ne a atta iis tit puntata aie ae ii RH ih a ttt | aia Ha : HPHeane eC aeaaCGE ts } Hk ea seh ate hit ot tt eeegtgtes t iyi thi) Het basen btt ity He ‘ata ieee - | Msotpanee Sancecetynntnt rt itt : it un ( f ut ti plot Partai pes int intel Plot ol ais ; Dipl sfotal eran ai oe iat 0 ia peittt itt a | i t? } Wareieieiriels ‘ 7 ” she a SSE Sees SS5e3s _~ eoete 2 —e~ 33 26= t iit duet bese ss bitte : tt sah it ti i ti HL Soto sters =st. a) Misistel : ey —— <-? Al ' eberehtht tet sabcectttt tg it digit iit Hh i Ha unt Tt ie pitt Tart atFtN Hi it ut i?) ye 4 ee ee Hn - Ks 52525! it ot) { i i tiie abegites { Ut NST taT ee etek teSeaes heat ata a here cz het { it : } f ad a te be we ae Pees erate. -~ 7 cht fitit ; 5 shite t if it thr rerbee tees it bd nese tite hr hha iit Hint uit : } a : ft * beh if bit thst Abit a ae tit A uu itt tt Ti ' at {! it i ii the ie ty hi C BiH wii ptt att itt tite itt i474 {iy ihe ee 3 . ripiels rie : ft itl a ae eee oe ee tee oe ee pe ee Epa wie aad eet at pe hela te Ht Hib th tt sl Se ttt cit tat — peat ate ts co ep esos eres i Hiren ate i | pinta i a Heertthth ti i pt lo tit} } hh 919) tt : ot of : bere iit if it ut ; git i istic rath eters bestest jess sectggt eh bistbibehtt ieee HSUseaey ge dtm : Het } t 4 Het HTP eLe rora te t . Hi sath ht ih scinutrea Ate saat d ; eeaettheats soto sa ah i itt “Hts i! i ibferbiel Hol oi bt otalotsty tit i Uheheh hts aria t duane Re ‘ Feet tee , ahi Aphett iit tH if t i ify it Ut ch : Mita ° Sk att bei da Hi eta) stanabatyt nts toh vaae di Ste i itt nite sitehas 19) 7 aheh? | {4 Tals! att ren tit Oo heh i t Cleat tit} ah Toto Hit he i S x | : Ctitih rate itn thas 4 ) ut j at { ae f t t t ; fit thegl iy : ' bat ft it it stat bial ce i i Bsa hehe th +t 1 Hhatts ‘ ti eletes if {tit oh) ? ’ nh seitet Thi Al 7 ahe t hebyhed tt | piript any th) tebsastyotet Hi nt he : a pen es 4 Math tit ; z ree Het at ate tetity it feat ith it ith i Co tthe - ne hit t “ ah tt besa eda hihdS teehee tt nity SE TOUR N Ade OF EXPERIMENTAL ZOOLOGY EDITED BY WILLIAM K. BROOKS FRANK R. LILLIE Johns Hopkins University University of Chicago WILLIAM E..CASTLE JACQUES LOEB Harvard University University of California EDWIN G. CONKLIN THOMAS H. MORGAN University of Pennsylvania Columbia University CHARLES B. DAVENPORT GEORGE H. PARKER Carnegie Institution Harvard University HERBERT S. JENNINGS CHARLES O. WHITMAN Johns Hopkins University University of Chicago EDMUND B. WILSON, Columbia University AND ROSS G. HARRISON Yale University Manacinc EpitTor VOLUME IV THE JOURNAL OF EXPERIMENTAL ZOOLOGY BALTIMORE 1907 2.0 CONTENTS No. 1—February, 1907 CHARLES RussELL BARDEEN Abnormal Development of Toad Ova Fertilized by Spermatozoa Exposed to the Roentgen Rays. With Five Plates Wit1amM B. Herms An Ecological and Experimental Study of Sarcophagidz with Relation to aice Debris: Wath Seven Freres: <4 ..44 54404 a eee eee 45 Sara WHITE CULL Rejuvenescence asthe Result of Conjugation. jy...) 4-). 9.0008 - 19) eee 85 GEORGE LEFEVRE Artificial Pathenogenesis in Thalassema Mellita. With Six Plates....... gI Jacques Lore Concermng the Vheory of Tropisms® 2) ici weer st, ach. ia ae eee eee I5I Frank W. BANCROFT The Mechanism of the Galvanotropic Orientation in Volvox............. 157 No. 2—June, 1907 CuHar.Les R. StocKarD The Influence of External Factors, Chemical and Physical, on the Development of Fundulus Heteroclitus. With Seventeen Figures...... 165 On@. Graser Movement and Problem Solving in Ophiura Brevispina. With Five Birparicer 2 Soe Serene Soh Po eauen AEM OIRY Gute Re ene = WOE aE OE 203 IsaBEL McCracken Occurrence of a Sport in Melasoma (Lina) Scripta and its Behavior in Ferediaya,* with OnevPlate: sy. eet crate tins Sr ayes ok arte eee, ee 221 Ross GRANVILLE HARRISON Experiments in Transplanting Limbs and their Bearing upon the Prob- lems of the Development of Nerves. With Fourteen Figures.......... 