromodoris is primarily muscular, not ciliary, the active part being the outer lateral margins of the foot, which suck locally. Progression is strongly polarized in THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 29, NO. 2 308 W. J. CROZIER AND LESLIE B. AREY the anterior direction. The foot is positively stereotropic, and when removed from a surface folds together laterally. These latter peculiarities enable the animal to creep upon narrow blades of eel grass where it feeds. The stereotropism of the anterior end of the foot is responsible for righting behavior; there is no apparent statolythie control of dorsoventral body orientation. Dyer Island, Bermuda, June, 1918. LITERATURE CITED Arny, L. B. 1917 The sensory potentialities of the nudibranch ‘rhinophore.’ Anat. Ree., vol. 11, pp. 514-516. 1918 The multiple sensory activities of the so-called rhinophore of nudibranchs. Amer. Jour. Physiol., vol. 46, no. 5, pp. 526-532. Argy, L. B., AnD Crozier, W. J. 1919 The sensory responses of Chiton. Jour. Exp. Zo6l., vol. 29, pp. 157-260. Bercu, R. 1884 Report on the nudibranchiata collected by H. M. §. Chal- lenger, during the years 1873 to 1876. Rept. Sci. Res. Voy. Chall., Zooél., vol. 10, pp. 1-154. 14 pl. Berne, A. 1903 Allgemeine Anatomie und Physiologie des Nervensystems. Leipzig, vii + 487 pp. CrossLanp, C. 1911 Warning coloration in a nudibranch molluse and in a chameleon. Proc. Zoédl. Soe. Lond., 1911, pp. 1062-1067. Crozipr, W. J. 1914 Note on the pigment of a Bermuda nudibranch, Chromo- doris zebra Heilprin. Jour. Physiol., vol. 47, pp. 491-492. 1915 a The sensory reactions of Holothuria surinamensis Ludwig. Zool. Jahrb., Abt. Physiol., Bd. 35, pp. 233-297. 1915 b On cell penetration by acids. Science, N.S., vol. 42, pp. 735-736. 1915 ¢ Regarding the existence of the ‘common chemical sense’ in vertebrates. Jour. Comp. Neur., vol. 26, pp. 1-8. 1916a Cell penetration by acids. [I]. Jour. Biol. Chem., vol. 24, pp. 255-279. 1916 b On the immunity coloration of some nudibranchs. Proe. Nat. Acad. Sci., vol. 2, pp. 672-675. 1917a The nature of the conical bodies on the mantle of certain nudibranchs. Nautilus, vol. 30, pp. 103-106. 1917 b On the periodic shoreward migration of tropical nudibranchs. Amer. Nat.; vol. 51, pp. 377-382. 1917 ¢ Some structural variations in Chromodoris zebra. Nautilus, vol. 30, pp. 140-142. 1917 d Evidence of assortive mating in a nudibranch. Proc. Nat. Acad. Sci., vol. 3, pp. 519-522. 1917 e Fusion of ‘rhinophores’ in Chromodoris. Amer. Nat., vol. 51, pp. 756-758. SENSORY REACTIONS OF CHROMODORIS ZEBRA 309 Crozinr, W. J. 1918 a Sensory activation by acids, I. Amer. Jour. Physiol., vol. 45, pp. 323-341. 1918 b Cell penetration by acids, IV. Jour. Biol. Chem., vol. 33, pp. 463-470. 1918 ¢ On tactile responses of the de-eyed hamlet (Epinephelus stria- tis). Jour. Comp. Neur., vol. 29, pp. 163-173 1918 d Assortive mating in a nudibranch, Chromodoris zebra Heil- prin. Jour. Exp. Zodél., vol. 27, pp. 247-292. Exrot, C. 1904 On some nudibranchs from East Africa and Zanzibar. Part V. Proc. Zoél. Soc. Lond., 1904, vol. 2, pp. 83-105, pl. 3-5. 1910 A monograph of the British nudibranchiate mollusea, Pt. VIII (Suppl.). London, Ray Soc., 198 pp. GarstaneG, W. 1890 A complete list of the opisthobranchiate mollusca found at Plymouth; with further observations on their morphology, colours, and natural history. Jour. Mar. Biol. Assn., N.S., vol. 1, pp. 399-457. Heriuprin, A. 1889 The Bermuda Islands. Phila., vi + 231 pp. Herpman, W. A. 1890 On the structure and functions of the cerata or dorsal papillae in some nudibranchiate mollusea. Quart. Jour. Micros. Sci., N.S., vol. 31, pp. 41-63, pl. 6-10. . HeRpMAN, W. A., AND Ciuss, J. A. 1892 On the innervation of the cerata of some nudibranchiata. Quart. Jour. Micros. Sci., vol. 33, pp. 541- 558. pl. 32-34. Horrman. F. B. 1907 Gibt es in der Muskulatur der Mollusken periphere, kontinuierlich leitende Nervennetze bei Abwesenheit von Ganglion- zellen? I. Untersuchungen an Cephalopoden. Arch. ges. Physiol., Bd. 118, pp. 375-412. Los, J. 1918a The stoichiometrical character of the action of neutral salts upon the swelling of gelatin. Jour. Biol. Chem., vol. 34, pp. 77-95. 1918 b The influence of neutral salts upon the viscosity of gelatin solutions. Jour. Biol. Chem., vol. 34, pp. 395-413. Otmstep, J. M. D. 1917 Notes on the locomotion of certain Bermudian mol- lusks. Jour. Exp. Zodl., vol. 24, pp. 223-236. OsterHouT, W. J. V. 1916 Permeability and viscosity. Science, N.S., vol. 43, pp. 857-859. Parker, G. H. 1908 The sensory reactions of amphioxus. Proc. Amer. Acad. Arts and Sci., vol. 43, pp. 413-455. 1911 The mechanism of locomotion in gastropods. Jour. Morph., vol. 22, pp. 155-170. 1917 The movements of the tentacles in actinians. Jour. Exp. Zodl., vol. 22, pp. 95-110. PELSENEER, P. 1894 Recherches sur divers Opisthobranches. Mém. couron. Acad. roy. Belgique, T. 53, 157 pp., 25 pl. , 1911 Recherches sur l’embryologie des Gastropodes. Mém. Acad. roy. Belgique, 2™ Sér., T. 3, 167 pp., 22 pl. Ranp, H. W. 1909 Wound reparation and polarity in tentacles of actinians. Jour. Exp. Zodél., vol. 7, pp. 189-238. 1915 Wound closure in actinian tentacles with reference to the prob- lem of organization. Arch. f. Entw.-mech. Org., Bd. 41, S. 159-214. 310 W. J. CROZIER AND LESLIE B. AREY Smroru, H. 1914 Untersuchungen an marinen Gastropoden. Pigment, Loko- motion, Phylogenetisches. Arch. f. Entw.-mech. Org., Bd. 39, S. 457-515, Taf. 21. SmaLLwoop, W. M. 1910 Notes on the hydroids and nudibranchs of Bermuda. Proc. Zo6l., Soc. Lond., 1910 (1), pp. 187-145. SmaLLwoop, W. M., anp Cuark, E. G. 1912 Chromodoris zebra Heilprin: a distinct species. Jour. Morph., vol. 23, pp. 625-636. Vuiirs, F. 1907 Sur les ondes pédieuses des mollusques reptateurs. Compt. Rend. Acad. Sci. Paris, T. 145, pp. 276-278. Resumen por el autor, Matsuziro Takenouchi. Instituto Wistar de Anatomia y Biologia. Estudios sobre la supuesta funcién endocrina de la glandula timo (rata albina). El suero procedente de conejos immunizados con la substancia del timo de la rata albina presenta una reacciOn de precipitina positiva con el extracto del timo, pero esta reaccién no es estricta- mente especifica. El suero anti-timico de conejo no produce ninguna accién hemolitica positiva con los corptisculos san- guineos de la rata cuando se usa como complemento el normal del conejillo de indias o el suero de la rata. Los sueros anti- timicos inyectados en ratas no producen sintomas de ‘‘anafilaxia primaria’’ ni tampoco afectan al crecimiento de estas. El autor no ha observado modificaciones en las visceras. Ha observado casi los mismos resultados con el suero anti-testicular de conejos inyectados con emulsién de testiculo de rata.. Ha intentado provocar la produccién de hemolisinas en los conejos mediante la inyeccién de corptisculos sanguineos de la rata lava- dos, pero el suero normal de la rata con su complemento no puede activar la hemolisina contra los corptisculos sanguineos de ésta, a causa de la presencia de una substancia inhibidora. El pollo no es animal adecuado para la produccién de hemolisina capaz de actuar sobre los corptsculos sanguineos de la rata. Nuestro intento de producir suero anti-timico enérgico en el conejo mediante inyecciones de timo de rata fracas6é probable- mente por la produccién de anticuerpos en el conejo y porque las células del timo de la rata estan protejidas. No podemos admitir hasta el presente la existencia en el timo de una funcién endocrina bien establecida. Translation by José F. Nonidez Carnegie Institution of Washington AUTHOR'S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, AuqusT 11 STUDIES ON THE REPUTED ENDOCRINE FUNCTION OF THE THYMUS GLAND (ALBINO RAT) MATSUZIRO TAKENOUCHI The Wistar Institute of Anatomy and Biology TWO CHARTS CONTENTS bed 1. Introduction.. Ns es een ce ne en eee Ben eee Bee eT 2. General plan onl Age minim we 2g LO 3. Immunization of the rabbit tite the tiytas Pbad i the nipne age eee 317 4. Immunization of the rabbit with testes of the albino rat. ia: . B27 5. Immunization of the rabbit with the red corpuscles of the alba —_ . 330 6. Immunization of the chicken with the red corpuscles of the albino ra . 335, Sees Mera TCS CUSHION Sse. og unin ccs vokeeocmap reset eee ge «ied ten ... 40 grams of thyme Pie ONG SS. ioe ul Se Rice eee ees... -ss. Lue blood was taken The autopsy findings were as follows: Spleen somewhat smaller than usual, without marked macroscopical alteration, liver normal; kidneys both normal. On the right side, in the middle part of the abdomen, directly inside of the peritoneum, there was a small fibrinous body about 12 mm. in length, 6 mm. in breadth, which is nothing else than an abscess enclosed in a strong fibrinous membrane hanging between the convolutions of the intestine. This abscess was most probably caused by the injection of the thymus emulsion, which though strained, had in it some small particles of connective tissue which were dis- solved with difficulty. The separated serum was tested for the presence of the anti- bodies by the precipitin and hemolytic reactions and then in- jected into rats to determine the action in vivo. 318 MATSUZIRO TAKENOUCHI b. Test of the antithymus serum in vitro 1. Precipitin reaction. For the precipitin reaction, we followed the procedure given by Ricketts and Rothstein (’03) for the action of neurotoxic serum. They used for their precipitin reaction an emulsion of nervous tissue as a precipitinogen (antigen). Different precipitinogens (antigens) from different kinds of tissue were prepared in the following manner: Weighed tissue, first washed with saline solution, was ground thoroughly, emulsified with sterile saline solution and shaken six hours at room temperature, with preliminary carbolization for the purpose of avoiding bacterial contamination and alterations in composition. Then it was strained through sterile cotton gauze. The strained emulsion was centrifuged for a long time at high speed to free the supernatant fluid from macroscopic particles. The supernatant fluid was diluted again with sterile saline solu- tion and carbolized. The following tissues were prepared in this way. TABLE 1 TISSUES Kidney} Spleen | Testes | Brain Thy Weight (grams).. weseeeeee| 2.075} 0.842) 0.742) 1.500) 0.899 Salt solution addiedls in ina (ce! 5 wees e| 20.0') 5.0" 15:0 eee Salt solution added to the fiperatant fluid (CC.). oo san aveeleh yl. teu eg bepieeeeprennt Veen) 1) «apne OOD Oe 1G To six test-tubes each containing 1 ce. of the diluted extract there were added from one to five drops of serum, the sixth test- tube being held as control. The quantity of the liquid in each test-tube was made the same with saline solution. The tubes were then put into the ice-box for sixteen to eighteen hours, avoiding any evaporation. The first reading was made after four to six hours and the second sixteen to eighteen hours later (table 2). Control tests with normal rabbit serum give a little precipi- tation in some tubes, but this is plahta and inconspicuous in others. ENDOCRINE FUNCTION OF THE THYMUS GLAND 319 TABLE 2 Showing the precipitin reaction of the antithymus serum withthe extracts of different rat organs TEST-TUBE 1 2 3 4 5 6 Hxiraet (del ire, eee ot eee EP LOY 1.0) 1.0) Rooney eae Antithymus serum in drops............. 1 2 3 4 5 First reading (4 to 6 hours) CRONE Cyto +. Shcekst dn. +o pustd seevolede Syed ITE - — - mi oe PT aos LB sic nn ors Ae eae boda og gical oe _ — = = bat MME ROG otha ya oe hei Pa essa e a'Fie ste o La] > — — =| Oe FE ee ee. J HORS EEE RAE ROTC) aati _ = = 1 if enna ite xpos onreh tre NAT as ake le pope site|) ces. |. || Second reading (16 to 18 hours) eadney 1a Gls. ehh es eee ES CBR ee eT ee = Spleen...... 0.0.0... cece ee ee ee ee ee eee] EH [F4++)/4+4+4+]/++4+/4+4+4) -— ERR B WN A th ioc. s ahs 5 Sictteaaconeesse Bend St Led + +4 |4+4]/t44/+44) — OETA S Geel es ARI Ps ears ire tl aaIR = = a Je Thymus..............0...00 eee ee eee eee | BH [4+] 44+4)/4+4+4+/4+44] - +++, marked sedimentation. ++, less marked, but easily recognizable precipitation. +, slightly positive precipitation. —, negative precipitin reaction. Another technique for the precipitin reaction which is gener- ally used in serological work (namely, mixing the antiserum with about an equal amount of some dilution of antigen injected for the production of the antiserum, with resulting turbidity and rapid flocculation) was tried with the material used in the above “tests, but the reaction was inconspicuous in all cases. Using as precipitinogen, serum from a normal rat, diluted ten times, instead of organ extract, all the test-tubes, to which from one to five drops of the antithymus serum had been added, failed to show either precipitation or sedimentation. From the result mentioned in table 2 we can tentatively con- clude that the antibodies which are present in the antithymus serum are not so specific as we are taught to believe by some 4 320 MATSUZIRO TAKENOUCHI investigators. A similar lack of specificity of the antibody in the antiserum produced by injection of rat testis tissue to the rabbit will be mentioned later. 2. Test of the hemolytic power of this antithymus serum. The antithymus serum from the rabbit does not possess any appreciable amount of hemolysin against the red corpuscles of the rat in vitro. For the test in vivo, 1 cc. of that serum was injected subcutaneously into three rats thirty days of age, taking three more rats of the same litter as controls. All three rats grew perfectly well and reached normal size. Just the same effect was observed with the antithymus serum from the second and third rabbits (rabbits B and C). c. Test of the antithymus serum in vivo Three litters of healthy young rats of the same age were taken for this test (series no. I, II, and III). To one half of each litter antithymus serum (from 0.3 cc. to 0.5 cc.) was injected subcutaneously, while to the other half of each litter correspond- ing amounts of normal rabbit serum were injected. All the rats were put under the same conditions and the growth was con- trolled by weighing each rat almost every day. Six to twenty- five days after the last injection, some of the rats, both test and control, were examined for the weights of each organ. Histologically, studies of the thymus gland from all the rats were carefully made to see whether there was any histological change caused by the injection of the antithymus serum. As an example of the growth of the test and control animals, I will give one chart which shows the average curve of the rats in, series II (chart 1). In no case of the injection of either the antithymus serum or the normal rabbit serum did we observe any pathological symp- toms like the so-called primary anaphylaxis due to the serum injection, or any sudden decrease of the body weight of rats, such as those reported by Shimizu (’13) in his dogs, after the injection of antithymus rabbit serum. ENDOCRINE FUNCTION OF THE THYMUS GLAND 321 The slight fluctuations of body weight in our charts depend very much upon the amount of stomach contents of the rats: Though the weighing was done mostly at a definite time in the morning before the regular feeding of the rats, circumstances sometimes obliged us to weigh our rats at other times, In any BERR SER S80R0850 PET ene ae red HH Hae LAT ret ret tt PERE EEA ee EE ae a8 BRBaaae ESE EaSeOLUsaneoeee BRT EEERE ERD E2E Loo eee A eee eee eee BEERS SEOUCS eRe eee eee BGR BEE EEE EE EEE eee Poo UGG 58 (SSS ES REE Pees Os ae SeS OS oOo eae eee ee SSeS SSeSRSeeeeOeees 25 30 35 40 45 50 55 60 65 10 5 80 85 Chart 1 The average growth curve in body weight on age, series II. The number of test rats, 4; the number of control rats, 5. The inverted black-dot arrows ({) denote the injection of the antithymus serum into the test rats. The ring arrows ({) indicate when each control rat was taken for examination. case, all the individuals in a litter were weighed at the same time. A special analysis of the weight of the thymus gland of all rats is worth recording here (tables 3, A, and 3, B). The calculated values are based on table 72 in The Rat (Donaldson, ’15). The data on the weight of the other organs do not show any definite difference between the test and control animals, and therefore are omitted entirely. “ THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 29, No. 2 322 MATSUZIRO TAKENOUCHI TABLE 3, A Showing the details regarding the studies on the weight of the thymus of the test rats a ee ath Faris) sg INITIAL FINAL gaters Le ene had OBS. THY. W.? esi THE THIRD Oe Lae? WEIGHT | WEIGHT OF OF SEX ukaeTdee 100% INATION INJECTION THYMUS GLAND days days grams grams grams grams per cent 35 4 1, 3, Mao 36 0.073 | 0.123 73 36 5 Il, 2) Miles. 38 0.091 | 0.159 87 ‘40 9 “\ ILE, s, Melee 33 0.062 | 0.236 59 At 10 LL, Miles 22 43 0.125 | 0.382 102 ‘54 23 Ii; 4; Fl 19 45 0.067 | 0.007 49 155 24 1 Me ink eT 61 0.101 | 0.597 56 58 7) ee 21 57 0.100 | 0.423 59 64 33, UG, Seno a3 71 0.173 | 0.0221 81 ' 64 23 Ge] 22 64 0.157 | 0.011: 83 83 52 L4M.) 21 120 0.200 | 1.689 70 84 53 at M.| 22 95 0.236 | 0.764 94 $4 53 feo F. | 21 102 0.129 | 0.074 49 84 53 Meer) 18 99 0.217 | 0.077 85 85 FARK: | 28 101 0.250 | 0.071 96 85 Saeed, F.| | (22 94 0.244 | 0.067 98 85 54 L2.M.| -20 114 0.266 | 0.716 98 AVOTA BET NYS mee areca. «fe ice heen eee keretts: @ oe cetacean CW ‘1 The weight of the ovaries of these two rats, of same age, differ as 2:1. This kind of difference may be caused often by ovulation and by the formation of corpus luteum. . ? The calculated values are based on table 72, The Rat (Donaldson, 715). d. Repetition of thesame plan of experiment with the second and the third antithymus serum (rabbit, group A) ‘The same plan of experiment just described was repeated with the second and third antithymus serum, and the respective data will be given here briefly. The second antithymus serum was obtained by the immuniza- tion of a healthy male rabbit (rabbit, group A, no. 2) with the thymus of albino rats, beginning with an injection on August 27, 1918. The injection was repeated three times until the blood was taken September 21, 1918. The autopsy of this rabbit showed nothing pathological. The separated serum, carbolized as usual, was used for the experiments. ENDOCRINE FUNCTION OF THE THYMUS GLAND 323 TABLE 3, B Showing the details regarding the studies on the weight of the thymus of the control rats AGE AT EXAMINED OBSERVED|OBSERVED or inte [rae fatap| MEMBER An] WUTAr | rmmat, | weronr | waranr [O38.7HE.: x 190% NATION INJECTION THYMUS GLAND days days grams grams grams grams per cent 35 4 I,2;M.| 20 32 0.054 | 0.152 67 36 5 II, 4,M 24 43 0.071 | 0.294 59 40 9 TET, 5, .M.|* 22 37 0.048 | 0.276 48 41 10 Bie Wer 123 46 0.147 | 0.283 109 54 23 IL 2;)F: 18 48 0.195 | 0.010 139 55 24 II,3,M.| 18 59 0.090 | 0.600 51 58 27 III,4,M.| 22 60 0.121 | 0.181 67 64 33 ETT ay. 23 72 0.162 | 0.020 76 64 33 III,3,M.| 24 73 0.193 | 0.516 90 83 52 15-4, Mi.|.- 20 121 0.282 | 1.100 99 84 53 TEAL F. 23 96 0.231 | 0.050 91 84 53 II,5,M.; 19 95 0.187 | 1.193 73 85 54 Tit, 2, F: 22 111 0.305 | 0.072 111 85 54 Lak. 21 109 0.190 | 0.068 70 85 54 5, Ek. 20 99 0.211 | 0.043 83 fo, GIRLS POO8 SRM ea aan ie 5” a a a 82 Precipitin reaction tests with the extracts of several rat organs gave Just the same results as were obtained with the first anti- thymus serum. For the experiment with the second antithymus serum, three litters of twenty-cight-day rats were used (Series no. X, XI, and XII). The injection of the antithymus serum was made on twenty-eighth (0.3 cc.) and on the thirty-sixth, forty-second, and forty-seventh days of age (each 0.5 ec.). Into the control animals, representing almost one half of each litter, nothing was injected. The average growth curve is given (chart 2). Owing in part to the parasitation of the rats by mites (Lelaps echidninus) and to the effects of dipping to kill these parasites, the growth curve of the rats of all three series was very low as compared with the standards (Donaldson, ’15), especially in the latter part of October. We observed a sudden drop of body weight of the rats in series no. XI between October 31 and No- 324 MATSUZIRO TAKENOUCHI vember 1, 1918, when the examination was made. This was due to the ‘dipping’ of the rats in soap-kerosine emulsion on the afternoon of October 31st, just after the body weight had been taken. This matter must be taken into consideration in the interpretation of the histological findings in various organs of all rats in these series. The third antithymus serum was obtained from one rabbit (rabbit, group A, no. 3) immunized by four injections of rat thymus emulsion, beginning October 7, 1918, and the blood was taken on November 8, 1918. The serological tests on precipitin 60 COPEL Perea) jBODY WEIGHT GRAMS [cece RESCUES Sia CI an 50,7 ARRESP ZEEE EOEoe 47 p poaapeesass 25 30 35. 2~«~O«4O 45 50 55 Chart 2 The average growth curve, series no. X, XI, and XII. The graph with dots marks the test series and the arrows show the age at which injection was made. The graph with circles marks the controls. and hemolysin yielded the same results as in the cases of the first and second serum. Three litters of albino rats at the age of fifty days were taken for the experiments. The antithymus serum was injected into one half the members of each litter five times, with several days’ interval, while into the other half nothing was injected (controls). The relation of the weight of the thymus of all these rats to the body weight and to the age at the time of examination is given in table 4. The computed relative values for the thymus given in the last column of the table are based on the standard values in table 72, of The Rat (Donaldson, 715). ENDOCRINE FUNCTION OF THE THYMUS GLAND 325 TABLE 4 Showing the details regarding the studies on the weight of the thymus of the test and control rats of series No. XIII, XV, and XVI. C = controls ace ar | TION AF = a ee ee THE EXAM-| TER THE | UTTER, ek, aoF Sa aes “ce oeree }s pens oan ae —s = X 100% INATION LAST THYMUS INJECTION days +o ae | grams grams grams per cent per cent 77 10 ULL tsk. 34 70 0.161 56 77 10 PS ULE 2, By 30 72 0.223 78 We 10 PORTE ie 28 75 0.168 59 77 10 XIE, CC; 2, BF. 30 68 0.128 45 77 10 MILL, C3, F 30 69 0.161 56 87 13 XV, 1, M. 50 105 0.133 46 87 13 DOV, 2 be 59 95 0.131 46 87 13 XOVi sah 52 75 0.108 36 87 13 Ve C— 15 Be 48 65 0.101 39 87 13 XV, C, 2, F 55 100 0.141 49 95 25 SVP 43 120 0.220 $2 95 25 VIE 2. Fy 35 66 0.145 53 95 25 XVI, 3, M. 43 105 0.132 49 95 25 XVI,.C, 1, F. 32 69 0.162 60 95 25 XVI, C, 2, M. 32 89 0.149 55 95 25 EVE, G, 3, Msi eaO |. 