239 E. G. SPAULDING The Energy of Segmentation. An Appplication of Physical Laws to Organicdivents:: taste, ook con a ee eee ee Oe ieee Se hr ee 283 No. 3—September, 1907 A. J. GoLDFARB Factors in the Regeneration of a Compound Hydroid, Eudendrium amosunn..” WithiewosPigares 235, 6 ame catcher ate athe oh ht ates 357 C. M. Cuitp Studies on Regulation. XI. Functional Regulation in the Intestine of Cestoplana. “With Twenty Text iomnes’ ay cee ee ee eee 357 S. J. Hotmes The Behavior of Loxophyllum and its Relation to Regeneration. With Sever Fiewres, ... 02, ¢ 0.0 eo wh oes an em me ers Soe ee 399 Regeneration as Functional Adjustment. With One Figure............. 419 GerorGE L. STREETER Some Factors in the Development of the Amphibian Ear Vesicle and Further Experiments on Equilibration. With Six Figures............ 431 Benjy. C. GRUENBERG Compensatory Motions and the Semicircular Canals. With Two Figures 447 No. 4—October, 1907 AxicE M. Borinc A Study of the Spermatogenesis of Twenty-two Species of the Membra- cid, Jassidz, Cercopidz and Fulgoridz. With Nine Plates.......... 469 Witi1am Morton Barrows The Reactions of the Pomace Fly, Drosophila Ampelophila Loew, to Odorous.Substances: With Five Figutes 3. e eo ee ee ae ee 515 Epcar Davipson CoNGDON The Effect of Temperature on the Migration of the Retinal Pigment in Decapod Crustaceans: Wath Seven Figures: 6:,2—2- eo. eee 539 S. Moreu tis Observations and Experiments on Regeneration in Lumbriculus. With nie rene: 78 vets. 963 Saket as Se. ae se ne ee eee 549 Wo. E. KeLuicotr Correlation and Variation in Internal and External Characters in the Common Toad (Bufo lentiginosus americanus, Le C.) With Six . ENB IMGs poring bl tre tolne ie hs jist seca otra mi et eee ev 575 ABNORMAL DEVELOPMENT OF TOAD OVA FER- TILIZED BY SPERMATOZOA EXPOSED TO THE ROENTGEN RAYS BY CHARLES RUSSELL BARDEEN, M.D. Professor of Anatomy, University of Wisconsin, Madison, Wis. Wiru Five Pirates REVIEW OF THE LITERATURE The marked alterations in structure and function produced in living tissues by the Roentgen rays and the rays of radium have excited much interest in the medical profession. Professional biologists have paid less attention to the subject, although it is one which promises to be of great importance to those engaged in experimental morphology. While probably all living tissues may be injured by sufficient exposure to these rays, there are great differences in the degree of susceptibility of different tissues and organisms. The power of assimilation of foodstuffs and their transformation into living structures complex in character seems «to be that which is most altered in exposed tissues. ae Bacteria are less easily affected than are most of the higher organ- isms. Many of those who have exposed cultures of bacteria to the Roentgen rays have failed to discover any checking of the nor- mal multiplication of the bacteria.t Rieder,’ on the other hand, found growth of bacteria inhibited by sufhcient exposure to the rays. [hose who have studied the action of radium on bacterial cultures have usually found the growth of the bacteria inhibited 1Park, Medical News, 1896; Mink, Miinchener med. Wochenschrift, 1896; Lyon, Lancet, 1896; Delepine, British Medical Journal, 1896; Atkinson, Nature, lvi, p. 600, 1897 (negative results with mucor, bacteria and oscillaria). ?Rieder: Miinchener m d. Wochenschrift, 1902. THE JOURNAL OF/ #ERIMENTAL ZOOLOGY, VOL. IV, NO. I. 2 Charles Russell Bardeen by its rays. Koernicke* found the development of yeast and bacteria inhibited by exposure to the rays but that the exposed cultures, if transferred to fresh unexposed gelatine, could begin to grow again. The action of the Roentgen rays and radium on the higher plants has been recently studied by M. Koernicke,* who gives ref- erences to the previous literature on the subject. Koernicke finds that the Roentgen rays serve to check the growth of germinating seeds. Immediately after exposure there may be a slight quicken- ing