113 | 0.131 48 miyerage) tieahiamaud.. 174 U0. Se sees ai (Sik 54 Uke EeR 56 DMerAge; CONC lspa:* 14; \ 9, eas ogeeme area elds . . oss, sah? cauricrsianorat ra oe ae ae 343 TTL, TOT HGTaGY "ie ere eee eR a AP «1! IE = 344 Beer AGAraius and meynodgs..). s.7/aeaeees eae | esse de. ake es eee 350 VeeViateritalk ie tP.) A eth ee oe ee iced od. Shit See 358 Vee DenaviOriOl NOGA tDEGss aie. carne een els cicyses ois o1cys1sfavaiss ae 360 VI. Behavior of bees with one eye blackened......................20000: 371 Wirt Varability of photiewespomse..;...-2.802 2-2 <.-+.04.+-.. 5.30 one 391 Willer Natureot-photic: orlentattoneeeaeae sect eene oe oes a: a0). > See 403 Pe General summary and, COMelUsONss. 2...) de wtls qc 2+ decile ols» ag COR 410 XG Bibliography 3. c-.i.0: See eRe ere ee yrs fo Tae sc © see 412 aaa ake) 110 1b. age em Aa eo, oh a REE 415 I. INTRODUCTION The circus movements produced by blackening one eye in cer- tain arthropods have long been familiar to zoélogists. It was not, however, until the advent of more recent interpretations of behavior, that they received any considerable attention. Then for the first, the significance of their relationship to normal orien- tation was recognized. It became apparent that the nature of the stimulus involved in the two cases was the same. Obviously, therefore, the application of any general theory of photic orien- tation to those forms in which circus movements occurred de- pended upon its ability to explain this phenomenon satisfactorily. This consideration has led, within the last few years, to a number of more or less extensive investigations of these reactions. When the present researches were begun there had been no attempt to study circus movements quantitatively. During the ‘ Contributions from the Zodlogical Laboratory of the Museum of Compara- tive Zodlogy at Harvard College, no. 320. 343 . 344 DWIGHT E. MINNICH progress of the experiments, however, Dolley (’16) has published a contribution to this phase of the subject. His methods as well as bis results, on Vanessa, differ widely from those to be de- scribed for the honey-bee. Although in his experiments, as in mine, the illumination employed is defined as non-directive, it was very unlike in the two instances. The results obtained by Dolley are described in terms of circus movements of greater or lesser ‘angles of curvature;’ those obtained by the writer, in terms of degrees turned per centimeter. The conclusions drawn in the two papers are also widely divergent. It is a pleasure here to acknowledge my deep indebtedness to Dr. G. H. Parker, at whose suggestion this research was under- taken and with whose helpful criticism it was carried on. I wish also to express my gratitude to Dr. E. L. Mark for the courtesies and privileges of the Zoélogical Laboratory. II. LITERATURE As early as 1796, Goeze? (p. 42) recorded the fact that a hornet in which one eye had been painted over with an opaque varnish always flew toward the uncovered eye. Some years later, Tre- viranus (’32, p. 194) described an experiment in which the lower half of the right cornea of a dragon-fly was carefully cut away from the optic nerve, with the result that the animal moved toward the left side. . Decidedly the most interesting of the earlier observations are those of Dubois (’86) on a phosphorescent elaterid beetle of the genus Pyrophorus. This insect responds positively to at least certain intensities of light, and according to Dubois (p. 209) it is most affected by the yellow-green rays, which also predominate in the spectrum of its own light. The photogenic organs are three in number, one occupying a median ventral position on the first abdominal segment, the other two being situated on opposite sides of the prothorax near its dorsolateral edges. Whenever the beetle begins to creep spontaneously in the dark, the pro- thoracic organs become luminous. During flight the abdominal organ does likewise. 2 T have not had direct access to this work. The above reference is taken from a footnote in Treviranus (’32, p. 193). PHOTIC REACTIONS OF HONEY-BEE 345 Dubois (p. 208) found that upon completely obscuring the light from the prothoracic organ of one side of the body with a covering of black wax, the beetle no longer crept in a straight line. Smoked paper records, made in a dark room, showed that such individuals crept in circles toward the functional eye. A check experiment, moreover, showed that the results obtained were not due to the weight of the wax. If instead of eliminating one of the prothoracic organs, the cornea or the entire eye of one * side was destroyed with a red-hot needle (p. 211), very similar results were obtained. When, however, both photogenic organs of the prothorax were obscured or both eyes were destroyed, the animal crept in a hesitant, irregular fashion, presently stopping altogether. Dubois has interpreted these results from an anthropomorphic viewpoint, as evidenced by his original paper and by a more recent comment (’09). To the present writer, however, these responses of Pyrophorus afford not only a typical case of circus movements, but one of considerable theoretical importance as well. The tendency to circle attendant upon the suppression of one photogenic organ or the destruction of one eye may be attrib- uted to the unequal stimulation on the two sides of the body. If this be correct, the case is indeed unique, for the beetle is ori- ented by its own luminosity. This, of course, in nowise affects behavior in the normal animal. With a photogenic organ on each side of the prothorax producing light of the same quality and intensity, it is always perfectly oriented with respect to its own light. But if the source of light or the photoreceptor of one side be eliminated, the beetle promptly orients toward the oppo- site side, the side which is receiving the greater stimulation. In recent years, a steadily increasing number of arthropods have been shown to exhibit circus movements when one eye is blackened or destroyed. The researches of Bethe (’97 a), Axen- feld (’99), Holmes (’01, 05), Radl (’01, 03), Parker (’03), Had- ley (08), Carpenter (’08), Brundin (’13), Holmes and McGraw (713), Dolley (’16), and Garrey (’17), have demonstrated conclus- ively that among phototropic arthropods generally, unilateral photic stimulation results in a more or less asymmetric response. 346 DWIGHT E. MINNICH These investigations have covered between fifty and sixty spe- cies, including the four chief classes of arthropods. Among the insects, where most of the work has been done, representatives of most of the larger orders have been experimented upon. These embrace Orthoptera, Blattoidea, Hymenoptera, Coleoptera, Odonata, Lepidoptera, Diptera, Homoptera, and Hemiptera. The phenomenon of circus movements—or perhaps better, asym- metrical response—must, therefore, be regarded as general rather than exceptional for the members of this phylum. The form of response naturally varies with the peculiarities of locomotion in a given species. It is not the same for a sidewise moving crab, such as Carcinus, as it is for an insect which moves forward. With the usual type of forward locomotion, however, arthropods with one eye blackened generally circle toward the functional eye, if they are positively phototropic; toward the non- functional eye, if they are negatively phototropic. It is true there are cases which, on first examination, do not appear to conform to this generalization. Thus Holmes (’05, pp. 332-336) has demonstrated clearly that an animal with one eye blackened may at first perform circus movements in creeping toward a light, only to modify its behavior after a time and creep in a straight path. Such was true of both Ranatra and Noto- necta. Axenfeld (99, p. 375) had previously made similar observations, and more recently Brundin (’13, pp. 337, 346-348) and Dolley (16, pp. 371-382) have demonstrated the same phe- nomenon in the species with which they worked. There can be no doubt, therefore, that many arthropods with one eye blackened are able in time to modify their behavior to light. This, however, in nowise lessens the significance of the initial tendency of the animal to perform circus movements. In fact, this initial tendency is the all-important one as far as the question of orientation in the normal animal is concerned. I do not believe, therefore, that the presence of modifiability in an animal warrants considering its behavior as an exception to the general occurrence of circus movements. A second difficulty in the way of any generalization concerning circus movements has been encountered in the behavior of cer- PHOTIC REACTIONS OF HONEY-BEE 347 tain flies. Thus Raédl (03, p. 62) says, ‘“‘ Die Calliphora vomi- toria bewegt sich fast ebenso gerade mit einem geschwiirzten Auge, wie wenn sie aus beiden sieht, und es ist mir nicht leicht, diese Erscheinung zu erkliiren.”’ Carpenter (08, p. 486) states that Drosophila with one eye blackened ‘crept in a fairly direct path toward the light, although a tendency to deviate toward the side of the normal eye regularly occurred.” It is possible, I believe, to interpret these cases as merely more extreme instances of modifiability, in which regulation occurs very rapidly instead of after a more or less prolonged experience. That modifiability is operative, at least in the case of Droso- phila, is evidenced by the following statement of Carpenter (p. 486). ‘‘The tendency to diverge from the direct path toward the side of the uncovered eye was overcome by a series of short, quick turns in the opposite direction, which kept them headed toward the light.”” Further evidence in the case is afforded by the be- havior of one fly which, according to Carpenter, persisted in per- forming circus movements. This fly, however, (p. 486) ‘‘had long been active, and showed signs of fatigue.”’ As will be shown later, very similar phenomena were observed in the honey-bee. In conditions, such as that of weakness, induced by long experiment, the bee frequently circled much more toward the functional eye than it had formerly done. It seems probable that in such states the animal approximates more nearly to a simple, reflex behavior. Factors effective in modifying behavior in the vigorous animal have ceased to be operative. If these interpretations be correct, the conspicuous absence of circus movements in Drosophila is only an extreme case of modi- fiability, and offers no real objection to the general conclusion to be drawn from these reactions. However, further work is neces- sary upon both Drosophila and Calliphora before they may be disposed of with certainty. Responses of still another kind have seemed perhaps the most formidable obstacle to any general conclusion as to the occur- rence of circus movements. Thus Hadley (’08, p. 197) has shown that whereas the ‘progressive orientation’ of the lobster larva after the blinding of one eye is positive, the larva performs circus 348 DWIGHT E. MINNICH movements or turns toward the injured side. Brundin (713, p. 346) states that in positive specimens of Orchestia traskiana, cir- cus movements will occur as often toward the blackened as toward the normal eye, while Holmes and McGraw (’13, p. 370) report the case of a positive skipper butterfly which almost inva- riably circled toward the blackened eye. A very plausible explanation of these apparent anomalies, however, has been offered by Dolley (’16, pp. 394-399), who has shown that the contact stimulus afforded by the material cover- ing the eye is sufficient to cause Vanessa, when in the dark, to turn continuously toward the covered eye. This tendency, more- over, exhibits little, if amy, modification from day to day. The effect of such a contact stimulus is continuous. But in the pres- ence of photic stimulation of moderate or high intensity, it is quite overwhelmed by the strong phototropism of the butterfly. In the case of animals of less certain phototropic index, this contact stimulus is, in all probability, frequently strong enough to over- come the effect of ight. An examination of the cases cited above shows that the phototropism of these animals is not of the une- quivocal kind exhibited by Vanessa. It seems likely, therefore, that their apparently exceptional behavior was due to contact and not to photic stimulation. Suppressions of photic circus movements by responses to other stimuli are not surprising, when it is recalled with what facility even the stereotyped circus movements produced through uni- lateral lesions of the central nervous system may be altered in a similar manner. Thus Bethe (’97 b, p. 507) states that the ten- dency of bees to circle toward the normal side after the removal of one half of the brain or the severance of one of the oesophageal commissures, may be arrested, and the animal may even be com- pelled to deviate toward the injured side by stimulating the legs of the normal side. Moreover, in a general statement concern- ing the several crustaceans and insects subjected to similar opera- tions (p. 541), he says, ““ . . . nach Aufhebung der Hem- mung der gesunden Seite durch angebrachte Reize aber auch spontan bei allen Versuchsthieren gerader Gang und Kreisgang nach der operirten Seite eintritt.”’ PHOTIC REACTIONS OF HONEY-BEE 349 Whether the effect of contact stimulation also accounts for certain of the phenomena observed by Axenfeld (’99) is not so clear. Axenfeld reports that nocturnal lepidoptera with one eye blackened turned toward the blackened eye during the day. In the same paper he makes the following general statement: “‘ En- fin on peut observer que ces mémes animaux photofuges, qui tour- nent en pleine lumiére du soleil du cdté de l’oeil couvert, offrent le mouvement contraire au soir ou méme de jour, quand ils sont transportés dans une chambre mal éclairée; . . . .” It may be that: such animals, being attuned to a low intensity, re- spond positively to it, whereas a stronger intensity evokes a nega- tive reaction, somewhat according to the idea of Davenport (97, p- 197). Certainly, if the circling of the nocturnal lepidoptera toward the covered eye was a light response, it is not in harmony with the statement of Loeb (90, p. 51) to the effect that all ‘day and night butterflies’ are without exception positively photo- tropic. J am led to suspect, however, that some of the reactions noted by Axenfeld were the results of contact stimulus, for Hess (13 a, p. 651) has shown that Coccinella, which Axenfeld reports as circling toward the blackened eye, is not negative to light. Axenfeld’s experiments, therefore, need careful repetition before any final conclusions may be drawn from them. . It seems quite certain, therefore, that what have appeared to be exceptions to the general occurrence of circus movements among phototropic arthropods are not really incompatible with this view. Taken as a whole, the investigations of these reac- tions demonstrate rather conclusively that, although they may be modified through experience or obscured by responses to other than photic stimuli, they are, nevertheless, to be considered as characteristic of phototropic arthropods. Photic orientation in this group of animals, therefore, cannot be accounted for by any theory which fails to offer a satisfactory explanation of circus movements. 350 DWIGHT E. MINNICH III. APPARATUS AND METHODS 1. Directive Light In the experiments of the present paper, both directive and non-directive light were employed. Those involving directive illumination were carried on in a circular area (fig. 1) 2.44 m. in diameter, which was laid out in black lines on the concrete floor of a dark room. Sixteen centimeters? above the center of this area, an incandescent lamp was suspended. The lamp employed was a 100-watt, 115-volt, stereopticon, Edison mazda lamp. Of several bulbs used in the course of experimentation, only the last was determined photometrically, its candle-power being approxi- mately 80. These lamps when new are calculated to furnish 100 c.p., but their efficiency decreases considerably with usage. In making tests in the directive light area, bees were started creeping at the outer circumference. The course of the animal as it traveled toward the light was then traced as accurately as possible on a record bearing a plan similar to that of the light area and drawn to scale. Such a record is shown in figure 1. 2. Non-directive light a. Construction. The apparatus employed to furnish non- directive light consisted essentially of a white-walled, cylindrical chamber. This chamber was illuminated by an incandescent lamp, the ight of which was diffused through a thin, white screen, suspended a short distance below the lamp. Bees were admitted to the apparatus through a small, circular opening in the center of the floor, and the course of their creeping was then traced as accurately as possible on a record. The apparatus was espe- cially designed to afford a creeping animal a continuous photic stimulation of uniform intensity over the entire surface of the eye. A more detailed description is presented in the following paragraphs (see figure 2). ® Distances from lamps to creeping surfaces were measured from the center of the filament in all cases. PHOTIC REACTIONS OF HONEY-BEE 351 The cylindrical chamber, which measured approximately 84 em. in height by 87 cm. in diameter, was constructed on a light wooden framework covered on the exterior with heavy, corru- gated cardboard. On the interior it was lined with a thickness Fig. 1 Plan of directive light area, showing two trails of anormal bee. Note the deflection of the courses in the non-directive region near the lamp and directly beneath it. of dead white, cotton cloth, backed by a layer of heavy white paper. On one side of the cylinder, and extending from its bot- tom edge, a rectangular opening 58 em. high by 32 cm. wide was cut through the cardboard and paper layers. The white cloth lining only closed this opening, and it was here slit from top to bottom, the bottom edges being left: free. The two flaps thus aoe DWIGHT E. MINNICH formed allowed free access to the interior of the cylinder. In one! of them a small opening (fig. 2, 0), 3 by 4 em., was cut for purposes of observation. The top of the cylinder was similar in construction to the side walls except that the cardboard layer was omitted. Near oppo- site edges of the top, two circular openings, 8 cm. in diameter, 0 s a A e eee CLM MMMM hh h : c Fig. 2 Diagrammatic section through non-directive light apparatus. c, transferring cage; e, entrance to light chamber; h, handle of slide opening and closing e; 0, opening for observation; s, light screen; v, ventilators. were cut (fig. 2, v). These were covered with a thin, white gauze of coarse mesh, and served as ventilators, preventing any undue rise of temperature within the apparatus. The bottom of the cylinder was formed by a layer of heavy, dead white paper, which covered the table on which the cylinder stood. This paper was especially selected to afford a good creeping surface. On it was drawn a plan, similar to that shown in figure 3, by means of which the course of a creeping bee could be accurately followed. PHOTIC REACTIONS OF HONEY-BEE 353 The illumination of the apparatus resembled the semi-indirect illumination of a modern house, the light from an incandescent lamp being diffused through a circular screen (fig. 2, s), 22 em. in diameter, of white bond paper. ‘Two intensities of illumination were employed. The less intense was produced by a carbon filament lamp of approximately 2 c.p.,4 66 cm. above the floor, and the more intense by the 80 ¢.p. mazda lamp previously described, 33 em. above the floor. The intensity of illumination in each instance was measured at three different points on the floor of the cylinder. One determination was made at the cen- ter; a second at a point 3 cm. from the right side wall, and a third, at a point 3 cm. from the left side wall. The results of these measurements are given in table 1. Hereafter, in referring to the TABLE 1 A B c D E F INTENSITY ON INTENSITY ON CANDLE-POWER | INTENSITY ON FLOOR 3 CM. FLOOR 3 CM. AVERAGE OFC | AVERAGE OF B OF LAMP FLOOR AT CENTER|FROM RIGHT SIDE| FROM LEFT SIDE AND D- AND E WALL WALL c.p. me.® me. me. me. mc. 2.36 25.9 17.93 25520 21.59 Aeato 79.45 | 1051.5 831.37 894.45 862.91 957.21 two intensities of illumination employed, the averages given in the table will be used in round numbers. ‘The less intense will be designated as non-directive light of 24 mc.; the more intense, as non-directive light of 957 me. The transference of bees to and from the apparatus was effected by means of a small, cylindrical cage of wire screen, 5 cm. in length by 2 cm. in diameter (fig. 2, c). This cage, one end of which was open, exactly fitted into a circular opening cut through the table top to the center of the chamber floor. By means of a 4'The lamp used throughout experimentation was, unfortunately, broken before being determined photometrically. Its candle-power was certainly between 2 and 4. * Throughout the present paper, the abbreviation me. will be used to designate meter candles. 354 DWIGHT E. MINNICH slide operated by a handle (fig. 2, h), it was possible, after in- serting the cage, to open or close the light chamber at will. The difficulties involved in direct manipulation of bees were thus entirely avoided. An individual to be tested in non-directive light was merely allowed to creep into the cage, which was then inserted into the opening in the table top. The slide was then pushed aside and the bee allowed to creep up on to the floor of the apparatus. As soon as the bee had entered the light chamber, the slide was pushed back, closing the entrance and leaving the floor of the apparatus complete. The ideal apparatus for studying the effects of continuous pho- tic stimulation of a constant intensity would be one so constructed that all the ommatidia of a compound eye would receive equal illumination, irrespective of the direction of locomotion. Such an apparatus is virtually a physical impossibility. However, the apparatus just described is perhaps somewhat of an approxi- mation to it, even if it does not afford an absolutely uniform light intensity over the floor of the light chamber. As table 1 shows, the illumination is more intense toward the center. Some fluc- tuation will, therefore, occur in the stimulation of the various ommatidia as the animal moves. However, in any position whatever on the floor of such a light chamber, all the ommatidia are receiving some stimulation. Moreover, the amount of stimu- lation received by those areas of the eye which are minimally affected does not differ vastly from that received by areas of maximal stimulation. b. Records. The method of recording behavior in non-directive light is illustrated in figures 3 and 4. The animal to be tested was transferred to the light chamber, and its course of creeping, observed through the ‘peep hole’ in the curtain, was traced as care- fully as possible on a record sheet. The record bore a plan simi- lar to that on the floor of the light chamber, drawn on a scale of 1 to 6. The duration of each trial was ascertained by counting the rings of an electric bell, attached to an electric clock regulated to seconds. How long the trial should last was determined by an interval previously decided upon or by the animal encountering the side wall of the chamber and creeping up. On completion of PHOTIC REACTIONS OF HONEY-BEE apo a trial, the bee was removed, and the remaining data called for on each record were entered. The tracing was marked with arrows to indicate its direction. Observations on the physical condition of the animal and others of importance, which were made from time to time, were also noted on the record. All the records of a single animal were then filed away together, thus affording a permanent record for further reference. Since it was desired to make a quantitative study of circus movements, it was necessary to adopt some method whereby the amount of turning exhibited by an animal, in a given trial or group of trials, might be expressed as a single value. These values have been stated in terms of average number of degrees turned per centimeter of progress, and were obtained in the fol- lowing manner. The length of the trail was first measured with a map tracer. Several readings were taken until two were ob- tained with a difference of less than 0.3 em. These were then averaged, and the result used in computations. Thus in figure 3, the length of the or’ginal tracing is 26.95 em. Since, however, the records were on a scale of 1 to 6, and these in reproduction have been reduced one half, the length of the text figure tracing must be multiplied by (6 x 2 =) 12 in order to obtain the dis- tance actually traveled by the animal. The various turns or angular deflections of the trail were next estimated by reference to the radii of the plan. It is obvious that in traveling a curved course, the direction of locomotion at any given instant is the tangent to the curve at that point. For example, in figure 3 the initial direction of locomotion is shown by the tangent ata. This direction is parallel to aradius. From a, the tangent to the curve or the direction of locomotion rotates continuously to the left until the point b is reached. At 6 the tangent is parallel to a second radius, which makes with the radius of initial parallelism an angle of § of 360° or 180°. (Each radius forms angles of 45° or + of 360° with its adjacent radii.) In other words, in traveling from a to 6 the axis of the animal’s body has rotated 180° to the left, or the animal has executed 4 of a complete sinistral loop. Similarly, from b to c the course of the animal makes 12 dextral loops; from ¢ to d, } of a sinistral 356 DWIGHT E. MINNICH loop, and finally from d to e, 13 dextral loops. The total amount of turning, or angular deflection, toward the right in this trail is, therefore, 1{ + 1§ or 2§ X 360°, while that to the left is 4 + 4 or 3 xX 360°. ; Since the honey-bee is positively phototropic and in this case — the left eye was blackened, the angular deflection toward the right or functional eye is designated as positive; that toward the left or covered eye, as negative. The algebraic sum of these Animal Pepa wullifine Experiment No. 12 ie 2 No. of Animal 3 eG i bes ra Times st" BoM. 30 a <2 pie Eye black Lape *<” Light 24 me No. dx. loops 1X+ 14 =21%~_ No. sn. loops 4 + K= % >> Trail fength, cm. 21.9 SUF ky. Fig. 8 Record or bee no. 123 in non-directive light. >+ pales angular deflections will give a result equivalent to the amount of continuous turning required to carry the animal from the start- ing point to the end of its course. Thus, in figure 3, the direction of locomotion at a makes with the direction at e an angle of 27 x 360° — 3 x 360° or 23 X 360°. Knowing the distance traveled in centimeters and the amount of turning in degrees, the average degrees turned per centimeter is easily computed. Denoting this average deflection, as I shall eall it, by D, we have for the trail in figure 3, PHOTIC REACTIONS OF HONEY-BEE 35 Vf 21 x 360° a 26.95em. X6 | + 5.01°/ em. D= It is to be emphasized that the value + 5.01°/cem., does not sig- nify that the animal turned only toward the functional eye. It merely shows that the algebraic sum of all its deflections aver- ages 5.01°/cm. toward the functional eye. Animal Apis metlifne Experiment No. 13 No. of Animal 5 2 Date 40K L ase Eye black Lept<” Light CleilMacresd POS No. dx. loops a % bR~~_ No. sn. loops a % b X¥ >> Trail Jength,em. a j23 b-y Ths he Th - - ~ ~ —— > == eee Al al ay Fig. 4 Two trails of bee no. 135 in non-directive light. A record is shown in figure 4 which represents two trials taken in rapid succession. This was necessitated by the animal’s en- countering the side wall of the light chamber so quickly that the — first trial was shorter than usual. The deflections in these trails are estimated as previously described. It will be noted, however, that in the trail marked a, the angular deflection between m and n does not amount to quite ? of a sinistral loop, although it is so counted. In such instances the angle was always estimated THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 29, NO. 3 308 DWIGHT E. MINNICH to the nearer $ of a circumference, no attempt being made to discriminate differences of less than 45°. In this record it is desirable to combine both trails into a single computation. Pro- ceeding as before, bye a7 38 OPE 60 2110 eee The negative sign of the average deflection here obviously indi- cates that the bee turned more toward the covered eye than toward the functional eye in these trials. In the course of experimentation, records of normal bees were also made in non-directive light. Since in such individuals neither eye was blackened, the positive sign was arbitrarily applied to the direction of greater angular deflection in each set of trials. Other- wise the computations for normal bees were performed in the same manner as those for bees with one eye blackened. These various examples will illustrate the method employed in all quantitative determinations. Upon the results thus obtained the chief conclusions of the present paper are based. IV. MATERIAL 1. General care of animals The bees used in all quantitative experiments were thoroughly active workers taken from the flowers of a near-by garden, and were, for the most part, individuals from a single large hive situ- ated there. The animals were trapped by simply inverting a long glass tumbler over the flower, and then transferring them to a small screen fly-trap. In some experiments, however, which were performed too late in the fall to obtain bees in this way, animals were used from a single comb of workers confined in an observation hive. The exit of the hive was kept securely screened, for such a colony quickly disintegrates if its members are per- mitted to leave the hive freely. Bees kept in this way remained in reasonably good condition, for at least a month. Bees destined to undergo experimentation were first subjected to having their wings clipped, an operation easily executed when PHOTIC REACTIONS OF HONEY-BEE 359 the animals were feeding. Each wingless individual was then confined in a small cylindrical cage of screen wire, the bottom of which was formed by a layer of tissue-paper over cotton to pre- vent injury in case of falling. In the same cage were also placed two friendly winged workers to counteract any possible effects of isolation. The cages of bees were kept in a darkened box when not directly under experimentation, since the influence of light often caused the animals to maintain a restless activity which appeared, in some cases, to shorten life considerably. In the dark, however, they usually remained more quiet. Each cage was supplied with water by a small wad of saturated cotton placed on its top. Small quantities of honey were also supplied on short wooden sticks stuck to the side of the cage. Early in the morning, at noon, and in the evening the cages were cleaned by removing excess honey, etc., and fresh honey and water were provided. Such operations were carried out at least a half-hour before any trials were made on the animals. The temperature of the laboratory in most of the experiments was kept above 20° to 21°C. This was found to be an important consideration, since at lower temperatures bees became torpid and inactive. In collecting the animals even, an attempt was made to take them on warm, sunny days which had, in general, been preceded by warm weather. It was found that bees brought in after a brief period of cold, wet. weather were apt to be either unresponsive or extremely variable in their behavior. 2. Blackening the eye Any technic for blackening the eye of a wingless bee requires, of course, the use of an anaesthetic. In the present experiments ether was used exclusively. Care was taken to administer it rather slowly and in minimal doses. When completely anaes- thetized, the bee was placed on one side, on a small cork pinning board. Here it was fastened down securely by the use of insect pins, with which the thorax, abdomen, and legs were securely braced against the cork. The blackening was then applied to the eye, the entire surface being covered with as thick a coat as pos- 360 DWIGHT E. MINNICH sible. ‘Two kinds of blackening material were used, viz., lamp- black in shellac and a dead black paint known commercially as ‘Jap-a-lac.’ The latter proved the more satisfactory and was used throughout the majority of experiments. Although bees under ether often began to recover in five to ten minutes, they were not removed from the pinning board for twenty to twenty-five minutes, when the covering of the eye was well hardened. Recovery from anaesthesia was usually com- plete in an hour and often much less. As a rule, however, opera- tions were carried out in the evening, and the bees were not sub- jected to further experiment until the following day. Ample time was thus allowed for the animals to recover as much as possible from the effects of the operation. V. BEHAVIOR OF NORMAL BEES 1. Kinetic effect of light The remarkable sensitivity of the honey-bee to photic stimu- lation must have long been patent to students of its behavior. Bethe (98, p. 83) says, ‘‘Das Licht ist bei diesen Tagthieren [bees, flies, etc.] der auslésende Reiz zum Fliegen; in einer dunklen Schachtel fliegt keine Biene auf, auch nicht, wenn man sie reizt. Das Licht gibt die Regulirung beim Fluge ab.’’ This observa- tion was repeatedly confirmed in the present experiments. When collecting bees from flowers, fifteen to twenty individuals were confined in a single cage, which was then placed in a closed box. Although at the height of activity when captured, a few minutes in the darkness of the box seldom failed to reduce these animals to a state of quiescence. If a little light was admitted to the box, however, by even partially removing the lid, there was a sudden resumption of activity. Precisely the same behavior was exhibited by wingless bees. If confined in a dark box, they were, as a rule, reduced to com- parative inactivity. A brief exposure to light, however, was usually sufficient to excite vigorous locomotion, and continued exposure not infrequently resulted in the maintenance of an intense activity for extended intervals of time. | PHOTIC REACTIONS OF HONEY-BEE 361 Individuals which had been subjected to operations of remoy- ing the wings and blackening the eye frequently responded some- what more slowly to this photic activation than did normal bees. In the former, locomotion was preceded by a more or less prolonged sequence of cleaning operations. The proboscis was extended and stroked with the fore legs. The eyes, particularly the covered one, were the objects of repeated and vigorous scrap- ings, responses no doubt largely attributable to the irritation of the blackening material. The abdomen was meanwhile bent from side to side, while the middle or hind legs were rubbed to- gether, or the hind legs assiduously stroked the dorsum of the abdomen. ‘These movements became more and more intense until at length they culminated in active creeping. Light, then, exerts a strong activating or kinetic influence upon the honey-bee, while darkness has the opposite effect. Es- sentially similar phenomena have been reported by Loeb (’90) for the plant louse, Carpenter (’05) for the pomace fly, and Turner (’12) for the mason wasp. Stockard (’08) has reported the case of Aplopus, the ‘walking-stick,’ which also falls into this category of behavior. In Aplopus, however, light inhibits ac- tivity, while darkness induces it. Hence the ‘walking-stick’ is nocturnal, whereas the plant louse, the pomace fly, the mason wasp, and the hive bee are diurnal. In diurnal animals this response is apparently due to the con- tinued action of light rather than a sudden change in it. Thus, while many bees respond almost, if not quite, at once to the presence of light, others may respond only after some minutes of exposure. According to Turner (12, p. 360), the same is true of the mason wasp. ~ 2. Directive light Not only does light induce locomotion in the honey-bee, but directive light regulates the course of locomotion. Bees brought into the laboratory direct from their foraging activities out of doors seldom failed to exhibit a most striking phototropism. Such insects when liberated in the laboratory flew almost im- mediately to the nearest window, where they remained fluttering 362 DWIGHT E. MINNICH against the glass. Or, if escaping in a darkened room, they not infrequently flew directly into the flame of the nearest gas jet. Observations of this sort were long ago reported by Lubbock (82, pp. 278, 279, 284). A few years later, Graber (’84) demon- strated the same thing experimentally by confining forty to sixty bees in a small box, one half of which was illuminated by direct sunlight, the other half being shaded, with the result that the majority of the bees soon collected in the illuminated end. More recently, Hess (’13 a, 713 b, ’17) has repeated this and a variety of other experiments. -As a result of these he has been able to show that in the presence of several sources of photic stimulation, which differ in color and intensity, bees always orient toward the one which to a totally color-blind person appears brightest. The positive phototropism of the honey-bee is thus demonstrable in a variety of ways. In the experiments just cited, winged bees were used exclu- sively. My own experiments, on the contrary, were confined en- tirely to workers from which the wings had been clipped. Such bees when creeping in the directive light area exhibited an orien- tation which was striking in three respects, viz., its rapidity, its precision, and its constancy. An individual to be tested was removed from the dark box and exposed to light for a few minutes until it was thoroughly active. It was then allowed to creep from its screen cage to a small, rectangular piece of black paper, and on this it was transferred to the edge of the directive light area. An effort was made to start the animal creeping at a right angle to the direction of the light rays by turning the paper just before it crept off. The rapidity of orientation was so great, however, that the intervening centi- meter or so was frequently sufficient to allow the animal to reorient perfectly. Since the velocity at which bees creep averages 3 to 6 em. per second, orientation in these cases occurred in considera- bly less than one second. I have also tried leading a bee by moving the light, now in this direction, now in that, with varying degrees of curvature. Always the animal followed, orienting rap- idly to even slight movements of the lamp. PHOTIC REACTIONS OF HONEY-BEE 363 The precision with which orientation was maintained was no less conspicuous. Once oriented, the animal generally moved in a nearly straight line toward the source of light. In figure 5 are shown two records of each of six bees in the directive light area. Of a large number of animals tested in the course of ex- perimentation, considerably over 25 per cent maintained their orientation as precisely as did bee no. 66. The deviations of 66 a J / \ le / an \ if 110 / 21 36 a o Fig. 5 Two trails of each of six normal bees in directive light. In this, as in subsequent figures of records in directive light, the clear circle represents the light source, and the straight lines from it, the direction of the rays. most of the animals would, moreover, easily fall within the lati- tude of that exhibited by bees nos. 66, 33, and 23. Results sim- ilar to those shown for animals nos. 110 and 21 were, on the con- trary, less frequently encountered, while trails such as those of bee no. 36 were seldom or never found among normal, healthy bees. The response to directive light is very constant in the bee. The oncoming of death itself seems often to intensify rather than to weaken this phase of its behavior. Bees occasionally escaped 364 DWIGHT E. MINNICH in the laboratory. Such individuals rarely survived the lack of food for more than a day or so. Yet it was not an infrequent occurrence to observe one of these starved animals, so weak that it was barely able to creep, slowly emerging from a hidden corner in a final struggle toward the light. Nevertheless, bees were discovered which in a few instances failed to exhibit the usual positive reaction to directive light. Such cases, however, are not to be construed as a total absence Fig. 6 Three successive records of a normal bee in directive light, showing a failure to orient in two cases. of phototropism, but rather its momentary suppression by other factors of behavior. This is well illustrated by the following example. Seven cages of bees were prepared from the stock in the observation hive, Oct. 30, at 2:45 p.m. When tested about an hour later in the directive light area, six of the seven animals exhibited the usual positive response. One animal, however, gave the records reproduced in figure 6. This bee when given its first trial at 4:06 (fig. 6, 1) did not ori- ent toward the light source. Instead it pursued a devious course PHOTIC REACTIONS OF HONEY-BEE 365 looping now to the right, now to the left, and finally turning al- most directly away from the light. In a second trial at 4:14, it exhibited a somewhat similar response (fig. 6, 2). One minute later, the animal was subjected to still a third trial, being started on this occasion some 30 cm. nearer the light. This time it ori- ented and moved in a fairly direct course toward the source of illumination (fig. 6, 3). What the temporary, inhibiting factors were which produced these very atypical responses could not be ascertained. In all other respects this bee was quite indistin- guishable from the other individuals in the experiment. This example, however, shows that even the constant response of the bee to directive illumination is not free from abrupt and appar- ently inexplicable departures. 3. Non-directive light The behavior of bees in non-directive light is no less charac- teristic than that in directive illumination. Since all quantita- tive experiments on circus movements were conducted in non- directive light, an intimate acquaintance with the behavior of normal animals under the same conditions was necessary. Every bee was, therefore, subjected to several trials in non-directive light before having one eye blackened. It was a matter of continual observation that a bee creeping in the directive light area ceased to move in a straight course upon reaching the area near and immediately beneath the lamp. Here, where the illumination was essentially non-directive, the animal deflected from its former, precise path and began to loop in a constant or varying direction (fig. 1). In other words, the bee was trapped; for directly it crept away sufficiently for the light to become directive again, it was forced to turn back. ‘Thus the animal continued to creep round and round in a limited area, occasionally rearing on its hind legs in an abortive attempt at flight, or finally ceasing locomotion to begin cleaning operations. In the non-directive light apparatus (fig. 2), the same tend- ency to loop was manifested, only on a much more extensive scale. Here the bee seldom crept in a straight line for any great 366 DWIGHT E. MINNICH distance. Each animal was subjected to two sets of trials, an hour or so apart. Usually a single trial only constituted a set. In case the bee quickly encountered the side wall of the light chamber, however, or exhibited unusual variability in its be- havior, additional trials were made. The aggregate duration of the trials of each set varied considerably, even in the same animal. Sometimes they were as short as thirty seconds; again, as long as two minutes. The average was in the neighborhood of thirty to sixty seconds. Preliminary to each trial, the bee was exposed to light until aroused to active creeping. The illumination used throughout in experimenting with normal bees was 957 mce.° The average deflection to the right or to the left has been computed for each set of records thus obtained, and the results of these computations presented in table 2 (appendix), columns B and C. On the basis of these data, the fifty-two bees experi- mented upon may be classified into three groups: 1. Bees whose average deflection in both sets of trials was over 2°/em. and in the same direction. 2. Bees whose average deflection in both sets of trials was small. 