of growth. The retardation of development appears only after a latent period. Ifthe exposure to the rays is not too intense or prolonged, the retardation of growth is merely transitory; if sufhciently intense, there is permanent inhibition of growth. Ex- posure of dry seeds and swollen, non-germinating seeds to the Roentgen rays had no immediate inhibitive effect on subsequent germination. Instead a slight quickening of germination was sometimes seen. Ultimately, however, the roots of the exposed seeds ceased to grow. According to Maldiney and Thouvenin* exposure to the Roentgen rays hastens germination of seeds. Lopriore’ found the germination of pollen inhibited by the rays. A pupil of H. Becquerel® found that while 24 hours’ exposure of certain seeds to radium rays caused no marked diminution in power of germination, an exposure of a week or more inhibited the process. Koernicke did not inhibit growth in the seeds he studied (Vicia faba, Brassica napus, Papaver somniferum), but he found that exposure to radium rays caused a subsequent check- ing or inhibition of growth in germinating seeds and that non- germinating seeds either dry or moist, if sufficiently exposed, 3Aschkinass and Caspari: Arch. fiir die Ges. Physiol., lxxxvi, 1901 (action attributed to a and B rays). W. Hoffmann: Hyg. Rundsch, xiii, s. 913, 1903 (bacteria killed). Dixon and Wigham: Nature, Ixix, pp. 81, 1903 (bacteria checked in growth). Danysz: Compt. Rend., cxxxvi, p. 463, 1903. H.v: Baeyer; Zeitschrift f. allgem. Physiologie, iv, p. 79, 1904. 4M. Koernicke: Berichte der Deutsche Bot. Gesellschaft, xxii, s. 163, 1904. °M. Koernicke: Ueber die Wirkung von Réntgen- und Radiumstrahlen auf den pflanzlichen Organismus. Berichte der Deutschen botanische Gesellschaft, xxii, pp. 148-166, 1904; xxili, pp. 404- 414, 1905. ®Maldiney and Thouvenin: Revue gen. de Bot., x, p. 81, 1898. 7Lopriore: Estr. dal’ Nuova Rassegna, Catania, 1897. ®H. Becquerel: Comptes Rendus, t. 133, pp- 712, Igol. Abnormal Development of Toad Ova 3 germinate normally at first and then show a temporary or per- manent inhibition of growth. Koernicke found in studying the tissues of exposed plants that it is the nuclei that seem especially affected by the rays. ‘There is no visible direct injury to the cytoplasm. ‘The effect on the nuclei is proportional to the length of exposure. ‘The nuclei of the vegetative cells are more resistant than the pollen mother-cells. Owing to injury to the nuclei of the pollen mother-cells the pollen cells may be abnormal in appearance but thisis due tg the actionof the injured nuclei on the cytoplasm. In the pollen mother-cells twenty-four hours after five hours’ exposure to radium rays the nuclear threads during mitosis fell into small double segments which were much smaller and more numer- ous than normal for Lilium martagon (the species studied). Division of the chromosomes took place somewhat in the normal manner but the daughter chromosomes seldom passed simul- taneously toward the poles. Occasionally two or three daughter nuclei were formed on each side of the equator. If the pollen mother-cells shortly before the diakinesis of the nuclei were exposed for twenty-four hours, they showed a day later a clumping of the chromosomes at the center of the nuclear cavity. ‘The spindle figure was strongly developed. Sometimes the spindle poles seemed split. Exposure to radium rays up to ten hours seemed to have little effect on the daughter cells of the pollen mother-cells, but a one to three days’ exposure had a marked effect. “‘Uhenuclei were brought into an abnormal stage which partially resembled a resting stage. The nuclei of the tetrads arising from exposed cells were also very abnormal. The effects of the Roentgen and radium rays on a protozoa have been studied by Schaudinn, Joseph and Prowazek, Zuelzer, and others. Schaudinn® showed that individuals of several species of protozoa may be killed by exposure to the Roentgen rays for a few hours, while others are not thus susceptible. Joseph and Prowazek” found that Paramecia and Daphnia show a negative "Schaudinn: Archiv. f. die gesammte Physiologie, Ixxvii, p. 29, 1899. 