3. Bees whose average deflection in the two sets of trials varied widely, either in magnitude or direction, or in both. The first class is composed of animals which exhibited a more or less pronounced tendency to turn in a constant direction (right or left). These animals, 29 in number, comprised 56 per cent of the total 52 bees. Fourteen of these were chiefly right-handed in their turns; 15, left-handed. In 14 of the 29 bees, or 27 per cent of the total number, the average deflections exceeded 4°/cm., while in 6 individuals, or 12 per cent, itrose toover8°/em. Sim- ilar right and left-handed tendencies of locomotion in non-direct- ive light have been reported by Walter (’07) for planarians and by Patten (’14) for the blowfly larva. A typical example of this behavior in bees is illustrated in figure 7, bee no. 101. In its first trial (fig. 7, 101, a), this animal showed an average deflection of 7.11°/em. to the left, and in the second trial (fig. 7, 101, b), a sim- ilar deflection of 6.10°/em. Since these records were made nearly an hour apart the left-handed tendency was not the result of a PHOTIC REACTIONS OF HONEY-BEE 367 10] Left Left a b 36 Left - Right Left - Right ays a Left Right | » b 2 Fig. 7 Records of normal bees in non-directive light. In this, as in subse- quent figures of records in non-directive light, a solid circle is used to indicate the center of the floor of the non-directive light apparatus. a, records of the first set of trials; b, records of the second set of trials. Bee no. 101 deflected constantly toward the left. Bee no. 36 varied its deflection in the course of single trials. 368 DWIGHT E. MINNICH brief, temporary condition, but was probably a more or less permanent feature of this animal’s behavior. The second class of animals includes those whose average de- flections were small in both directions. The results obtained here are attributable to either of two causes: a. The bee varied its turning from right to left, so that on an average, one tendency nearly or quite balanced the other. b. The bee exhibited little or no tendency to turn either to the right or to the left. . An example of the first type is seen in the records of bee no. 36, figure 7. In its first set of trials (a, 1, 2) this animal turned sometimes to the left, sometimes to the right, so that the result- ant average deflection was but 0.79°/cm. to the left. Similarly in the second set of trails (b, 1, 2), the average deflection amounted to only 1.94°/em. to the right. The second type of this class is illustrated by bee no. 134, figure 8. This animal showed no pro- nounced tendency to turn either to the right or left. The aver- age deflection for each set of records was, therefore, small, being only 1.52°/em. to the left for the first set (a, 1, 2, 3) and 1.22°/cm. to the left for the second set (b, 1, 2, 3, 4). In the third class of bees are to be found those which, although they exhibited fairly uniform behavior in a single set of trials, varied widely in different sets. For example, bee no. 82, in its first record (fig. 8, 82, a) showed a pronounced deflection which averaged 5.81°/em. to the left. In its second set of trials (fig. 8, 82, b, 1, 2, 3, 4), on the contrary, it showed little tendency to turn, and the average deflection was but 0.65°/cm. to the left. An even more striking case of variation, however, was afforded by bee no. 63. In a single record of fifty-three seconds’ duration (fig. 8, 63, a) this animal deflected, on the average, 5.58°/em. to the left. Approximately two hours later, in a record of sixty seconds’ duration (fig. 8, 63, b), the same animal exhibited an even greater average deflection in the opposite direction, viz., 7.50°/em. to the right. The range of variation presented by these two records is no less than 13.08°/em. In a uniform, non-directive light field, therefore, many bees exhibit a fairly constant tendency to turn toward a given side, PHOTIC REACTIONS OF HONEY-BEE 369 Naa Ny / a a b Fig. 8 Records of normal bees in non-directive light. a, records of the first set of trials; b, records of the second set of trials. Bee no. 134 exhibited little tendency to deflect in either direction. Bees nos. 82 and 63 varied widely in their average deflections in the two sets of trials. 370 DWIGHT E. MINNICH others display little or no such tendency, while still others vary widely in their deflections from time to time. Since the animal moves in a uniform environment, the conspicuous asymmetry of response so frequently noted must be attributed to internal factors. Such factors are, for the most part, probably quite in- dependent of light. A more detailed discussion of these will be presented in a subsequent section of this paper. 4. Total darkness If internal factors are responsible for the asymmetric responses of bees in non-directive illumination, a similar behavior should be exhibited in the total absence of photic stimulation. Such was indeed the case. Animals creeping on smoked paper, in total darkness, showed the same conspicuous tendencies to loop and turn as did animals in non-directive light (fig. 16). The data here referred to were taken in connection with experiments conducted for a different purpose. They are, therefore, not sufficiently extensive to establish more than a similarity to the behavior exhibited in non-directive light. Responses essentially like those of bees in total darkness have also been described by Pouchet (’72) for the larvae of Lucilia caesar, Davenport (’97) for the amoeba, and Frandsen (’01) for the garden slug Limax. Frandsen’s observations in particular bear a striking resemblance to those which I have just described for the honey-bee in non-directive light. Thus he found that while most of his animals looped in a fairly constant fashion to the right or left, a few were extremely variable, while still a few others moved in rather straight courses. The responses of creep- ing bees in the total absence of photic stimulation are, therefore, very similar to those observed for other animals under the same conditions. 5. Summary In the preceding pages certain responses of normal bees have been described in considerable detail, but only as a prerequisite to an adequate understanding of the behavior of the animals PHOTIC REACTIONS OF HONEY-BEE ote when one eye is blackened. The features of behavior which are important in this connection may be summarized as follows: 1. In the honey-bee, light tends to induce activity; darkness, to inhibit it. This response is dependent upon the continuous action of photic stimulation. 2. Isolated worker bees in an active condition exhibit strong positive phototropism when flying or creeping. ‘Temporary sup- pressions of this response may occur, however. 3. Normal bees when creeping in non-directive light usually exhibit pronounced asymmetrical responses of constant or vari- able index. Since essentially the same responses occur in total darkness they are not fundamentally dependent upon photic stimulation. They are probably, therefore, conditioned largely by internal factors. VI. BEHAVIOR OF BEES WITH ONE EYE BLACKENED 1. Directive light The previous investigations of circus movements have pointed unmistakably to the generality of these responses among photo- tropic arthropods. Positive animals with one eye covered tend to circle toward the functional eye; negative animals, under the same conditions, tend to circle away from the functional eye. The honey-bee exhibits a striking positive phototropism. When one eye is blackened, therefore, we should expect the bee to circle toward the remaining functional eye. Such is indeed the case, as Axenfeld (99, p. 374) has previously shown. In my own experiments, bees thus operated upon were no longer able to creep in a straight course toward a source of il- lumination. Instead, their progress thither was marked by re- peated loops. If the right eye was blackened, the bee looped to the left; if the left eye was blackened, it looped to the right. Moreover, it was possible by blackening one eye, then removing the black and blacking the other eye, to cause a single individual to perform circus movements first in one direction, then in the opposite direction. 372 DWIGHT E. MINNICH The above experiment was carried out on five bees. In gen- eral, all of these animals looped more or less markedly toward the functional eye as they crept toward the light source. This tend- ency, moreover, was not confined to the period immediately subsequent to the operation of blackening the eye, as the experi- ment clearly demonstrates. The first records of these bees with their left eyes blackened were taken in the evening between 6 and 7 p.m. No further tests were made until the following day at 11:30 a.m. Yet the behavior at the end of this seventeen- hour period was practically the same as it had been before. One bee, it is true, showed considerable improvement. In the other four animals, however, the two sets of records were indistinguish- able. In the absence of experience, therefore, the performance of circus movements remains a permanent feature of behavior. Of the five bees tested, the most pronounced and uniform exhi- bition of cireus movements was displayed by bee no. 5. Its records are almost diagrammatic in their close approximation to the theoretical expectation. Records of this animal are repro- duced in figure 9. Not all of these animals, however, yielded such striking results. Some individuals were found which manifested little or no tend- ency to deviate toward the functional eye, except in the area immediately beneath the lamp, where the illumination was es- sentially non-directive. Thus, bee no. 4, with its right eye black- ened, circled toward the left in the usual manner. But a few hours later, when the black had been removed from the right eye and the left eye painted over, it exhibited little or no tend- ency to circle toward the right (fig. 10, B). The explanation at once suggests itself, that in such cases the eye was imperfectly covered, and hence not absolutely free from stimulation. This may be correct. As will be shown later, however, there are also a variety of other circumstances which might account for such behavior. The tendency to circle toward the blackened eye was not fre- quently encountered in the reactions of bees to directive light. No instance of it occurred in the experiment described above, al- though it was occasionally met with in other experiments. A PHOTIC REACTIONS OF HONEY-BEE 373 ‘4 ip Normal Right Eye Black | I @ Left Eye tg U \ x Ps M4 Fig. 9 Consecutive records of a bee in directive light, showing the effect of blackening first one eye, then the other. Left Eye Black THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 29, NO. 3 374 DWIGHT E. MINNICH single record of this kind is shown in figure 10, A. This was ob- tained from an animal which subsequently became wholly unre- active. Its aberrant tendencies may, therefore, have been due to an abnormal condition. In any case it is significant that, al- though the bee looped toward the covered eye, yet it progressed toward the light source. Consequently, this was not a ease of reversal of phototropism. Instances somewhat similar to the one last mentioned have been described by Dolley (16, p. 373) for the butterfly Vanessa. | Neg * Fig. 10 A. Record of a bee in directive light, showing loops toward the blackened eye. B. Two records of a bee in directive light which showed no deflection, although the left eye was blackened. Although positive to light, this insect with one eye blackened occasionally turned toward the covered eye instead of toward the functional eye. Possibly results of this sort are to be attributed to the effect of contact stimulus afforded by the covering of the eye, as indicated by Dolley (’16, pp. 394-399). This will be dis- cussed more fully in a subsequent portion of the paper. 2. Non-directive light a. Amount of turning. In non-directive light, bees with one eye blackened tended, in general, to turn more or less pronouncedly toward the functional eye. As was to be expected, the course PHOTIC REACTIONS OF HONEY-BEE 375 taken by the animals under these conditions assumed no specific direction. They either continued to circle in a fairly limited area or proceeded in a looping fashion in any direction whatever. The variability of this response, moreover, was much greater than in the case of the response to directive light. Thus, a num- ber of animals circled almost continuously toward the covered eye in non-directive light, while still others varied, circling some- times toward the covered eye, sometimes toward the functional eye. This was doubtless true for much the same reasons that normal bees also exhibited a greater variability of response in non-directive light. Circus movements attendant upon the elimination of one pho- toreceptor undoubtedly represent the orienting process of an asymmetric animal. The specific photic stimulus, therefore, which produces these reactions must be identical with that which effects orientation in the normal animal. Whatever the nature of this stimulus be, moreover, it is afforded by both directive and non-directive light, since circus movements occur in either situation. What is the nature of this orienting stimulus? Per- haps the best method of demonstrating the dependence of a par- ticular response upon a certain stimulus is to show that the in- tensity of the response varies with the intensity of the stimulus in question. It seemed possible to attack this problem, therefore, through a study of the relationship existent between the amount of turning displayed by an animal with one eye blackened and the intensity of the illumination to which it was subjected. Non-directive illumination was chosen in preference to direc- tive illumination because of the simpler experimental conditions which the former affords. In directive light, every movement of the entire animal is accompanied by more or less complicated changes not only in the intensity of the stimulation received, but also in the area of the eye stimulated. Asan animal with one eye covered moves toward a light source, the stimulation of the functional eye steadily increases. As it loops toward this eye, however, this steadily increasing stimulus is subjected to rapid and transitory fluctuations. When the animal begins to loop, the functional eye is first turned away from the light, resulting 376 DWIGHT E. MINNICH in a rapid decrease of photic stimulation. As the loop is com- pleted, the photoreceptor in turn experiences an increase in stimulation. In non-directive illumination such as was employed in the present experiments, however, these complications are largely avoided. Photice stimulation here is maintained at a fairly uniform and constant intensity over the entire surface of the compound eye. Two slightly different types of experiment were performed. The procedure in the first type was as follows. Bees were col- lected from flowers in the morning between 8 and 10 o’clock, brought into the laboratory and prepared for experimentation. One and two hours later, respectively, they were given single trials in the directive light area. On a basis of these records, individuals of abnormal tendencies were discarded, and those evinc- ing the greatest accuracy of orientation were selected. An hour or so later, the selected bees were tested in non- directive light of 957 me. Two sets of records, about one hour apart, were made of each animal. Each set was composed of one to several records, the aggregate duration of which, in gen- eral, was between thirty and sixty seconds. An examination of the records showed clearly whether the individual was normally right-handed, left-handed, or variable in its deflection in non- directive light. These results determined which eye should be blackened. If, for example, a bee normally circled to the right, the right eye was covered. Whatever influence was exerted by photic stimulation, therefore, would tend to force the animal toward the left. In this manner, responses which might other- wise have been mistakenly attributed to photic stimulation were to some extent eliminated. The operations of blackening the eye were carried out in the late afternoon of the first day of experimentation, in accordance with the technic previously described. On the following morn- ing, before resuming experimentation, it was not infrequently necessary to discard a few additional animals either because of extreme weakness or occasionally death as a result of the operation. The majority of bees usually appeared quite normal, however, and were subjected to several series of trials in non-directive PHOTIC REACTIONS OF HONEY-BEE 377 light. Throughout a single series of consecutive trials, or, as I shall call it, a determination, one intensity of light only was employed. But in the total number of determinations the more intense illumination of 957 me. and the less intense of 24 me. were used an equal number of times. ‘The animal to be tested was first removed from the dark box and exposed from half a minute to several minutes in the intensity of light in which it was to be tried. This was usually sufficient to activate the animal thor- oughly, and several records were then made in the non-directive light chamber. In case the bee failed to respond to photic acti- vation, recourse was had to mechanical stimulation. The cage was tapped or even shaken fairly vigorously until locomotion was induced. This precedure seldom failed to elicit activity. When it did fail, it was usually necessary to discard the animal altogether. The number and duration of the records comprising a single series or determination varied widely even in the same animal. If the bee quickly encountered the side wall of the light chamber, records were short, and a number had to be taken. If, on the contrary, the animal kept well toward the center of the floor of the apparatus, one or two records were quite sufficient. In cases of great variability of response or unusual departures from the general, expected behavior, additional trials were made, on the assumption that a greater number would more accurately ex- press the average tendency of the animal. Single trials seldom exceeded thirty seconds, and were often much shorter. Occa- sionally, however, records of forty-five seconds, sixty seconds, or even slightly greater durations were taken. The aggregate duration of the trials comprising a single determination, for one intensity of light, was usually in the neighborhood of thirty or sixty or ninety seconds. The adoption of any more uniform period for all animals, at all times, was quite impossible. Upon completion of a series of trials in one intensity of light, the bee was returned to the dark box. Here it was allowed to remain for a period of about fifteen minutes to one hour. In the earlier experiments the longer period was practiced; in subsequent experiments, the shorter. After this period in the dark, the ani- 378 DWIGHT E. MINNICH mal was subjected to a second set of trials of the same aggregate duration as the first, but in the other of the two light intensities. The order in which the two intensities of illumination were em- ployed was varied from time to time. Sometimes the first de- termination was made in the more intense light; the second, in the less intense. Sometimes the reverse order was observed. A single series of records in one intensity of non-directive light together with the corresponding series in the other intensity con- stitute what I shall term a pair of determinations. The protocol of such a pair of determinations on bee no. 42 is given in table 3. Four or five pairs of determinations were usually made on each in- dividual of an experiment in the course of a day, beginning TABLE 3 DETERMINATION FOR NON-DIRECTIVE LIGHT DETERMINATION FOR NON-DIRECTIVE LIGHT OF 24 Mc. OF 957 Mc. Number of Duration of Number of Duration of Fee Pee Hour of record aaanaey panei Hour of record randiel seconds seconds 4 1:47 p.m. 30 1 1:32 p.m. 31 5 1:47% p.m. 30 2 1 :322 p.m. 30 6 1:48 p.m. 30 3 1:38 p.m. 30 MTotalsy.::,: eyes eee 90 91 between 8 and 9 o’clock in the morning and concluding between 4 and 5 in the afternoon. The bees often seemed to become slug- gish in the late afternoon. Whether this was due to fatigue or a natural rhythm of activity from day to night, I am unable to say. This phenomenon, however, led me to abandon any attempt to continue experimentation much after 5 o’clock. On the third and concluding day of the experiment, the scheme of the second day was again carried out as far as possible. Bees usually survived the first two days of experimentation, and in case they did not, the data on them were discarded. A number of individuals, however, failed to survive in fit condition for the trials of the third day, and still others had to be discarded in the course of the day, although in both eases the results were counted. PHOTIC REACTIONS OF HONEY-BEE 379 Some of the more vigorous animals survived not only a third day of experimentation, but lived on for three or four days, and in a few instances even longer. Although no further trials were made with such bees, they were kept and, as far as possible, records of their subsequent longevity taken. Having described the first type of non-directive light experi- ment in considerable detail, the second type may be described very briefly. It differed from the first only in the method of making pairs of determinations. In this case, the two determi- nations of a pair were made during the same period of time, TABLE 4 DETERMINATION FOR NON-DIRECTIVE LIGHT DETERMINATION FOR NON-DIRECTIVE LIGHT oF 24 mc. OF 957 Mc. Number of Duration of Number of Duration of ieirieredl Hour of record er aiid Hour of record Sanne seconds seconds 1 1:41 p.m. 30 2 1:43 p.m 11 3 1:48 p.m. +4 4 1:51 p.m 23 5 1:53 p.m 40 6 1:56 p.m 30 7 1:59 p.m 30 | Watal ss. so74 3 As aS | 104 | 104 instead of an appreciable interval apart. The bee was first tested in one intensity of light, then within a minute or so in the other intensity, then again in the first, and so on untila series of one to five records had been completed for each intensity. Care was exercised, however, even with this rapid alternation of intensi- ties, always to expose the animal for thirty to sixty seconds in a given intensity before subjecting it to a trialin the same. The following protocol from bee no. 83 (table 4) will illustrate this method of making determinations. In both types of experiment, there were obtained for each bee a number of pairs of determinations, usually four to ten, depend- ing upon the longevity of the individual. The records of each 380 DWIGHT E. MINNICH determination have been computed collectively in the manner already described. Single values have thus been derived which express the average deflection, or tendency to turn, exhibited by the animal in each determination. When the turning was chiefly toward the blackened eye, the sign of these values is negative; when chiefly toward the functional eye, it is positive. If, now, the value of each determination in 24 me. light be subtracted from the corresponding one in 957 me. light, differences will be obtained which should answer conclusively the question of rela- tionship between the amount of turning and the intensity of photic stimulation. The differences obtained in the manner just described I shall designate asd. oO Sa PHOTIC REACTIONS OF HONEY-BEE XI. APPENDIX (TABLE 2)—Continued A B c D E F. G NORMAL | NORMAL ee a av saaie suiek ay.” + VALUEs |— VALUES NUMBER IN RIGHT | TO LEFT PER CM. PER CM. OFE—D/} OFE—D OF BEE 957 Mc. 957 uc, | TURNED 24 |TURNED 957, ord ord LIGHT TnGrEn ||) Mer eee | MC. EaSEr 42 + 8.71 | +11.20 2.49 + 8.45 | +10.47 2.02 + 9.49 | +12.47 2.98 a 78 |) Loe Oo tk 22 + 9.03 | +13.70 4.67 +11.18 | +10.23 0.95 + 9.34 | =-13748 4.14 43 2.05 — 9.99 | — 2.55 7.44 ell? — 4.44 | —12.61 8.17 — 8.48 | — 8.48 0.05 — 9.57 | —11.06 1.49 44 0.00 + 2.04 | +16.10 14.06 0.57 | + 1.438 | +138.73 12.30 + 4.53 | +10.36 5.83 + 2.05 | + 8.76 6.71 45 5.20 + 1.97 | + 2.39 0.42 8.01 + 1.03 | + 3.80 ili No rec- | —11.17 ord — 4.68 | — 8.76 4.08 51 3.41 | — 4.49 | + 1.48 5.92 3.75 | — 3.96 | — 0.44 3.52 + 2.87 | + 9.05 2.18 — 8.30 | — 5.05 3.20 52 5.00 | + 6.22 | + 9.06 2.84 2.35 + 2.37 | + 4.59 2.22 + 3.30 | + 5.02 ae + 8.44 | + 7.57 0.87 + 6.94 | +11.07 4.13 53 2.08 | — 5.65 | — 3.41 2.24 2.61 | — 9.92 | — 3.79 6.13 — 1.19 | — 6.19 5.00 — 9.72 | — 3.638 6.09 54 6.91 — 0.79 | + 8.23 9.02 H 417 I DURATION | DURATION OF REC- ORDS IN SECONDS 24 mc. LIGHT OF REC- ORDS IN SECONDS 957 Mc. LIGHT 111 418 DWIGHT E. MINNICH XI. APPENDIX (TABLE 2)—Continued a er . oe tN B Cc D EB F G H I NORMAL | NORMAL DURATION | DURATION BEE AV.° | BEE AV.° OF REC- | OF REC- NUMBER | PERCM. | PERCM. ONE EYE ONE EYE ° ° = BLACK AV.°| BLACK AYy.° |+ VALUES VALUES | opps IN ORDA Ir OF BEE | ToRIGHT| ToLEFT |_ PROM. FEE CM. .| OF ND) OF aD’! C scone ee ‘nome | rome | seuteur ['sectione) O°" 7 OS | ae | oars 54 3.11 + 5.77 | + 8.25 2.48 51 59 + 8.20 | + 6.63 SG y/ 50 31 + 3.37 | + 3.54 0.17 76 67 + 1.70} + 3.71 2.01 112 105 55 7.69 | — 0.35 | + 4.36 4.71 66 55 4.17} + 8.14] + 8.89 0.75 72 64 + 9.88 | + 7.89 1.99 84 119 + 8.64 | + 6.78 1.86 80 76 56 0.382 | + 1.50 | —12.86 14.36 61 64 1.45 | + 1.14 | —10.27 11.41 35 30 — 3.73 | + 1.838 5.06 ie 69 —10.56 | —18.47 7.91 100 115 62 1.38 + 1.55 | + 1.36 0.19 83 73 2.05 | + 4.54] + 2.98 1.56 133 122 + 1.88 | + 9.18 7.30 98 99 + 4.42 | + 8.71 4.29 84 80 +10.73 | +11.61 0.88 77 63 + 6.05 | + 6.33 0.28 110 110 + 5.79 | +10.79 5.00 42 40 + 5.58 | + 5.87 0.29 112 113 + 7.59 | + 7.45 0.14 116 120 + 4.50 | + 4.47 0.03 103 99 63 5.58 | + 0.48 | +11.83 35 52 49 7.50 + 3.94] + 8.56 4.62 78 72 +11.64 | + 2.61 9.03 35 38 +10.69 | + 8.45 2.24 48 50 + 4.13 | + 4.50 0.37 87 74 + 3.04] + 3.41 0.37 48 61 + 3.59 | +12.00 8.41 41 30 + 0.18 | + 8.74 8.56 56 60 + 2.99 |} + 9.21 6.22 50 48 + 7.00 | + 9.87 2.87 46 50 +12.58 | +14.74 2.16 60 54 + 5.59 | + 7.79 2.20 80 85 66 1.82 | + 6.78 | + 6.48 0.30 46 46 PHOTIC REACTIONS OF HONEY-BEE 419 XI. APPENDIX (TABLE 2)—Continued A B c D E F G H I NORMAL NORMAL DURATION | DURATION ONE EYE ONE EYE BEE AV.° | BEE AV.° ° ° OF REC- OF REC- BLACK AY. | BLACK Ay. VALUES |— VALUES NUMBER PER CM. PER CM. us ORDS IN ORDS IN PER CM. PER CM. OFE—D | OFE-—D OF BEE | TO RIGHT | TOLEFT |, penep 24|\rURNED 957| ord Gua SECONDS | SECONDS 957 mc. | 957mc. | ic rienr | mc. LIGHT 24 mc. 957 Mc. LIGHT LIGHT - . LIGHT LIGHT 66 0.18 | + 6.46 | + 4.07 2.39 85 93 aS | 4e2F 2.54 68 68 + 4.36 | + 8.27 3.91 54 55 + 7.40 | + 8.68 1.28 96 86 + 9.45 | +'9.87 3.58 60 60 + 5.23 | + 4.93 0.30 142 136 + 4.98 | + 4.60 0.38 100 104 “+. 3.02 |w-f-81209 8.07 83 80 + 5.59 | +10.32 4.73 101 122 68 6.08 + 4.35 | +13-12 8.77 97 94 1.94 +11.65 | +16.40 4.75 66 67 ac .2- 00) | 14-01 9.35 54 59 a 8-008 |-Fli.23 2.26 60 69 -e. 8.83) |°-+--9.82 0.99 87 100 +10.57 | +12.21 1.64 53 63 + 5.62 | +12.06 6.44 75 60 +10.48 | +13.52 3.04 75 87 + 5.32 | +11.79 6.47 91 99 + 4.37 | + 4.76 0.39 100 104 72 5.73 ++ 7.13, |o-89 711 1.98 60 60 7.30 + 1.02 | 424.35 | 23.33 20 35 +- 2.60) |G 70529 2.94 77 58 + .0.87 |periss 2.26 87 86 — 5.92 | — 0.66 3.26 103 100 +; 6.77, Ronee 0.55 120 120 +10.17 | +12.94 2.77 100 91 73 4.94) + 0.92 | + 8.27 7.35 39 40 2.24 | + 1.53 | + 4.22 2.69 34 34 — 1.42 | +13.04 | 14.46 60 65 + 2.45 | +10.89 8.44 100 100 + 2.00 | +12.55 | 10.55 61 75 + 3.29 | +13.79 | 10.50 93 101 +5 2. SBF 79.49 6.61 104 87 — 0.12); — 8.89 9.01 147 147 + 6.07 | +13.82 7.75 135 136 — 1.69 | 415.55 | 17.24 110 110 420 DWIGHT E. MINNICH XI. APPENDIX (TABLE 2)—Continued —_——_——_—_— aw Qa» eee A B Cc D E F G H I NORMAL NORMAL DURATION| DURATION ONE EYE | ONE EYE BEE AV.°| BEE AY.° 4 : OF REC- | OF REC- BLA ; oly ms NUMBER | PERCM. | PER CM. ee | buses AY VALUES|— VALUES! ops in | ORDS IN or BEE | TO RIGHT] TO LEFT | PPR CM. FER'CM. | OF B~ D| OFE.—)| seconps | seeoNNE sof we. | $97 wc. |"elaaeun | we uion?| °° | OFS | 24 ac. | 967 we. 77 4.22 410.71 | 4-84243 | 3.72 57 75 7.46 + 9.62] +16.59| 6.97 8s 65 +13.53.| +18.39 | 4.86 48 40 +12.43 | +14.74| 2.31 104 80 + 8.45] +14.92| 6.47 90 80 +11.64 | +16.89} 5.25 120 130 +12.60 | +15.38| 2.78 75 70 +15.29 | +17.98| 2.69 40 35 +11.57 | +18.36] 6.79 80 80 81 2.09 + £93 '|°-10 8000" 6.39 114 117 2.59 4 1203 |4-93 98 a) 2195 53 53 35 6.08!) 4-178 le 1.75 116 106 + 5.76} + 5.86] 0.10 108 101 + 3.30] + 8.35] 5.05 70 65 + 4.59 | 4°9.89| 5.30 56 61 + 5.03|+ 7.49] 2.46 63 63 +5 4 S10eis-52 | Siva 57 80 + 2.76 | + 8.21 | 5.45 110 122 4971 0105-12.60 | 8.89 67 67 82 5-81 | Sees) Ft .33 7] 0) Ser 124 119 0.65 | 2237 a= °5-01//0 0:36 110 123 =10730 || °— 6-431) 3.87 60 60 — 6.95 | — 4.88| 2.07 59 59 ae7859'| 4-1 ASHEN A. Ot 66 64 = 5.29 || 0-B7 aes 74 85 277-06 )| +4 2.96171 See 54 56 = 60] — 5.201) Wert 79 80 S80 || = 5247 4.67| 108 105 83 3.59 + 5.62 | +15.93 | 10.31 81 76 5.85 oe S287 1 3-15.98 07.38 90 90 + 8.89 | +10.53| 1.64 104 104 + 6.93 | +11.49| 4.56 71 71 +12.84 | +21.53 8.69 58 60 + 9.39 | +17.47 8.08 60 60 + 4.93 | +16.39 | 11.46 94 90 + 4.53 | +12.50 7.97 41 33 + 3.52 | +14.42| 10.90 133 128 PHOTIC REACTIONS OF HONEY-BEE 421 XI. APPENDIX (TABLE 2)—Continued A B Cc D E F G H I NORMAL NORMAL DURATION| DURATION ° o| ONE EYE ONE EYE BEE AV. BEE AV. c ° OF REC- OF REC- : LACK AV. ALU — NUMBER PER CM. PER CM. eee: | BUAG + VALUES eee ORDS IN ORDS IN PER CM. PER CM. OF E—D/| ORE-—D OF BEE TO RIGHT TO LEFT 9 SECONDS SECONDS es? uc. | 957 we. |"Yicluionr | we uicer | "| | O*% | 24 wc. | 957 atc. 85 0:84 | + 6.46 | +12.52 6.06 0.93 | + 1.41 | + 7.78 Grau + 7.08 | + 3.74 3.04 + 6.87 | + 6.42 0.45 + 4.34 |} + 3.54 0.80 + 1.73 | + 3.63 1.90 + 4.61 | + 8.19 3.58 + 6.13 | + 7.45 132 + 4.68 | + 7.99 3431! 91 0.48 — 2.42 | — 4.98 2.56 6.99 —10.05 | — 4.00 6.05 — 6.44] — 6.05 0.39 — 7.33 | — 3.05 4.28 + 2.18 |} + 3.52 1.34 + 0.73 | + 6.69 5.96 +10.19 | +27.54 eon 92 10.66 | + 4.42 |] + 5.77 ESS 8.67 | + 4.28 | + 9.86 5.58 + 7.41} + 9.19 1.78 + 9.64 | + 7.64 2.00 + 7.49 | +17.34 9.85 +14.19 | +23.82 9.63 93 0.21 | + 1.61 | + 6.48 4.87 2.44 + 0.19} + 1.42 ies: + 2.87 | — 0.46 3.300 + 2.37 | + 2.92 0.55 95 2.29 | + 2.14 | + 4.89 PATE 6.52 | + 1.13 | + 3.07 1.94 0.00 | + 5.67 5.67 + 0.32 | + 6.58 6.26 + 1.29} + 4.82 3.53 + 0.93 | +11.94 11.01 + 2.39 |} + 7.18 4.79 + 3.67 | + 7.11 3.44 + 0.85 | + 6.35 5.50 + 0.21} + 1.93 LZ THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 29, No.3 422 DWIGHT E. MINNICH XI. APPENDIX (TABLE 2)—Continued A. B c D E F G H I A xonnay,| NORMA | ons ere | oxn Exe >a | ee x * |BLACK : : ae NUMBER | PER CM. | PER CM. CK AV." BLACK AV.” |-+ VALUES VALUES | orpsIN | ORDS IN OF BEE |T0 RIGHT| To LeFT |_P2% CM; EDR CM. || OF B- D'| OF E— > | suconDs | SECONDS ssf a. | Tue. Geiaene | acans | | | eae 96 2.76 | +10.46 | +19.08 8.62 90 92 5.54 + 9.29 | + 8.50 0.79 195 192 +11.89 | +15.99 4.10 60 60 +14.09 | +17.84 3.70 60 60 + 5.33 | +10.15 4.82 35 29 101 7.11 | — 5.47 | — 2755 2.92 62 63 6.10 | — 1.67 | — 2.07 0.40 88 79 — 2.09 | — 0.61 1.48 60 59 — 1.02 | — 1.00 0.02 67 70 — 3.34] — 1.58 1.76 46 49 102 3 67 — 1.58 | +14.37 15.95 49 49 5.30 + 2.98 | +15.02 12.04 50 54 + 3.75 | +11.08 4.30 7G 78 — 2.23} + 1.45 3.68 62 66 — 0.32 | + 6.46 6.78 64 53 — 3.38 | +11.11 14.49 57 58 — 1.98 | + 4.52 6.50 41 60 — 2.23 | +10.19 12.42 88 88 + 3.55 | +13.35 9.80 95 95 — 1.28 | +12.23 13°51 70 60 103 0.00 — 2.46 | +13.48 15.94 GY ( 57 6.34 — 4,29 | +11.25 15.54 ia 60 + 0.16 | + 8.49 8.33 60 69 — 1.26 | +11.36 12.62 61 61 — 0.44 | + 2.15 2.59 84 — 3.75 | +11.04 14.79 60 60 + 3.89 | + 6.47 2.58 60 65 + 2.27 | + 6.99 4.72 63 68 — 0.18 | + 0.22 0.40 105 104 + 0.86 | + 4.22 3.36 62 64 105 5.938 + 8.46 | +14.64 6.18 68 60 8.81 + 9.19 | +13.7 4.59 50 60 + 6.89 | +14.10 7.21 80 80 + 6.55 | +15.20 8.65 60 60 + 6.74 | +14.83 8.09 93 93 + 6.21 | +12.73 6.52 60 60 +10.34 | +16.05 all 60 60 PHOTIC REACTIONS OF HONEY-BEE 423 XI. APPENDIX (TABLE 2)—Continued A B Cc D E F G H I NORMAL NORMAL DURATION! DURATION BEE AV.° | BEE AV.° OF REC- OF REC- NUMBER PER CM. PER CM. ORDS IN ORDS IN ONE EYE | ONE EYE BLACK AV.°| BLACK AV.°|-+ VALUES |— VALUES OF BEE | TO RIGHT TO LEFT EBE/ON: ERE OM. oe | SECONDS | SECONDS tae | Bae |uttcom [wean | O° | OS | "SE | ere 105 +10.03 | +17.05 7.02 60 60 + 9.04 | --11.01 1.97 100 100 + 7.46 | 4-12.68 5.22 163 163 106 6.43 — £2851 --4- 149 2.67 70 68 1.71 + 0.44 | + 3.22 2.78 81 79 + 2.190 Gee 3.93 41 41 — | 40 | tele 6.68 30 35 + 0.59 | — 5.49 4.90 77 77 + 0.53 | + 5.33 4.80 103 115 121 1.61 + 2.96 | +24.04 | 21.08 60 60 6.87 | -+11.57 | +-19.71 8.14 60 60 age. te je-t-top 1) |. 1198 92 88 tegtok tepalesi2 1 10.21 87 91 — 4.71 | +17.50 | 22.21 60 60 122 2.31 a E-8e) | 5, 96 7.08 69 73 2.93 + O.Od lean. 0e 0.12 118 111 + 3.25 | + 1.99 1.26 90 85 — 4.39 | + 6.50 | 10.89 56 59 123 17:86 | + 7.16 | --18.55 | 11.39 54 60 3.60 | + 1.79 | — 1.21 4.00 65 69 —10.76 | + 0.59 | 11.35 69 67 8.08 | + 2.29| 10.37 79 79 — 7.61 | — 2.14 5.47 60 60 — 3.46 | + 1.12 4.58 140 135 <= e009 8.94 60 58 — 3.53 | + 0.56 4.09 58 58 — 2.21 | + 1.9% 4.18 90 101 124 2.59 | + 3.88 | +11.18 7.30 73 68 3.00 | + 3.94] + 6.45 2.51 53 59 + 2.54 | + 8.38 5.84 79 71 * iO ae: ete 0.83 80 80 11.04 | --17- 61 6.57 90 90 10.39 | 4-17-44 7.02 60 60 + 5.81 | +12.54 6.73 58 58 + 6:50)| =F 6.98 2.40 74 60 + 7.67 | +-10.19 2.52 60 60 424 DWIGHT E. MINNICH XI. APPENDIX (TABLE 2)—Continued A B Cc D E F G H I NORMAL NORMAL DURATION |DURATION ONE EYE | ONE EYE BEE AV.° | BEE AV.° ° , 0 OF REC- | OF REC- NuMBER | PeRcM. | PER CM, | BUACK AV."| BLACK AV.°|-+ VALUES |— VALUES| Grpgin | ORDS IN orpes | rorieat| toxerr |, PER CM. EOR CM || OF HD) | Or?) | ‘snbonpsl| aaeaeee cero | Oe. | Geeaomr| emer | |)? || ae || 126 6.51 | — 1.39 | + 3.11 4.50 105 100 1.19 | + 4.51 | — 2.89 7.40 48 50 + 3.73 | + 6.14 DEA 78 (ih — 1.44 | +23.37 24.81 45 45 + 1.83 | + 4.41 2.58 107 107 + 4.57 | + 3.65 0.92 148 153 + 5.51 | + 4.32 1.19 72 72 — 2.98 | — 2.26 0.72 83 83 — 1.49 | + 3.73 5.22 25 28 133 2.82 + 2.39 | + 3.18 0.79 70 74 5.86 + 1.98 | + 3.26 1.28 69 68 + 1.54) + 4.31 Dati, 60 61 + 4.79 | -— 9.29 4.50 60 60 + 1.76 | + 4.92 3.16 82 76 + 3.30 | +11.54 8.24 38 40 + 3.35 | + 7.77 4.42 64 68 134 1.52 | + 8.29 | +19.62 il st; 69 69 1.22 | + 9.94 | +19.60 9.66 57 yi + 5.67 | +18.19 | 12.52 80 80 + 0.22 | +18.93 18.71 60 60 + 3.83 | +12.53 8.70 58 58 135 10.77 | — 3.25 | — 0.54 Peril 63 61 10.88 | + 0.74 | + 2.03 1!) 95 89 + 1.29 | + 3.35 2.06 109 109 — 9.49 | — 5.85 3.64 58 59 + 2.28 | — 1.37 4.65 7h! 71 — 1.05 | — 3.03 1.98 94 94 — 0.33 | — 7.67 7.34 79 76 + 0.44 | — 6.66 7.10 58 59 137 5.99 | +11.73 | +19.09 17.36 60 60 2.94 | +15.82 | +16.09 0.27 49 49 +10.59 | +16.45 5.86 70 70 +15.78 | +16.57 0.79 60 60 +14.31 | +15.40 1.09 78 78 +14.71 | +22.57 7.86 60 60 +14.69 | +17.02 2.33 70 70 PHOTIC REACTIONS OF HONEY-BEE 425 XI. APPENDIX (TABLE 2)—Concluded A B Cc D B F G H I NORMAL NORMAL =, DURATION | DURATION woumen, | DEE AY.” | BEEAY.” | piace av.°| BLACK Av.°|+ vans |~vaxune| OF BEC | OF BEC OF BEE | TO RIGHT | TOLEFT | pouNep 24 [rennED 957; ond | ond | SECONDS | SECONDS LIGHT. LIGHT. Bio TEN en2 BS dees LIGHT LIGHT. US 7 +13.38 | +15.64 2.26 60 60 + 8.43 | +16.71 8.28 60 60 +15.92 | +18.08 2016 ‘ 60 60 138 6.55 |) 0.18 |) Pas oy 4.49] 129 129 8187 | = 0:37.) = OrrsHorsa 62 58 — 1.12 | +138.52 14.64 65 60 — 9.00 | — 3.52 5.48 70 70 — 3.07 | + 1.60 4.67 115 115 — 1.97 | + 2.57 4.54 t 74. — 6.19 | — 1.08 a 69 69 Resumen por el autor, R. W. Hegner. Universidad John Hopkins. Los efectos de los factores ambientes sobre los caracteres heredi- tarios de Arcella dentata y A. polypora. El protozoario rizépodo Arcella dentata ha sido sometido por el autor a la influencia de varios factores ambientes. Hjemplares de tamafio y ntimero de espinas conocidos han sido tratados con soluciones de silicato sédico y alcohol etilico; tambien fueron sometidos a varias temperaturas y nutridos insuficientemente. En los descendientes de animales asf tratados aparecen cambios en el didmetro de la concha y en el nimero, tamafio y forma de las espinas, pero recobran la condicién normal cuando se crian de nuevo bajo condiciones normales. Translation by José F. Nonidez Carnegie Institution of Washington AUTHOR'S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, SEPTEMBER 29 THE EFFECTS OF ENVIRONMENTAL FACTORS UPON THE HERITABLE CHARACTERISTICS OF ARCELLA DENTATA AND A. POLYPORA! ROBERT W. HEGNER Department of Protozoology and Medical Zoology, School of Hygiene and Public Health, The Johns Hopkins University SEVEN FIGURES While carrying on a series of experiments for the purpose of testing the efficacy of selection as a means of isolating heritably diverse lines within a clone of Arcella dentata, and later while studying the nucleocytoplasmic relations in this species and in A. polypora, several experiments were performed with the pur- pose of determining the effects of environmental factors upon the heritable characteristics of these organisms. If it is possible to modify organisms by means of external fac- tors in such a way that the diversities produced will persist after the disturbing factors have been eliminated, we may account for the numerous heritable diversities that have been described among the lower organisms by the presence of such factors in their environment. UNDERFEEDING During the selection work on Arcella dentata (Hegner, 719) the organisms were supplied with an abundance of organic matter shaken from the leaves and stems of aquatic plants. The pond water in which this food material was suspended was then diluted with distilled water. At first the number of spines was used for purposes of selection, but later the diameter of the shell was also employed, since the organisms were found to be remarkably constant in size, and diameter of shell and spine 1 The studies presented in this paper are incomplete, but they are published at this time, since the writer will probably be unable to carry on further experi- mental work with Arcella in the near future. 427 428 ROBERT W. HEGNER number were shown to be closely correlated. It was also discoy- ered that a definite relation exists in these animals between nuclear number and size and between chromatin mass and cyto- plasmic mass. These facts led to the experiments on under- feeding described below. The data obtained furnish infor- mation regarding fission rate and variations in diameter of the shell. a. The effects of underfeeding upon the rate of fission Specimens of Arcella dentata were taken from cultures that were being used for selection experiments, and placed in a medium consisting of one-half distilled water and one-half filtered pond water. Instead of a gradual decrease in nuclear and cytoplasmic material and the cessation of reproduction as was expected, sey- eral specimens proceeded to divide, and within a month three rather large families had been produced by them. Evidently, even after being filtered, enough food remained in the pond water to enable the arcellas to grow and reproduce. However, the amount of food present was much less than was ordinarily given to the specimens in laboratory cultures, and several in- teresting results appeared which were apparently due to under- feeding. There was a marked retardation of the division rate during the period when the specimens were underfed. Fission occurred at intervals of from two to ten days, with the following means: Family 2. Average interval between fissions, 5.25 days. - Family 3. Average interval between fissions, 3.47 days. Family 4. Average interval between fissions, 4.15 days. In contrast to this, the division rate of the parental lines under normal cultural conditions was approximately 2.50 days. Previous work on pure lines in Arcella dentata indicated that different lines differ in division rate, but these were all being reared under similar conditions and supplied with an abundance of food. The differences in division rate between these pure lines were, therefore, not due to differences in nutrition. The results just described, however, show that it is necessary to keep ENVIRONMENT ON ARCELLA 429 food conditions constant when undertaking experiments involvy- ing division rate. They also prove that Arcella dentata is able to grow and reproduce under adverse conditions with respect to the food supply, but at a much less rapid rate. b. The effects of underfeeding upon the diameter of the shell The offspring of parents that were underfed showed the effects of this treatment immediately, being smaller than their parents in every case. When these offspring were underfed, they like- wise gave rise to smaller offspring than the normal for the line, but not, on the average, smaller than themselves. When, on the other hand, these small offspring of underfed parents were TABLE 1 Arcella. dentata. Table showing the distribution in diameters (in units of 4.3 p) of specimens in families, 2, 3, and 4 during periods when they were being reared in normal medium and when they were underfed DIAMETER 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | 32 | 33 | 34 | 35 Family 2. Normal....... 2|6)2)1 Family 2. Underfed...... 1 Aes Sb]. 2, | Family 3. Normal....... Dish Des| plete Family 3. Underfed...... B judy kk LA% eDiletaied Family 4. Normal....... 2/1 )-2 Family 4. Underfed...... 1/4) 15)8|4/[2 returned to normal cultural conditions, their first offspring showed the effects of the abundance of food, becoming close to the normal. Also, when full-sized specimens that were produced under normal conditions and which had given rise to small off- spring when subjected to underfeeding, were again supplied with an abundance of food, the size of their offspring immediately attained that normal for the line. The six cases on following page present data characteristic of the series. Table 1 gives the distribution of the diameters of the speci- mens reared in families 2, 3, and 4, both when the parents were in normal cultural conditions and when underfed. The following means bring out the significance of these data: ¥ Table of means NUMBER OF|MEAN SPINE MEAN SPECIMENS NUMBER DIAMETER Family 2. Offspring produced while parent in normal COngIPiION. «chong Meabies or « ai pic eee 11 11.50 27.18 Family 2. Offspring produced while parent un- derfed iit es. 1 os +c ae. vs. 2. 12 10.82 24.25 Family 3. Offspring produced while parent in normal’condition.1)), Paseo.) Ye ee) 6 14.17 33.50 Family 3. Offspring produced while parent un- deried ey on. ce seh Meee: <> 9 fc ae ae 36 13.54 30.94 Family 4. Offspring produced while parent in normal condinaie. cp itn «|. s ure Vee nuee eee 15 13.20 33.00 Family 4. Offspring produced while parent un- erie... sake cuts ees re hte e ee eae oe eee 24 12.88 30.47 Siz cases one DIAMETER Pamily.2:. “Originalprogenitor’. aaetenccnee-cue edocs eee 10 27 First offspring while parent underfed......................- 9 23 First offspring when parent returned to normal............. 11 27 First offspring when parent again underfed................. 10 26 Second offspring when parent still underfed................. 9 24 Family 2. First offspring original progenitor............... 9 23 First offspring while parent underfed.....................-- 11 23 Second offspring while parent underfed..................... 11 25 First offspring when parent returned to normal............. LE 26 Family 2. First offspring of first offspring.................. 11 23 First offspring while parent underfed....................0-. 10 21 First offspring when parent returned to normal............. 10 26 Family's. Original’ progentpor:...:.. 2/2 See ee 15 35 First offspring while parent underfed.....................-- 14 31 Second offspring while parent underfed....................- 14 30 Family 3. First offspring of original progenitor............ 14 31 Fourth offspring while parent underfed..................... 12 28 Sixth offspring when parent returned to normal............. 14 33 Family 4. Original progenitor........2...6. /..2.2ee see 14 32 Fourth offspring while parent underfed....................- 11 29 First offspring when preceding specimen placed in normal... 14 33 First offspring when preceding specimen underfed........... 14 29 First offspring when preceding specimen underfed........... 