10Joseph and Prowazek: Zeitschr. fiir allg. Physiol., Bd. i, 1902. 4 Charles Russell Bardeen tropism toward the Roentgen rays, and that the protoplasm of Paramecia seems injured by the rays. I have found paramecia very resistant to the Roentgen rays. ‘“[welve hours’ exposure to powerful rays made no difference in the form or rate of division in P. aurelia or P. candatum. Exposure to the rays seemed not to influence conjugation. M. Zuelzer in protozoa which were exposed to radium rays under the microscope noted primarily an injury to the nuclear substance. ‘The cytoplasm appeared affected later than the nuclei. Great variation in susceptibility was noted in different species. Zuelzer gives a brief summary of the previous literature on this subject. _ Zuelzer™ found insects, and Danysz® insect larve affected by radium rays. The action of Roetgen and radium rays on the fertilized eggs of Ascaris megalocephala has been studied by Perthes.'® Perthes found that there is a retardation of the cleavage of eggs exposed to the rays and that the later divisions are either inhibited or are abnormal. In the latter case the eggs give rise to either irregular masses of cells or to abnormal embryos. ‘The effects depend largely on the degree of exposure. ‘The nuclei of the exposed eggs are markedly affected. “The chromosomes of the dividing nuclei are irregular in shape and sometimes seem to be divided abnor- mally into smaller parts. ‘The spindle figures appear normal. In fresh water Planarians Bardeen and Baetjer showed that exposure to the Roentgen rays destroys the power of regeneration. Schaper® has shown that exposure to radium rays produces sim- ilar effects. The action of Roentgen rays on the developing eggs of sea urchins was found to be negative by G. Schwarz.’ In the spring of 1903 the writer failed to get any positive results on exposing the eggs of sea-urchins and teleosts to the Roentgen rays, but these MM. Zuelzer: Archiv fiir Protistenkunde, v, p. 358, 1905. Danysz: Compt. Rend., cxxxvi, 1903. 18Perthes: Archiv fiir klinische Chirurgie, Ixxi, 1903; Deutsche med. Wochenschrift, Nr. 17-18, 1904. MBardeen and Baetjer: This journal, i, p. 192, 1904. 1Schaper: Anat. Anzeiger, xxv, p. 298, 1904. 6G. Schwarz: Wiener klin. Wochenschrift, xvi, s. 714, 1903. Abnormal Development of Toad Ova 5 negative results he attibuted to the use of an apparatus from which rays of merely moderate intensity could be obtained and which could not be used for prolonged exposures. G. Bohn‘ has reported the production of arificial parthenogenesis in Strongy- locentrotus lividus by exposure to the rays of radium. The effects of the exposure of the fertilized eggs and the larve of Amphibia to the Roentgen and radium rays have been studied by a number of investigators. P. K. Gilman and F. H. Baetjer* have shown that the eggs of Amblystoma exposed to the Roentgen rays exhibit a brief period of accelerated growth and then mark- edly abnormal development. If the exposure is not too severe the tadpoles may recover; if sufficiently severe they develop into monstrosities and soon die. A. Schaper® obtained somewhat similar results with frogs’ eggs exposed to radium rays, although he failed to find a period of accelerated growth immediately follow- ing exposure. Schaper also found that regeneration of the tail and limbs of ‘Triton larve is inhibited by exposure to radium rays. The wound heals and a mass of cells is accumulated in the region of the lost part but no specific regeneration takes place. O. Levy,” who has studied microscopically the specimens prepared by Schaper just before his untimely death, comes to the following conclusions : 1 In the period of cleavage of the ovum the rays may serve to check or inhibit cell division but cause no cell degeneration. Death may follow. 