12 30 430 ENVIRONMENT ON ARCELLA 431 The mean diameter of the line from which the progenitor of family 2 was taken was 26.40 units and the mean spine number 10.90. The mean diameter of the line from which the progenitors of familes 3 and 4 were taken was 34.00 units and the mean spine number 14.59. In connection with these experiments in underfeeding it may be noted that a large majority of wild specimens, when collected from the vegetation in ponds, possess very little cytoplasm. This is probably due to the struggle these minute organisms must undergo in their natural habitat. On the other hand, it is not unusual after being brought into the laboratory for the offspring of such wild specimens to average much smaller than their orig- inal progenitors, although they are supplied with an abundance of food. Thus,in Arcella polypora the progenitors of twenty-six families ranged in diameter from 23 to 35 units of 4.3 « each, with a mean diameter of 30.42 units, whereas their first offspring ranged in diameter from 21 to 32 units with a mean diameter of 27.50 units. No definite reason can be given for this decrease in size under laboratory conditions, but perhaps the abundant food supply is responsible, resulting in the initiation of fission before the cytoplasm has increased to the amount normally present when the animals are in their natural habitat. The results of these experiments prove that size and spine number in Arcella dentata are affected by the food supply. Selection experiments involving these characteristics in this organism and probably in other similar organisms, as well, must therefore be carried out so as to provide a constant food supply. This factor, however, was carefully controlled in the selection experiments that have been reported on Difflugia (Jennings, ’16), Centropyxis (Root, 18), and Arcella (Hegner, ‘19 a). THE ADDITION OF SODIUM SILICATE The fact that Whitney (’16) caused the transformation of the rotifer Brachionus pala into the variation Brachionus amphiceros by the simple addition of sodium silicate to the medium in which they were reared, led to the use of this substance in experiments on Arcella dentata. It seemed probable that the presence of an 432 ROBERT W. HEGNER excess of sodium silicate might facilitate shell production and bring about the formation of variations, such as longer spines. The method employed was to make up daily or every other day culture media as usual and then add one drop of sodium silicate to 100 ce. of the medium. Solutions of greater strength were tried, but the organisms did not thrive in them. The experiments were begun with thirty specimens taken from the lines that were being used at the time for selection experiments. From these, five families were reared as follows: Family 14 with 135 specimens Family 15 with 20 specimens Family 20 with 11 specimens Family 25 with 9 specimens Family 26 with 19 specimens No difficulty was experienced in obtaining large families, the number of specimens recorded being limited only by the amount of time available to care for them. a. Fission rate The rate of fission of specimens grown in the sodium silicate medium decreased immediately from an average of one division in 2.50 days to one division in about four days. Evidently the presence of sodium silicate affected adversely the food material upon which the Arcellas were feeding or else hindered the feeding process or the rate of digestion and assimilation. b. Size and spine number Instead of an increase in spine number and length and in the size of the shell as was expected, the immediate result of the changed me¢ium was a decrease in all these characters. The cise bal cen in figures 1, 2, and 3 illustrate the differences in diameter and spine number very clearly. Many of the specimens were badly crinkled, which, no doubt, accounts in part for their smaller diameter; in others the binucleate condition was lost and uninucleates appeared. Wherever this occurred a 1 in paren- ENVIRONMENT ON ARCELLA 433 theses is added in the figures (figs. 1, 2, and 3). In another place (Hegner, ’20) the writer has shown that this change from the binucleate to the uninucleate condition is accompanied by a decrease in size, as indicated also by these experiments. The 11-28 (n) (1) 13-33 (n) —————_ 14-35 (n) ee tae) 45-20(n) aa (s)———12-30(s) 17-32 (g) ————— 14-31 (s) —_———13 28 (s) 15-32 (3) ————— 16-31 (s) 2-16(s)(o) 15-33 (n) —————_ 1432 (s ) ——————- 2#255(8) (c) 128M, gee ——_——15-34(n) 2 eee 16- 2 ae UP rae 34(s) 14-33 (gs) ———_ 2=30(s) ee” 8-24 E ae 2=24(5) (0) 2=25(s) (co), 2=26(s) (0) Noe | ae 14-33 (s) —_______?#31(s) 2=22(s)(c) \ cas tants 12+30(s)—______14-33 (n) Fig. 1 Arcella dentata. Part of the pedigree of family 14. The numbers indicate the number of spines and the diameter in units of 4.34. For example, the original progenitor (16-34) possessed sixteen spines and was 34 units in diame- ter. The letters and numbers in parentheses should be interpreted as follows: (s) = specimen produced while parent was in sodium silicate solution; (n) = specimen produced while parent was in normal medium; (c) = specimen with crinkled shell; (1) = specimen with only one nucleus; (?) = indeterminate number of spines. outlines in figures 4 and 5 show the decrease in the length of the spines. In many cases the spines did not extend beyond the edge of the shell, being represented only by ridges on the dorsal surface of the shell. The experiments extended over the period from March 18 to May 15, 1918. The largest family, no. 14 434 ROBERT W. HEGNER (fig. 1), was begun on March 21, and its original progenitor lived throughout the rest of the period, giving rise in that time to ten offspring. Specimens from branches of this family that had been constantly subjected to the sodium silicate solution were at intervals transferred to the normal culture medium, but in every case the size, spine number, and spine length character- istic of the parent line were immediately regained. Family 14 was continued until it contained 135 members and included specimens of the eighteenth generation, but no heritable variations were noted that could be attributed to the changed medium. Figures 4 and 5 illustrate the changed condition of the offspring 12-27(s8) Wee ciate ?-20(s8)(o) 13-27(n) eo, aes 10-26 (n) 9-26(n) 12-27 12-27(s) 12-27 (s) — 11-28 (a) Fig. 2 Arcella dentata. Part of the pedigree of family 15. For description see figure 1. when the parents of this family and of family 15 were subjected first to the sodium silicate solution and later were replaced in the normal medium. c. Color of the shell Another character that was modified by the presence of sodium silicate was the color of the shell. Arcellas that are reared in a normal medium have a shell that is at first almost transparent, but graduaily changes to a pale yellow, and in time becomes a very deep brown. ‘The offspring reared from parents that were kept in the sodium silicate solution were pale yellowish green, in color, as long as they remained in this medium, but became of the ENVIRONMENT ON ARCELLA 435 normal brown color as soon as they were transferred to normal cultures. The color mechanism of Arcella seems to be very sensitive to environmental factors. During the work on selection it became evident that pure lines probably exist with respect to the length of time it requires for the young to reach their defin- itive color and also with regard to the intensity of the color attained. Changes in the character of the food were likewise found to affect the color. For example, when food was obtained from vegetation taken from a cement tank in the botanical garden on the Johns Hopkins University campus, the offspring became dark brown almost immediately after fission; specimens reared aha 9-25(n) (1) {9-23 (a) (1) 9=23(n) (1) P20(s)(0)(1)— 8-23 E 14-29(8) (ce) za -27{a) 2-24(8) (oc) (1)—— 8-25(n) 10-29(n) 13-32 (n) 9-27 E Z|\\ 11-28(s) 11 E 11-30(n) ~~ 15-35 (n) Fig. 3 Arcella dentata. Complete pedigree of family 26. For description see figure 1. 10-28 (n) in hay infusions were more yellowish in color than normal, and those fed on material collected from a spring-water fish pond at Cold Spring Harbor, Long Island, were characteristically pale. Arcella thus resembles the many other organisms that are modified by changes in the environment; they remain so as long as they are in this environment, but return to their former con- dition when transferred back to the original medium. THE ADDITION OF ALCOHOL One of the substances to which Protozoa have been found to be resistant is alcohol. Experiments were begun to determine the effects of alcohol on Arcella dentata, but were terminated 436 ROBERT W. HEGNER because of lack of time to keep them going. Sufficient data were obtained, however, to prove that these organisms are able to live and reproduce in a medium containing from 0.25 to 1 per cent of alcohol. ‘Three offspring were obtained from one of the E D Fig. 4 Arcella dentata. Family 14. Camera-lucida sketches (X 207) showing: A = the original progenitor; with sixteen spines and diameter of 34 units. B = the first offspring after A was placed in a solution of one drop of sodium silicate to 50 cc. of normal medium. C = the third offspring of A, produced while A was in a solution of one drop of sodium silicate to 100 ce. of normal medium. D = fourth offspring of A, produced immediately after A was transferred from the sodium silicate solution back to normal medium. E = seventh offspring of A, produced immediately after A was transferred from normal medium to sodium silicate solution. F = a specimen of the seventh generation after continuous subjection to sodium silicate. specimens that was kept in a 0.50 per cent solution of alcohol; two of these and the parent were still alive thirty-five days after the experiment was begun. The rate of fission was very slow, probably because of the effects of the alcohol upon the food. ENVIRONMENT ON ARCELLA 437 One of the progeny was irregular in shape and another was without definite spines. It is evident from the results that alcohol is injurious to these organisms, although they are able to with- stand the presence of a considerable amount for a long period. TEMPERATURE A number of specimens of Arcella dentata were collected on December 27, 1917, from pond weeds under a layer of ice eight inches thick. Many of these had no well-developed spines, but COnCOY A B E Fig. 5 Arcella dentata. Family 15. Camera-lucida sketches (X 207) showing: O, D A = the original progenitor with twelve spines and diameter of 27 units. B = the first offspring after A was placed in a solution of one drop of sodium silicate to 100 cc. of normal medium. C = the first offspring of B, after B had produced two offsprings while in sodium silicate solution and was then transferred to normal medium. D = the first offspring of C while C was in normal medium. their progeny, when reared under laboratory conditions, exhibited a complete set of fully formed spines. This suggested that per- haps the low temperature might have been responsible for the lack of spines in the ‘wild’ parents. Several experiments were begun to test this hypothesis, and although they were not ex- tensive they indicated that length of spine and temperature may be correlated since, as shown in figure 6, the spines of offspring reared at a temperature of about 10°C. are smaller than those of their parents which were reared at room temperature. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 29, NO. 3 438 ROBERT W. HEGNER SPECIMENS WITH BENT, OVAL SHELLS Among the specimens of Arcella polypora that were collected at Cold Spring Harbor during the summer of 1918 were many with a shell that was oval in outline and contained an oval mouth opening and at the same time was bent slightly as indicated in figure 7. Most of these were obtained from duck weed taken from a small pond that was gradually drying up, but others were found in a small lake. Leidy (79, pl. XXVIII, figs. 36 and 37) figures shells similar to these. A number of specimens were isolated and their progeny examined. The offspring of the first generation were either circular in outline or nearly so and almost ae B 2 A Fig. 6 Arcella dentata. Camera-lucida sketches (X 450) of three of the spines belonging, respectively, to normal specimens (A, 1 and B,1) and to the: first offspring produced by them when subjected to low temperature (A, 2 and B, 2). flat, and those of later generations were entirely normal when reared under laboratory conditions. A few measurements are given below. DIAMETER OF FIRST DIAMETER OF PARENT OFFSPRING SHAPE OF FIRST OFFSPRING 28 x 25 25 Flat, almost circular 29 x 25 27 Almost flat, almost cireular 28 x 25 29 Almost flat, almost cireular A large number of specimens were placed in Syracuse watch- glasses and examined every day for a week. Many offspring were produced by them, but in every case they were quite normal or nearly so. Since the descendants of these bent, oval speci- ENVIRONMENT ON ARCELLA 439 mens were normal and since no specimens of this sort have ap- peared among the thousands reared in the laboratory, it seems safe to conclude that some environmental condition is respon- sible for this peculiarity and that as soon as the controlling factor is removed, the normal characteristics are regained. SUMMARY AND CONCLUSIONS a. When specimens of Arcella dentata are underfed, the inter- val between successive divisions increases from an average period of 2.50 days to a period of about 4 days; the shell decreases in diameter on the average 2.68 units of 4.3 u each, and the spine IO) Fig. 7 Arcella polypora. Camera-lucida sketches (X 207). A = dorsal view of ‘wild’ specimen with oval shell and mouth. C = side view of same showing bent condition of the shell. B = dorsal view of first offspring of A under labora- tory conditions. D = side view of B. number also decreases slightly. The offspring of underfed par- ents produce progeny normal in size and spine number when given an abundance of food, and parents that have been underfed likewise give rise to normal offspring under similar conditions. Wild specimens are often poor in cytoplasmic content, but their offspring are frequently smaller than the parents when reared under laboratory conditions. b. Arcella dentata will grow and reproduce in a medium con- taining one drop of sodium silicate to 100 ce. of water. The fis- sion rate decreases, the average interval between fissions increas- ing from 2.50 to 4 days. The size of specimens produced while the parents are in the sodium silicate solution is less than that of progeny formed under normal conditions. The most con- 440 ROBERT W. HEGNER spicuous change brought about by the presence of sodium sili- cate is the almost complete loss of spines. The color of the shell, which becomes brown in a normal medium remains a pale greenish yellow in a sodium silicate solution. Specimens reared in sodium silicate and then returned to a normal medium regain the fission rate, size, spine length, and color or characteristic of the race. c. Specimens of Arcella dentata are able to grow and repro- duce in a medium containing from 0.25 to 1 per cent of alcohol. Alcohol, however, is shown to be injurious, as indicated by the retarded fission rate and irregularities in the shells of the offspring. d. Several experiments seem to indicate that temperature influences the length of the spines of Arcella dentata, and that the lower the temperature the smaller the spines become. e. Wild specimens of Arcella polypora that possessed a bent, oval shell with an oval mouth opening gave rise under labora- tory conditions to offspring with a flat circular shell and a circular mouth opening. The bent, oval condition is probably due to an unknown environmental factor. f. The environmental factors to which specimens of Arcella have been subjected cause distinct variations from the normal racial conditions, but these modifications persist only so long as the modifying factors are operative. No heritable diversi- ties were observed that were due to the changed conditions. The experiments bring out no data that affect the results obtained by Jennings, Root, and the writer in isolating heritably diverse lines within families of fresh-water rhizopods during vegetative reproduction. ENVIRONMENT ON ARCELLA 441 LITERATURE CITED Heener, R. W. 1919 Heredity, variation and the appearance of diversities during the vegetative reproduction of Arcella dentata. Genetics. Vol. 4, pp. 95-150. 1920 The relations between nuclear number, chromatin mass, cytoplasmic mass, and shell characteristics in four species of the genus Arcella. Jour. Exp. Zoél. (In press.) Jennines, H. S. 1916 Heredity, variation and the results of selection in the uniparental reproduction of Difflugia corona. Genetics, vol. 1, pp. 407-534. Lerpy, J. 1879 Fresh-water rhizopods of North America. Report U. 8. Geol. Survey of the Territories, 12, 324 pages. Root, F. M. 1918 Inheritance in the asexual reproduction of Centropyxis. aculeata. Genetics. Vol. 3, pp. 174-206. Wuitnry, D.D. 1916 The transformation of Brachionus pala into Brachionus amphiceros by sodium silicate. Biol. Bull., vol. 31, pp. 113-120. Resumen por la autora, Mary Elizabeth Collett. Universidad de Pennsylvania. La toxicidad de los Acidos sobre los infusorios ciliados. La autora ha determinado la toxicidad del Acido clorhfdrico y la de unos 15 acidos orgénicos en varias concentraciones y temperaturas sobre Paramoecium caudatum, Stylonichia pustu- lata, Euplotes patella y Vorticella nebulifera. En concentra- ciones de 0.001 y 0.00008 N el 6rden téxico es groseramente paralelo al érden de disociacién, indicando esto que el ién H es un factor importante en la toxicidad. El érden t6xico varia algo con la temperatura, concentracién y con la especie. Si se disminuye la ionizacién (como sucede con las mezclas de acido clorhidrico con un Acido débil) algunos de los dcidos son menos efectivos, indicando esto que el anidén es t6xico, asi como elién H. Los coeficientes de temperatura para intervalos de diez grados entre los 10° y 33°C indican que tanto las reacciones quimicas como las fisicas tienen influencia en la accién t6xica. Los acidos acético y butfirico son irregulares en lo referente a su toxicidad, puesto que esta funcién en Euplotes, aunque no en Paramoecium, aumenta con temperaturas inferiores y también superiores a los 20°. Translation by José F. Nonidez Carnegie Institution of Washington AUTHOR’S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, SEPTEMBER 29 THE TOXICITY OF ACIDS TO CILIATE INFUSORIA M. E. COLLETT University of Pennsylvania SIX GRAPHS INTRODUCTION Many workers have studied the toxic effect of acids and have tried to arrive at an explanation of their mode of action. Very few of the experiments, however, are quite satisfactory. One difficulty lies in the choice of material: for example, in studies of the effect upon dogs of the feeding (Walter, ’77) or intravenous injections of acids (Szili, 09) or the effect upon fish or tadpoles of adding acids to the external medium (Loeb and Wasteneys, 712; Roaf-Whitley, 04; Unger, 716; Winogradoff, ’11), the material is too complex for a ready analysis of results. Some of the best work of the sort is that of Winogradoff (’11), who watched the process of injury to heart, corpuscles, muscle, etc., in small transparent tadpoles. More satisfactory material has been found in tissue or isolated cells, such as rootlets, muscle, ciliated cells, and erythrocytes (Kahlenberg-True, ’96; Loeb, 98; Harvey, ’14; Landsteiner and Prasek, ’13) where conditions are much less complex. Aside from obvious crudities, such as not guarding against evaporation or not indicating time accu- rately, the experiments have most often been unsatisfactory be- cause of lack of data for more than one concentration of the acids used or because of faulty temperature control. The usual method has been to determine upon some one tissue the toxic order of a series of acids, all at the same concentration (or at the concentration necessary to kill in a fixed time) and then to try to explain their relative toxicity by correlating the results of this one experiment with the physical properties of the acids. 443 444 M. E. COLLETT This is not a satisfactory procedure, as was pointed out by Cro- gier in his work upon permeability. Lack of uniformity in concentration and temperature makes a comparison of different experiments almost impossible, even when the reactions involved are the same, and there are relatively few experiments in which several kinds of material or several physiological processes have been studied all together. The aim of the present experiments is to determine the relative toxicity of a series of acids, at sev- eral concentrations each and at various temperatures, to differ- ent sorts of nearly related material, and from these more com- plete experiments to draw what conclusions are possible as to the nature of the toxic action. J am greatly indebted to Dr. M. H. Jacobs, who suggested the problem, for helpful advice and criticism. METHOD The organisms observed in the following experiments were Paramoecium caudatum, Stylonichia pustulata, Euplotes pa- tella, and Vorticella nebulifera. They were grown in small jars of boiled hay infusion made with pond-water and seeded with decaying leaves, etc. Though many forms appeared in the cul- tures, only these four were constantly present in sufficient num- bers to be useful for a long series of experiments. The acids used were hydrochloric, formic, acetic, propionic, butyric, valeric, lactic, oxalic, malonic, tartaric, citric, benzoic salicylic, and phthalic. All were made up at 0.1 N and titrated with litmus against 0.1 N NaOH. These stock solutions, as well as the weaker solutions made from them for use in the ex- periments, were kept in paraffined bottles, for they deteriorate rapidly in unprotected reagent bottles, even those prepared by thorough steaming. The organisms were always washed overnight before use in order to eliminate fluctuations due to the changing composition of the culture medium. They were collected from the culture in as concentrated masses as possible and pipetted into about twenty times their volume of fresh pond-water. After twelve hours’ washing they were concentrated again by means of a TOXICITY OF ACIDS TO CILIATE INFUSORIA 445 hand centrifuge. This treatment does not injure them. If they are washed less than five hours, the results are apt to be irregular, but five hours seems to be almost as satisfactory as a longer period. Pond-water was used because ordinary distilled water seems to be somewhat injurious. After being washed and reconcentrated, the organisms were transferred to a watch-glass, in which they could be kept in apparently psn condition for several hours. In order to test their resistance to the acids, the following procedure was adopted. A test-tube was filled with about 15 ee. of the solution to be tested, and less than 0.1 ce. of the concen- trated culture was introduced with a capillary pipette. After being inverted once or twice for mixing, the test-tube was corked and placed in a water-bath, where the temperature could be held at the desired level throughout the experiment. Samples were poured at intervals into watch-glasses and observed under the low power of the compound microscope. It was found desirable to use watch-glasses with curved rather than vertical sides in order to prevent collection of the organisms at the margin where they are difficult to see. It is not enough to note the time at which the organisms cease to move about, for the cilia usually beat for some time afterward. Since there is often considerable variation in individual resistance, the time at which the cilia of just over half cease to beat was adopted as a better measure of average resistance than the time at which all are dead. Every solution was tried five or even ten times over, usually on different days, in order to check deviations due to biological differences, which, however, were relatively slight. In the fol- lowing tables the averages of all the determinations are given. DISCUSSION OF RESULTS A. Concentration In these experiments each acid of the series was used at four of the following concentrations, viz., 0.001, 0.0005, 0.0002, 0.0001, 0.00008 N. Full data on the effects of these solutions are given in table 1, but the important points are perhaps 446 M. E. COLLETT clearer in, graph 1. It will be observed that for Paramoecium at 20° salicylic and HCl are at all concentrations the most toxic of the acids studied and acetic and butyric the least so, with the other acids scattered in between. But although in a general way the toxic order is similar at the different concentrations, there are several important exceptions. For example, the toxic order at 0.0002 N is formic > tartaric > oxalic = malonic, while at 0.0001 N the order is tartaric > malonic > formic > oxalic. This change in order is indicated by intersection of the curves. There is a similar though less important change at the next concentration, where the curve for HCl crosses that of salicylic and the tartaric curve crosses those of malonic and formic. With Euplotes change in concentration produces less variation in order, but instead a considerable variation in the degree of difference existing between the acids. Thus at 0.0005 N HCl formic and benzoic acids kill in 13, 24, and 3 minutes, re- spectively, while at 0.0001 N the relative times are 9, 40, and 35 minutes. Obviously, if conclusions were based solely upon the results for a single concentration, 0.0005 N for instance, they would not be valid, since the results at other concentrations are in so many respects different. This conclusion agrees with the results obtained by Crozier (’16) in studying the rate of penetration of acids into the mantle cells of Stichopus, which contain a natural indicator. He found that the relative order obtaining at a concentration of 0.01 N was in many respects unlike the order at 0.001 N, and that the order of penetration found by Harvey (714) and by Haas (’16) at 0.01 N with different material closely resembled his results at 0.01 N, but not his results at other concentrations. If other KEY TO GRAPH SYMBOLS 1. Hydrochloric ———— 8. Citric ——|—— 2. Formic —|-|——|-|— 9: Maloniey.