2 In the period of formation and early differentiation of the organs (generative self-assimilation) marked degenerative abnor- malities appear in many of the organs, especially in the neural tube, retina and nose. ‘The optic lens, the pigment layer of the retina, the aural vesicle, the chorda dorsalis, and the myotomes appear comparatively little affected. The heart is frequently rudimen- tary. lhe tubules of the pronephros are frequently dilated. In general, the effects are the most serious in those tissues in which growth and complex differentiation are normally most rapid. %G. Bohn: Comptes Rendus de l’Acad. des Sciences, Paris, cxxxvi, pp. 1012, 1085, 1903. 8p. K. Gilman and F. H. Baetjer: Amer. Jour. of Physiology, x, p. 222, 1904. 194. Schaper: Anat. Anzeiger, xxv, p- 298, 1904; Deutsche med. Wochenschrift, xxx, 1904. 00, Levy: Archiv f. Entwicklungsmechanik, xxi, P- 130-152, 1906. 6 Charles Russell Bardeen 3 In the period of the finer differentiation of the organs (func- tional development of Roux) the primary effect of exposure appears to be on the blood vessels, the tissues suffering because of the effects on the blood vessels. G. Bohn, who studied the effects of radium rays on the eggs and larve of the frog and turtle, found that if growth is slow exposure to radium rays prevents the attainment of full size; if rapid and associated with tissue differentiation, radium causes degenera- tion of the tissues and although at first accelerating, ultimately stops development. He considers that everything leads one to think that the rays of radium affect the chromatine because it is fromthe activity of this substance that there results assimilation and growth. The action of the Roentgen rays on the hen’s egg has been studied by Gilman and Baetjer.” These investigators found a preliminary period of accelerated development followed by retard- ation of development and the production of abnormal embryos. J. Tur? studied the action of radium rays on developing hen’s eggs and obtained various deformities. He found the embryonic area of the germinal disc more sensitive to the rays than the periphery and obtained some germinal discs without embryos. He found the cells of the ectoderm more easily affected than the yolk cells. G. Schwarz,” on the other hand, from his experiments onthe action of radium rays on the hen’s egg, concludes that the action of the rays is due to a decomposition similar to that of a dry distillation brought about in the albumenoid bodies of the cell. He explains the effect of the rays on rapidly growing tissue as due to their special power to decompose lecithin. Perthes®® found that the wing of a chick exposed to radium rays was checked in its develop- ment. The experiments with radium and Roentgen rays on plants, invertebrates and the lower vertebrates, though of great scientific 1G. Bohn: Op. cit. 2Gilman and Baetjer: Op. cit. 37. Tur: Comptes Rendus des Séances de la Société de Biologie, t. lvii, 1904. "4G, Schwarz: Archiv f. gesammte Physiologie, C. 532, 1903. *6Perthes: Archiv f. klin. Chirurgie, lxxi, 1903. Abnormal Development of Toad Ova 7 value, have been comparatively few in number. ‘The practical application of the rays in medicine has led to a much more exten- sive series of observations on the effects of the rays on man and mammals. The physiological effect of the Roentgen rays first noticed was the skin burn which after an interveniug latent period usually follows much exposure to the rays, and which may give rise to great thickening of the skinor to ulceration. Similar lesions were found to follow exposure to the salts of radium and like sub- stances. Clincal experience as well as experiments on various mammals soon showed that the more deeply seated tissues, as well as the skin, are affected by the rays, but that the different tissues are variously