-asseee 3. Acetic ————<— 10. Tartaric ——o0o0—— 4, Propionie —-|——+|— 11. .Benzbie =e 5. Butyric ~—+~—--—. 12. Phthalic —— : ——. 6. Lactic —— - —— 13. Salicylic — — — — — 7 . Oxalie ——o—— TOXICITY OF ACIDS TO CILIATE INFUSORIA 447 Min : PA j. A 70 : o ! 60 is 10 | ; : vat a re i] 50 hears ot es y. We ee ? - o ¢ A _ 4 ge 30 . io ee vie yee Vg are : 6 0° A A A J 10 SE / 5 ae vf ee oes eo 7 o ee i Se N x to73 8° 20 10 Min. B / I, 50 3 7 1° oF ° fe ® 40 / r ° ha : ed ° oe ” i! ; 20 fo / i 0,5 10 U = —— bi = 5 eee ae = mm ai eee) ee ee 3 Nx 163°Go 20 10 7 Graph 1, A, B_ Toxicity of equinormal acids at 20°C. (Euplotes above, Paramoecium below.) Should be N~* and not N°. 448 Effect of temperature on toxicity of equinormal acids. M. E. COLLETT TABLE 1 £ re) 2 | Z 2 © Nien | 5 | 2g =x ° a a fe < Paramoecium 0.001 10 0.0005 5 8 if 7 4 | 17 12°. ee OFO0OZF 910) 16 ae vie2b 15) oO 0.0001 | 22 40 |120 | 38 0.00008 (| 0.001 6 0.0005 4 +1 5 i 4 | 12 152. 2 Roe 0.0002 6 | 12% | 14 | 25 OW ikol 0.0001 15 35 {100 | 40 0.00008 {| 0.001 4 0.0005 13 | 3 3 24 | 8 PAR cen art iets. te Be 0.0002 44 | 9 9 | 10 GO laze 0.0001 Oi someseoO) 235" |) 40 0.00008 | 18 (a) Meta, |) 4910, 0.0002 6 9 5 Danae e ee ee ck 0.0001 6 13: | 28ra 20 0.00008 | 10 22 0.0002 3 33 | Il ; i 0.0001 5 @ N20) 14 0.00008 | 8 10 20 Stylonichia 0.001 2 0.0005 13 i | 3% 23 | 4 DORE. 0 Ac pe eee 0.0002 5S Pat See lilt ee 0.0001 93 50 | 90 | 40 0.00008 | 20 Euplotes 0.001 Z. 12° | 0.0005 | 12 10 15 18...) 12) sie25 cage or 0.0002 | 22 | 40 | 60 |100 | 25 0.0001 | 48 100+ BuTyYRic! a1 17 60 Time in minutes o 2 5 ro) cal S iS are T | Para- moecium caudatum > Euplotes charon > Stylonichia pustu- lata > Urostyla grandis > Euplotes patella > Vorticella. Additional evidence of specific differences is presented in the sections following. Although several experiments upon the cilia of other organ- isms have been reported previously, it is almost impossible to compare them, for they were made by different observers under widely varying conditions. Thus in Barratt’s experiments (04, 452 M. E. COLLETT 705) upon Paramoecium aurelia, the acids were not used consist- ently at the same concentrations, the temperature varied from 16° to 22°, and the method of washing injured the organisms. Weinland (’94) studied the oral epithelium of the frog at 18°, but made up his solutions at 0.001 M in physiological NaCl. Harvey (14), in studying the cilia of the giant clam, used acids made up to 0.01 N in van’t Hoff’s solution and worked at a temperature of about 27°. We do not know in these eases whether the salts antagonize the acids (as was found by Loeb, 12, and by Osterhout, ’14) or whether they reinforce its action (as found by Paul et al., 710), and consequently are not sure how far these experiments are comparable with other experi- ments in which the acids were made up in distilled water. Neither are Koltzoff’s studies (714) upon the cilia of Carchesium available for comparison, for, unlike the previously mentioned experiments, they deal with concentrations which do not kill the cilia, but only increase their viscidity. There are, however, reports upon other material capable of showing specific differences clearly. The experiments of Land- steiner (713) on acid agglutination of blood corpuscles and of Walbum and of Cunningham (’16) on haemolysis give clear evidence of specific differences in resistance. Ritter (712) and Clark (’96) found striking differences in resistance among vari- ous mold spores, as did Heald (’96) with seedlings. Taylor (17) found that the concentration of any acid necessary to clear a wound of bacteria varied with the organisms concerned and that the toxic order likewise varied. This is precisely the condi- tion observed in the present experiments. Similar results have been obtained by Kopazcewski (714) with enzymes. His obser- vation that the optimal Ps was not the same for his usual prepa- ration of maltose and for a well-dialyzed preparation indicates the importance of salts in determining acid resistance and may perhaps lead to an explanation of certain specific differences. TOXICITY OF ACIDS TO CILIATE INFUSORIA 453 C. Temperature In order to find out something of the nature of the toxic action, acids were tried at several temperatures between 10° and 30° as is indicated in table 1. The conditions are more readily seen in the curves shown below. In almost every case an increase above room temperature produces an increase in toxicity both to Paramoecium and to Euplotes and a decrease below room temperature generally produces a decrease in toxicity. If the length of life at, let us say, 10° is divided by the length of life at 20°, the figure so obtained (the temperature coefficient for 10°) can be used as a measure of the degree of influence exerted by these temperatures upon the toxicity of the acid. Table 2 gives the coefficients found for Paramoecium and Euplotes. It will be observed that the coefficient for any one acid is not the same at all temperatures nor at all dilutions. For example, benzoic for the range 20° to 30° has coefficients of 2 at 0.0002 N and 3.88 at 0.0001 N, and for the range 10° to 20°, coefficients of 3 and 2.1, respectively. Often the more dilute solutions have higher coefficients than the more concentrated. It isalsoremark- able that the temperature coefficient for any given range is not the same for different acids even at equinormal concentra- tions. Thus at 0.0002 N the coefficients for the range 10° to 20° are salicylic 1.5, hydrochloric 2.2, benzoic 3, malonic 2.5, etc. All this would indicate that the mode of action of dif- ferent acids is by no means the same and that the action of any one is markedly influenced by concentration and by temperature. Neither do the various acids affect Paramoecium and Euplotes in precisely the same way; if the coefficients for the two organ- isms be compared at corresponding temperatures and concen- trations of the same acids, one finds that they are almost never the same. It is impossible at present to interpret these results fully, but some few conclusions suggest themselves. As is well known, a coefficient of from 2 to 3 for every increase of 10° generally indicates a chemical reaction, while a coefficient of less than 2 or over 4 is frequently associated with physical processes. Most THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 29, No. 3 454 M. E. COLLETT TABLE 2 Temperature coefficients PARAMOECIUM EUPLOTES N/ 12-20% | 15-25° | 20-30° | 12-20° | 15-25° | 20-30° 0.0005 | 4.3 1.7 0.0002 | 2.7 1.5 | 1.66 HCl. - 0 up;0cbe. one | PE os ie Va 0.00008 a2 ik J | 0.0005 | 3.3 1.1 Onaga ha ea | 0.0002 | 1.9 10 0.0005 | 2.8 1.2 Eee 0.0002 | 2.2 2.57 | 1.0 eee) eee aie 0.00008 6.0 0.0005 | 2.4 1.38 Malonies: (aans. eee 0.0002 | 3.2 10:4 be 0.0001 | 4 1.7 0.0005 | 1.8 2.4 a 0.0002 | 3.2 By ae eB Frere rete, | PAO 2 a ae 0.00008 2.5 | 0.001 | 2.5 2.0 | 0.41 Acetic tee. sia 0.0005 | 2.1 0.55 0.0002 | 2.2 0.001 | 1.88 0.36 Butyrie: (228. Pc) eee OOS: SS 0.60 0.0002 | 2 0.0005 | 2.8 1.6 Benzoic 0.0002 | 3.9 2.0 | 1.0 DIG ae stg cee PO IE 9 88 0.00008 2.88 [ 0.0002 | 1.6 13 Salicyligwe. oils es Be es :} 0.0001 | 2.1 2.5/4.0 | 0.00008 3.84 ‘ Coefficients for 10° calculated from observations at 12° and 20°, except in the case of acetic and butyric acids, which were used at 10° and 20°. TOXICITY OF ACIDS TO CILIATE INFUSORIA 455 of the coefficients for a change of 10° obtained with Paramoe- cium lie between 2 and 3, whether the temperature range is 10° to 20° or 20° to 30°, so chemical reactions would seem to be im- portant. Euplotes also has coefficients between 2 and 3 for the range 20° to 30° which are to be explained in the same way. At lower temperatures, however, the condition for Euplotes is different. Over therange 15° to 25° and 10° to 20° the coefficients are generally little more than unity. A few coefficients of less than 2 were observed over this range for Paramoecium, but were decidedly the exception, instead of being the rule as with Euplotes. These low coefficients suggest complicating physical Minutes. 10 20 30 40 Graph 3, D Effect of temperature upon the toxicity of hydrochloric and of butyric acids to Euplotes. A. HCl, 0.0005 N; B. HCl, 0.0002 N; C. Butyric, 0.001 N; D. Butyric, 0.0005 N. factors perhaps masking a chemical reaction. Especially remark- able are the results obtained with Euplotes in acetic and butyric acids. LENGTH OF LIFE IN MINUTES COEFFICIENTS 10° 15° 20° 30° 10-20° 20-30° nn 10 17 Gruiedial bas 3 0.0005 N...........| 25 20 45 20 | 0.55 | 2.2 ODEN. .......... 9 2 25 SL 0.96.4, 31 Su fe eas 33 35 55 2 | 0.60 | 2.2 456 M. E. COLLETT In the other acids (formic, tartaric, HCl, etc.) lowering the temperature either lessens the toxic action slightly or leaves it quite unaffected. Butyric and acetic, however, are more toxic at 15° than at 20° and still more toxic at 10°. Indeed, they are almost as toxic at 10° as at 30°. If they consistently become more toxic with falling temperature throughout the range 30° to 10°, this might be explained as due to rate of adsorption, since this is a process favored by low and diminished by high temperatures. But this cannot be the case, since at 20° they are less toxic than at 30°. Probably some further processes are involved, such as solubility, etc. A striking example of this sort has been described by Hans Meyer (quoted by Lillie, ’16). The solubility of ethyl alcohol and chloral hydrate is increased by rise in temperature, while the solubility of salicylamid and monoacetin is decreased, and corresponding with this difference in solubility is the fact that the narcotic power of the former group is increased by rise in temperature, that of the latter decreased. Such conditions apparently complicate my results. It appears from these experiments that temperature defi- nitely influences the toxic action of acids. The degree of influ- ence, however, varies with the acid, its concentration, the organ- ism, and the temperature chosen. These irregularities show that the action is by no means simple and that it probably involves both physical and chemical factors.’ D. Importance of the H ion The general order of toxicity obtaining in equinormal solu- tions gives some indication of the mode of action. Full data are given in table 1 and graph 1. The order runs: HCl > dibasic 1 Lillie (17) found that long exposure to sea-water at 2° to 6° would suffice to initiate development in starfish eggs, and that the rate of activation in butyric acid is greater at 6° than at 8° to 10°. These results are not strictly parallel with those obtained in the present experiments upon cilia, in which the lower temperatures (10° to 15°) increase the effectiveness of butyric acid, although alone they are not effective; but in both series of experiments Doctor Lillie’s explanation seems applicable, viz., that a change in the physical condition of the structural colloids due to cold may alter their permeability or other properties and so affect their resistance to certain reagents. TOXICITY OF ACIDS TO CILIATE INFUSORIA 457 and hydroxy acids > fatty acids except formic. This is roughly the same as the order of dissociation and indicates that the H ion is important. But if the H ion is the sole factor in deter- mining toxicity, the order of toxicity and the order of dissocia- tion should agree perfectly, which, as the following table shows, is not the case. Toxicity and dissociation Toxicity aT 0.0002 n DISSOCIATION CONSTANTS (SCUDDER) Paramoecium Stylonichia Euplotes Salicylic Salicylic Salicylic HCl HCl HCl HCl Salicylic LAS 1085 Formic Tartaric Formic Oxalic C010 Lactic Lactic Citric Lactic La xaos ai Peas Oxalic Formic 1296 oO" Tartaric Formic Benzoic Tartaric es { Malonic Lactic Malonic 1.64 X 10-3 Benzoic \ Oxalic Phthalic Benzoic 6.64 XK 10-5 Acetic Phthalic Tartaric Citric Sia, X 1044 Phthalic Acetic Malonic Acetic Ueesy >< 10ee Citric Citric Acetic Propionic 1.42 X 105 Butyrie Butyric “Prin Butyric LL els Valeric Valeric Valeric Valeric Le S10” Still clearer evidence should be given by solutions of the vari- ous acids at the same H ion concentration, for if the H ion is the only toxic factor, they must all be equally toxic. Solutions were therefore made (without buffers) of all the acids in the series at four of the following concentrations, viz., Psa 3.5, 3.85, 4.0, 4.15, 4.75. ‘The concentrations were determined by means of indica- tors tetrabromphenolsulphonphthalein and methyl red (Clark and Lubs, ’17) in the usual way by comparison with acetic acid Na acetate standards. The method is not perfectly exact, but upon determining the Px of a series of equinormal acids (0.0005, 0.0002, 0.0001 N) both colorimetrically and by calculation from the dissociation constants the indicators proved to be accurate to within about 0.1 Pu. The results obtained by this method are indicated in table 3 and graph 4. It will be seen at a glance that the toxicity of different acids of the same Pu is by no means identical. ‘This 458 M. E. COLLETT L Pp 3.5 3.85 4.0 45 474 Graph 4, E, F Toxicity of acids of equal Py at 20°C. (Euplotes above, Paramoecium below. Time given in minutes.) TOXICITY OF ACIDS TO CILIATE INFUSORIA 459 TABLE 3 Toxicity of solutions of equal Py at 20°C. Time in minutes 3.49} 11) 17 | 10} 11) 12) 33 3 (el | 84| 40) 60 | 35) 35) 40/42 | 20 19 | 15] 15/11 /10 | 35 3.99) 60 100} 60) 60\70 | 55 55 | 45 | 33/33 40 | 30 14 g B g|/e 3:49| 4] 23) 6] 5| 5) 2) <2 <2 <}| <1 3.84] 13] 11 | 12] 11) 15/10] 7 7 | -tol-o 6a aas Paramoecium.... 4 |3.99| 21] 35 | 35| 23] 23|16 | 113} 10] 11] 11/11 /13 | 10 4.14} 30/100 |105| 60} 55/20} 17 | 16| 20] 20 (|4.74] 35 BOY aa | 25°) 28 35 |35 | 30 3.49] 4| 5| 3] 3] 3] 23) <1 | <1d] <3] <1 3.84] 15] 20 | 23] 13] 20/10 | 6 7| 6{| 6] 83] 63 17 Stylonichia...... 4 |3.99] 35] 48 | 80] 25} 27/30 | 138 | 14] 15] 11}11 |15 | 10 4.14 75| 7550 | 30 | 28} 30] 23 | |4.74 30 3.49) 15] 15 | 15) 15) 20) 5 | <3 <2) salicylic > formic > dibasic and hydroxy acids > acetic and butyric; but in solutions of equal Ps acetic and butyric become equal in toxicity to the cyclic acids, and the dibasic and hydroxy acids together with hydrochloric become least toxic of all. Formic, which in equinormal solution is more toxic than the other fatty acids, probably because of its greater dissociation, becomes less toxic than the rest when the solutions have the same Ps. Somewhat the same conditions hold with Euplotes. In solutions of equal Px salicylic retains 460 M. E. COLLETT Order of toxicity PARAMOECIUM PARAMOECIUM 0.0005 nN 0.0001 n Px 3.5 4.0 Salicylic Salicylic Valeric free Hydrochloric Hydrochloric Butyric Salicylic Formic Tartaric Propionic Benzoie Pee ace Acetic eee Benzoic Malonic Formic Butyric | Ggsttie Formic Oxalic Valeric Malonic Oxalic Hydrochloric Formic Acetic ee ee Hydrochloric Butyric Butyric Malonic Tartaric (eee Oxalic EUPLOTES EUPLOTES 0.0005 N 0.0002 n Px 3.5 | 4.0 Salicylic Salicylic Salicylic ee ee hia HCl Butyric Benzoic Benzoic Formic penne Butyrie HCl Oxalic Acetic Propionic Oxalic Benzoic Formic Acetic Tartaric Tartaric Malonic HCl Malonic Malonie (Goi pee. Acetic Acetic HCl | Citric Butyric Butyric Citric Formic Oxalic Malonie its position at the top of the list, the other cyclic and the fatty acids (except formic) stand next, while the hydroxy acids and hydrochloric are much less toxic than the rest. Evidence of like nature has been obtained in experiments upon very different sorts of material. In every case the relative efficiency of equi- normal solutions failed to parallel their H ion concentration exactly, and solutions of equal Px were always unequal in effect. Only one conclusion is possible. The H ion is an important factor in the physiological effects of acids, but at least in organic acids some other factor or factors are involved which makes them more effective than their Px alone would lead one to expect. TOXICITY OF ACIDS TO CILIATE INFUSORIA 461 These experiments with solutions of equal Px also give evi- dence of specific differences in resistance. The order of resist- ance to every acid is Paramoecium > Stylonichia > Euplotes > Vorticella. There is considerable evidence of specific difference in resistance to H+ among other organisms, such as bacteria (Michealis, 714, 715; Kemper, 716) it is noteworthy, however, that here the limit of tolerance of a particular species is not the same in all acids (Wyeth, 718). This is true also of Paramoe- cium; thus, a concentration of Ps 3.5 kills Paramoecium in four minutes in HCl, five minutes in tartaric, less than one minute in valeric, etc., and so with the other organisms. The order of P, G 6. 3,5 3.85 4.0 4.75 Minutes. 10 205) 30,1640" 5:60: 60° § 70)! 780 Graph 5, C Relative resistance of Paramoecium, Stylonichia, Euplotes, and Vorticella to equal Ps of hydrochloric acid. Temperature 20°C.; time in minutes. toxicity of the different acids at the same Px also is slightly dif- ferent for each organism (table 3) as is the case with equinormal solutions. It will be remembered that Kopacweski (14) found the same condition in enzymes. E. Importance of anion and molecule Although there is ample proof from many sources that the effect of an acid is not determined solely by its H ion concentra- tion, there is very little evidence indicating exactly the part played by anion and molecule. One way of approaching the matter is to compare the relative toxicity of a series of acids 462 M. E. COLLETT with the relative toxicity of their salts. I have not experimented with salts, but there are interesting data to be had from other experiments. The order of toxicity to Lupinus seedlings (Kah- lenberg and True and True, ’96) is as follows: Acids equinormal, salts equimolecular. Acids: HCl > benzoic > salicylic > formic = propionic > butyric. Salts: Na— benzoate > salicylate > formate > propionate > butyrate > chloride. The order for equinormal acids is almost the same as that obtained in my experiments: that is, HCl is more toxic than benzoic, and benzoic in turn is more toxic than the fatty acids. Since there is little difference in dissociation among the salts, their order, unlike that of the acids, is a measure of the relative toxicity of the anions. Salts of the cyclic and of the fatty acids are more toxic than NaCl and must therefore have more toxic anions. ‘These are salts of the very acids which in solutions of equal ionic concentration proved to be most toxic to cilia. Practically the same conditions hold in the haemolysis of blood corpuscles (Fiihner and Neubauer, ’07, and Hoeber, 710). Acids: HCl > formic > acetic > propionic. Salts: Na—Salicylate > benzoate > formate > acetate > butyrate. The similarity between these experiments with equimolar salts and my experiments with acids in solutions of equal Ps makes it seem probable that the acids which are most toxie in these solutions owe their effectiveness to the anions as well as to the Hion. The fact that a different order of toxicity obtains in normal solutions may perhaps be explained as follows: If an acid is not highly dissociated, even though it has a slightly toxie anion, it is less toxic than a more highly dissociated acid with a non-toxic anion (for instance, butyric as compared with HCl). If, however, the anion is very toxic, as is the case with salicylic, the acid in spite of its slighter ionization may equal or even exceed a more completely ionized acid such as HCl in toxicity. Another method of investigating this point has been suggested by Klocman (714). To various concentrations of a weak acid a TOXICITY OF ACIDS TO CILIATE INFUSORIA 463 fixed amount of HCl is added; this depresses the dissociation of the weaker acid and produces a solution containing chiefly the organic acid molecule, H+ and Cl-. If the anion of the organic acid is toxic, this mixture should be less toxic than the original solution, provided the H ion concentration is not greatly altered. Klocman found that the acetic anion was toxic, but did not try the method with other acids. In the present experiments the method was somewhat modified and was applied to a num- ber of organic acids of the same concentration (0.0002 N). The Px of each solution was first determined by the use of indicators, then enough HCl was added to increase the Px slightly, by a known amount when the toxicity of each of these mixtures had been determined HCl was again added, increasing the Px decid- edly. With each addition the ionization of the organic acid was depressed further, until finally most of it was in molecular form and the H ions were derived chiefly from the HCl. By deter- mining the toxicity of the acids alone and again after each addi- tion of HCl it was possible to follow the effect of the decreasing dissociation of the acids and at every stage to compare them with HCl of the same H ion concentration. The figures obtained in this experiment are recorded in table 4 and expressed graphi- cally in graph 4. | It will be seen at once that some of the acids become less toxic as the H ion concentration is increased, formic lactic and acetic, for example, in the experiments with Paramoecium. This means that as the anions decrease in number the acid be- comes less harmful. In other cases, addition of HCl increases the toxicity, but only until it is equal to that of HCl of the same Pz. Thus phthalic by itself is more toxic than HCl, but as ionization is depressed approaches it more closely; that is, In graphic form, the curves are at first distinct, but as Px is increased through the addition of HCl the phthalic approaches the HCl curve and finally coincides with it. The same is true of oxalic, benzoic, citric, acetic, propionic, butyric, valeric. In these cases, too, the anion must be toxic, since depression of lonization does away with the greater toxicity of the organic acid as compared with HCl of the same Ps. Since these acids 464 M. E. COLLETT 36 4.0 415 4.75 Minetes 5 10 iS 20 25 30 35 90 100 Graph 6, H,I Effect upon toxicity at 20°C. of depressing ionization. (Para- moecium above and Euplotes below.) TOXICITY OF ACIDS TO CILIATE INFUSORIA 465 when in molecular form are not more toxic than HCl of the same Px, it seems probable that their molecules do not exert a toxic action. The curves for Euplotes present a slightly different picture. Certain acids (oxalic, malonic, tartaric) become more toxic as ionization is depressed. It will be noted that these are all dibasic acids. They are, however, so nearly like HCl that it is difficult to draw any definite conclusions from the results. With TABLE 4 Effect of depressing ionization Py LENGTH OF LIFE IN MINUTES ACID Alone Plus HCl Paramoecium Euplotes I Ii iil I II III I II Til Rormier ss 5 4 Se95riacO) | o. Sh 4 73% | 10 25 Zh | oD INCODIC HES <0, syne s | heat eo Ta leseS5:'| 722, De 14 60 90 | 60 Wii ee A aa osota oO 30 21 6 60 100 Ripe 12 IBMUVTIC. 02.73. 0----| 4-40 canoe | o.co | 60 ae 13 70 110 WeUleriGs: <5 .0, cee: 5 |) 4.6) WeoaSaaleo.oo. | --45 25 13 70 105 | 60 Sale yilter sk 2s): - .| 3.6 3.49 2 4 4 4 BOOZOIC...02. 5 25...| 4.05 | 3297 W738. 85 11 12 11 45 70 | 40 Picbalieess - -).,.<5..0 . ATs |) pics (Pais), | eer 18 14 70 60 | 80 WACTIC aa86 5-2 =. .2 | OM orSm tose n Us 7 20 50 60 Oa liGy tat veh ed TOLD 3.6 9 8 40 15 Mallanicn. shi 28s38295 3.85 || 10 11 85 65 PLAT GARIC 27 as isl LOnoOu| Os Ouse 9 9 9 60 45 Gite... sane scare l\pos0on|Los6 25 20 6 60 100 | 25 the first addition of HCI the toxicity of many of the other acids is decreased (lactic, formic, acetic, propionic, butyric, valeric, citric, benzoic), proving that their anions must be toxic. Indeed the decrease proceeds so far that the mixtures of these acids with HCl become much less toxic than HCl of the same P; alone. Phthalic does something of the same sort, viz., first becomes equal to HCl and then falls far below it in toxicity. When the second lot of HCl is added to some of the acids (ben- zoic, citric, acetic-valeric) a curious result is obtained. The toxicity, instead of being decreased as by the first addition, is 466 M. E. COLLETT sharply increased, until the mixture again becomes equal to pure HCl. This is indicated though much less clearly in the Paramoecium curves for citric and acetic, propionic, etc., but not for benzoic. Until further work is done I cannot with assur- ance offer a definite explanation of this condition, but the results so far suggest that the molecule is to some extent capable of antagonizing the action of the H ion. Thus as dissociation is depressed and the anion is removed from the sphere of action, antagonism between molecule and H ion may cause the mixture to become less toxic than HCl of the same Px, and this antag- onism may cease to be important only when the H ions largely out-number the molecules present. From this series of experiments we find that the following anions are toxic. To Paramoecium: formic, acetic, propionic, butyric, valeric, benzoic, phthalic, lactic, oxalic, malonic, tar- taric, citric. To Euplotes, all of these except oxalic, tartaric, lactic, and possibly malonic. These findings agree with the conclusions suggested by the previous experiments, viz., that the anions as well as the H ions are sometimes toxic, and that the same acid need not act upon different organisms in exactly the - same way. F. Nature of the toxic action At higher concentrations (0.0005 N) Paramoecium discharges trichocysts, and as it coagulates turns rapidly opaque without great swelling, but at 0.0002 N the swelling is pronounced and no trichocysts are discharged. At death the protoplasm be- comes granular and the nucleus stands out sharply. Vorticella and Euplotes, too, become granular and in all the vacuoles grow large and rigid before death. The cilia themselves swell and become sticky, the beat grows irregular and slows until finally the cilia dissolve (see also Koltzoff on Carchesium). Stylonichia swells and some of its vacuoles increase in size, then anteriorly, at a point near the edge, the protoplasm dissolves and releases large apparently insoluble droplets. The cilia stop only when the body is completely disintegrated and must there- fore be more resistant than the rest of the cell. Careful obser- TOXICITY OF ACIDS TO CILIATE INFUSORIA 467 vations under high power, such as Greeley made, of the changes induced in the physical state of the protoplasm might throw light upon the nature of the effect of various acids upon the tissue of different organisms. It is certain that the H ion is exceedingly important in the swelling of muscle (Loeb, ’98, no swelling in hypotonic salt solution unless acid is also present) as well as in the swelling of other hydrophilous colloids, such as fibrin, gelatin, ete. (Loeb, 19; Proctor, ’16; Fischer, ’18). It seems probable that the H ion is also important in the swelling of cilia, perhaps the specific differences observed are due in part to differences in the colloids present. Another factor seems to be surface tension. Working with salts, Clowes (’16) found that slight changes in surface tension produce great changes in the physical state of oil-soap emul- sions, and he suggested that the physiological effects of these salts were due to changes induced in the protoplasmic emulsion. There is evidence in other experiments of profound changes in the surface tension of protoplasm induced by acids. Ham- burger (713) found that ingestion of India ink or charcoal par- ticles by phagocytes was depressed in too high concentrations of fatty acids, but was stimulated in more dilute solutions, and this stimulating action he attributed to changes in surface tension. Koltzoffi (14) used Carchesium (a colonial relative of Vorti- cella) for the same purpose, and found that at some concentra- tions acids increased the ingestion of ink particles, and at shghtly higher concentrations produced visible evidence of change in surface tension in the accumulation of ink particles on the cilia. The optimum concentration for this softening effect as well as for stimulation of phagocytosis varied with the acid used. Another factor in toxicity is Bpoid solubility. * Benzoic acid is much more lipoid-soluble than the fatty acids and should therefore penetrate a lipoid rich membrane and attack the cell contents more rapidly. The fact that benzoic is more toxic than valeric to Euplotes but not to Paramoecium suggests the pres- ence of some lipoid in Euplotes which is not present in the same 468 M. E. COLLETT concentration in Paramoecium. In many experiments upon other material benzoic has been found very effective, and simi- lar specific differences in susceptibility have been observed which suggest differences in lipoid content (Harvey, ’14; Crozier, 16; Haas, ’16; Loeb, 713). Peters and Burres (’09) conclude from their experiments upon the toxicity of Cu salts to P. aurelia that toxic effect is due to injury to an essential enzyme and not to direct chemical injury of the protoplasm. If the normal metabolic processes of the cell are interrupted, as they would be by the failure of an im- portant enzyme, it is obvious that the chemical and physical balance of the whole cell would be affected. Possibly something of the sort may be involved in the toxic action of acids upon the cilia of Paramoecium, Euplotes, etc., and may account for the specific differences involved. This explanation is of course purely hypothetical at present. There are many other possible factors in toxicity, but conclusions are difficult and uncertain the moment one ventures beyond very simple and obvious comparisons. SUMMARY It is found that the relative toxicity of a series of acids varies decidedly with the concentration, and therefore it is unwise to base conclusions as to mode of action upon results obtained with only one concentration. ‘The fact that power of penetra- tion also varies greatly with concentration makes it probable that the same is true of many of the physiological effects of acids. There are also great differences in the effects of acids upon different species: even organisms so closely related as the infusoria used in these experiments show characteristic differ- ences. When tested in the same solution, one species may be two, four, or even twenty times as resistant as another; and in addition the order of toxicity of the series of acids is somewhat different for each species. Another factor in determining rela- tive toxicity is temperature. Ordinarily toxicity increases with increase in temperature and decreases with decrease in tem- perature, but the degrees of influence exerted by temperature TOXICITY OF ACIDS TO CILIATE INFUSORIA 469 varies with several factors—the species used, the acid, and the concentration. The acids are unequally affected by different temperatures. The coefficients for Paramoecium and Euplotes are scarcely alike at any point. Most of the coefficient of 10° for Paramoecium lie between two and three and suggest a chemi- cal reaction. The coefficients for Euplotes, however, are rarely so high and agree better with the idea of a physical factor. Indeed, with butyric and acetic, though not with any of the other acids tried, a decrease in temperature below 20° accelerates the toxic action upon Euplotes, so that their toxicity at 10° is nearly as great as at 30° and much greater than at 20°. © A comparison of the acids at various equinormal concentra- tions shows that a rough parallelism exists between toxicity and degree of dissociation, as would be expected if the H ion is the most important factor in toxicity. When, however, the acids are compared in solutions of equal Px, it is evident that other factors must enter, for the toxicity of the various acids is mark- edly different. The order of toxicity, unlike that of equinormal acids, is closely similar to the toxic order of salts of these acids. This suggests that the anions of at least some must be toxic. Additional evidence of the toxic action of certain anions (mainly of the fatty and the cyclic acids) is afforded by the action of acid mixtures, in which the ionization of the organic acid is progres- sively diminished by addition of a strong acid, HCl. All the acids used bring about swelling followed by precipita- tion of part of the cell contents and by solution of the cilia. There is clear evidence of change in surface tension, for the cilia always become sticky before they stop beating. It seems prob- able that the toxic effect of the various acids depends upon their solubility in the tissue as well as upon capillary activity and the changes in colloidal state wrought by the H ion. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 29, No. 3 470 M. E. COLLETT CONCLUSIONS 1. The order of toxicity of a series of acids varies with the concentration, the temperature, and the species. The action is therefore not simple. . 2. The H ion is an important factor, for the toxic order of equinormal solutions is roughly parallel with the order of dissociation. 3. The H ion is not the only factor in toxicity, for in solutions of equal Px the acids are not equal in toxicity. 4. Depression of ionization shows that the anions of certain acids are toxic to both Paramoecium and Euplotes, viz., formic, acetic, propionic, butyric, valeric, citric, benzoic, phthalic, salicylic. The anions of oxalic, tartaric, lactic, and malonic are toxic to Paramoecium, but not to Euplotes. 5. The temperature coefficients indicate that both chemical and physical reactions are probably concerned in the toxic effect of acids. 6. A most marked irregularity is shown by acetic and butyric acids, in that their toxicity to Euplotes (though not to Para- moecium) is greatly increased by temperatures below as well as above 20°C. TOXICITY OF ACIDS TO CILIATE INFUSORIA 471 BIBLIOGRAPHY Barratt 1904 Die Wirkung von Sduren und Basen auf Par. aurelia. Z. Allgem. Physiol., Bd. 5. 1905 Die Addition von Siuren und Alkalien durch lebendiges Proto- plasms. Z. Allgem. Physiol., Bd. 5. Cuark, J. F. 1899 Toxic effect of deleterious agents. Bot. Gaz., vol. 28. CrarKk-Lusss 1917 Colorimetric determination of H* concentration. J. Bacteriol., vol. 2. Ciowes 1916 The protoplasmic equilibrium. J. Physical Chem., vol. 20. Crozier 1916 Cell penetration by acids. J. Biol. Chem., vol. 24. 1917 Sensory activation by acids. Am. J. Physiol., vol. 45. Cummine 1916 Specific differences in haemolysis with acids and alkalies. J. Inf. Dis., vol. 18. Date anpD Mines 1911 Action of acids upon skeletal muscle. J. Physiol., vol. 42. FIscHER AND Hooker 1918 Swelling of gelatine in acids. J. Am. Chem. Soc., vol. 40. FUHNER UND NEUBAUER 1907 Haemolyse durch Substanzen homologer Reihe. Arch. Exp. Pathol. u. Pharm., Bd. 56. GREELEY 1904 Paramoecium and electrolytes. Biol. Bull., vol. 7. Haas 1916 Permeability of plant cells to acids. J. Biol. Chem., vol. 27. HAMBURGER UND DE Haan 1914 Wirkung von Fettsiuren auf die Phagozytose. Cent. Physiol., Bd. 29. Harvey 1914 Permeability of cells to acids. Intern. Z. phys.-chem. Biol., Bd. 1. Heatp 1896 Toxic effect of dilute solutions of salts and acids. Bot. Gaz., vol. 22. Horser 1910 Der Einfluss einiger organischen Alkalisalze auf Muskeln. Pfliiger’s Arch., Bd. 134. KAHLENBERGER AND TRUE 1896 Toxic action of dissolved salts and acids. Bot Gaz., vol. 22. Kemper 1916 Precipitation of bacteria by a definite Py. J. Inf. Dis., vol. 18. Ktocman 1911 Kinetik der Giftwirkung von organischen und unorganischen Saiuren. Dissertation, Miinchen. Kourzorr 1914 Phagozytose von Charchesium. Intern. Z. phys.-chem. Biol., Bd. 1. Kopaczewski 1914 Die Affinititsreihe und biologische Wirkung der Sauren. Intern. Z. phys.-chem. Biol., Bd. 1. LANDSTEINER UND PrasEK 1913 Siéureflockung der Blutstromata. Z. Immun., Bd. 20. Litre, R. 8S. 1916 Theories of anaesthesia. Biol. Bull., vol. 30. 1917 Temperature coefficients in the activation of starfish eggs by butyric acid. Biol. Bull., vol. 32. Logs, J. 1898 Physiol. Untersuchungen iiber Ionenwirkung. Pfliiger’s Archiv, Bd. 69, 71. : 1913 Artificial parthenogenesis. Chicago. 1917 Membrane permeability to electrolytes. J. Biol. Chem., vol. 32. 1919 Amphoteric colloids. J. Gen. Physiol., vol. 1. 472 M. E. COLLETT Lorsp unpD WasTENEYS 1912 Entgiftung von Saéuren durch Salze. Biochem. Z., Bd. 39. Micuartis 1914 Wasserstoffionenkonzentration. Berlin. 1915 Saéureagglutination von B. typhosus. Deutsch. med. Wochen- schrift, Bd. 41. OstrrHouT 1914 Effect of acid upon permeability. J. Biol. Chem., vol. 19. Pau, Brrnstein, Reuss 1910 Kinetik der Giftwirkung geléster Stoffe. Biochem. Z., Bd. 29. PrtTerS AND Burres 1909 The diastatic enzyme of Paramoecium and the killing concentration of CuSO,. J. Biol. Chem., vol. 6. Proctor AND Witson 1916 The acid-gelatin equilibrium. J. Am. Chem. Soce., vol. 109. : Ritter, G. E. 1912 Die giftige und formative Wirkung der Siuren auf die Mucoraceen. J. wissenschaftliche Botanik, Bd. 52. RoaF anD WHITLEY 1906 Action of acid and alkali on the tadpoles of R. tem- poraria. Biochem. J., vol. 1. Scupprr 1914 _Conductivity and ionization constants of organic compounds. Van Nostrand. Szitt 1906 Sdureintoxikation. Pfliiger’s Arch., Bd. 115. 1909 Vergiftigung mit anorganischen und organischen Sauren. Pfliiger’s Arch., Bd. 130. Taytor 1917 Specificity of antiseptics. Lancet, vol. I. True 1900 Toxic action of acids and Na salts. Am. J. Se., vol. 9. Watpum 1914 Bedeutung d. Py fiir Haemolyse. Biochem. Z., Bd. 63. WEINLAND 1894 Chemische Reizung des Flimmerepithels. Pfliiger’s Arch., Bd. 58. WIinoGRADOFF 1911 Beitrige zur Kenntnis der physiologischen Wirkung der organischen Siuren. Dissertation, Miinchen. Wyretu 1918 Effect of acids on B. coli. Biochem. J., vol. 12. Unecer 1916 Toxic action of acids and alkalies on fish. Chem. abstr. 10. Ste a kt 7 absasel suerte fis ie im -W i a Wr mu heey rr rin 4 : { (4 id °° im “athe eral Fa 4 o =) iv? i" , oil f ee 7y +) POET (wi Resumen por el autor, J. A. Dawson. Universidad Yale. Estudio experimental de un Oxytricha amicronucleado. I. Estudio del animal normal con una descripcién de su canibalismo. Durante 289 generaciones se ha cultivado Oxytricha hymen- ostoma Stokes—desde el 10 de Julio de 1917 hasta el 17 de Noviembre de 1917— y el autor conserva aun vivos cultivos en masa de los descendientes de estos cultivos. No ha visto un micronticleo en ningtin estado de los diversos animales, durante la historia de los cultivos. No ha habido singamia, si i bien el autor cree que el organismo en cuesti6én ha pasado frecuentemente por un estado fisico semejante al en que tiene lugar la singamia. En este momento: a) los animales se fusionan por parejas a semejanza de la unién de los conjugantes, y, b) tiene lugar el canibalismo. Cuando acaece la fusién en parejas, los animales o bien permanecen fusionados hasta su muerte o se separan. En caso de separacién los organismos en cuestién contintan repro- duciéndose y no presentan sintoma alguno de estar en una condicién de depresién. Cuando el canibalismo tiene lugar los animalesingeridosson digeridos rapidamente y los que sobreviven vuelven a adquirir el tamafio y estructura tipicos. El cani- balismo produce como efecto el aumentar ligeramente la cantidad de divisiones durante un corto tiempo. No se ha observado enquistamiento. Esta raza amicronucleada de Oxytricha puede vivir indefinidamente, en apariencia, bajo condiciones ambientes favorables sin conjugacion, autogamia o endomixis. Translation by José F. Nonidez Carnegie Institution of Washington AUTHOR’S ABSTRACT OF THIS PAPER ISSUED BY THE BIBLIOGRAPHIC SERVICE, OCTOBER 13 AN EXPERIMENTALSTUDY OF AN AMICRONUCLEATE OXYTRICHA I. STUDY OF THE NORMAL ANIMAL, WITH AN ACCOUNT OF CANNIBALISM J. A. DAWSON Osborn Zoological Laboratory, Yale University TWENTY FIGURES CONTENTS MeN LEOAUEhION A 22 saan ote si sojisk be RA pease ea tehlas ts else ee eee 473 Pee vinierial and anebhods.s 5 cc si ois. 3 eon, . te ae eee ae cae at 473 516 INDEX HOTIC reactions of the honey-bee, Apis mellifera da: “Dhe.s-casacesent caer 343 AT. Studies on inbreeding. IV. A fur- ther study of the effects of inbreeding on the growth and variability in the body weight of the albino.............. 71 Rat). Studies on the reputed endocrine func- tion of the thymus gland (albino......... 311 Reactions of Chromodoris zebra. Sensory... 261 Reactions of the honey-bee, Apis mellifera L. The photious. fess ch. b25 1 sehieie ots RR 8 343 Renewal of vitality through conjugation. Uroleptus mobilis Engelm. II........... Responses of Chiton. Thesensory....... 157 ENSORY reactions of Chromodoris zebra 261 Sensory responses of Chiton. The...... 157 Sex in the hermaphrodite molluse Crepidula plana. III. Transference of the male- producing stimulus through sea-water. SEUGIES Oley csc ewer cs nede tee ea. Tee 113 Stimulus through sea-water. Studies on sex in the hermaphrodite molluse Crepidula plana. III. Transference of the male- LOO UCINE pert escae cele e sere clelniayisiercs 113 AKENOUCHI, Marsuzrro. Studies on the reputed endocrine function of the thymus gland (albino rat).............. 311 Thymus gland (albino rat). Studies on the reputed endocrine function of the........ 311 Thyroidectomy. Growth and development of Amphibia as affected by............... Toxicity of acids to ciliate Infusoria. The... 443 ROLEPTUS mobilis Engelm. II. Re- newal of vitality through conjugation.. 121 ARIABILITY in the body weight of the albinorat. Studies on inbreeding. A further study of the effects of inbreed- ing on the growth and.................. Vitality through conjugation. Uroleptus mo- bilis Engelm. II. Renewal of EIGHT of the albino rat. Studies on inbreeding. IV. A further study of the effects of inbreeding on the growth and variability in the body.......... 71 — QL The Journal of experimental 1 zoology J68 V.e29 cop.2 Biological & Medical Serials PLEASE DO NOT REMOVE CARDS OR SLIPS FROM THIS POCKET UNIVERSITY OF TORONTO LIBRARY Paty ie iy cae 5 Sim —ler- dag a OMe teienes os sbviaiphs aint ee pee Tr Bott 939 arse eds em. % Leke ee ys. Te Se ginny ICY en Se Fol anvaga, * Y Nel Sg ey : . “~ > “ een ¥ aioe dete a * . . sateen Ayana i 2 Ae hey ‘eee STP et = : ores SOR ae Pr oe on ren Mahan peat omen ae : ener tapen ns way iteniphernc tipncey oe Leek ie vs a nctiabnpetete artrenncn gee te) Mea 2 Wee woe =, iat eal ated