affected. Some tissues seem to be affected directly, others seem to be affected only indirectly through alterna- tions produced in the general metabolism or in the blood supply. General toxic effects following the exposure to the Roentgen trays have been described in man by Seguy and Quenisset, *Walsh,” Kienbock,” Baermann and Linser,” A.S. Warthin,® D. Edsall,**and many others.” Similar effects have been described in many mam- mals. ‘larkhanoft* experimented not only with several small mammals (mice, rabbits and guinea pigs) but also upon frogs and birds. He found that when long exposed the animals died with symptoms of paralysis. Rodet and Bertin* attributed the death of animals exposed to X-rays to a meningo-myelitis. Numerous sub- sequent experimenters have described toxic effects, paralytic symp- toms and death in small mammals after prolonged exposure to the Roentgen and radium rays. Danysz** was one of the first to study this action of radium on small mammals. He found that a pro- 6Seguy and Quenisset: Bulletin de Acad. des Sciences, 1897. "Walsh: British Med. Journal, 1897. *Kienbock: Wiener med. Presse, 1901. ?°Baermann and Linser: Miinchener med. Wochenschrift, li, s. 918-994, 1904. 304. S. Warthin, International Clinics, 15th series, vol. iv, p. 243, 1906. 31D, Edsall, Journal American Medical Association, xlvii, p. 1425, 1906, 32For a list of the literature on this subject, see Warthin, op. cit. 38Tarkhanoff: Gaz. degli ospedali, 1897. (Cited by Warthin.) 34Rodet and Bertin: Gaz. des Hép., 1898. (Cited by Warthin.) 35Danysz: Comptes Rendus de l’Acad. des Sciences, Paris, 1903, 1904. 8 Charles Russell Bardeen longed application of the tube containing the radium salt to the head or spine of a small mammal was followed by paralysis, ataxia, convulsions and death. In the central nervous system marked hemorrhagic lesions were found after death. Similar alterations in the central nervous system have been described by Heineke,* Scholtz,*7 Obersteiner,® and others. Obersteiner, who paid espe- cial attention to the lesions of the central nervous system, con- cludes that “the various phenomena which are observed in the exposed mice, including the death which follows sufficient exposure to the rays, in greatest part are, directly or indirectly, merely an expression of a general disturbance of the circulation and of meta- bolism produced bythe radium rays.’ Obersteiner does not con- sider the nerve cells specifically susceptible to the rays although they are, more easily than many tissues, disturbed by altera- tion in the circulation or general metabolism. ‘The general dis- turbances produced by the rays are indicated by the increased elimination of nitrogen discovered to take place by Baermann and Linser® after severe exposure. Lepine and Bonlud* had previously shown that alterations affecting metabolism take place in the pancreas, liver and blood after exposure to the Roentgen rays. The great susceptibility of the nervous system to the indirect, if not to the direct, action of the rays, is shown not only by the lowering of the reflexes, apathy and paralysis which precede death in animals sufficiently exposed tothe Roentgen or the radium rays, but also by the injury of the retina and secondary atrophy of the optic nerve which Birch-Hirschfeld** has described. ‘Trophic disturbances may likewise possibly be due to the injured nervous system. Obersteiner” has described a severe panophthalmitis and a gan- grene of the tendons of the feet, the ears and the nose following exposure of mice to radium. While there is doubt concerning the specific sensibility of the 36Heineke: Miinchener med. Wochenschrift, 1, s. 2090, 1903. 37§choltz: Deutsche med. Wochenschrift, xxx, s. 94, 1904. 38Obersteiner: Arbeiten aus dem Neurologischen Institute, Wien, xii, p. 86, 1905. 3®Baermann and Linser: Op. cit, 4°Lepine and Bonlud: Comptes Rendus de l’Acad. des Sciences, Paris, t. xxxviii, 1904. “'Birch-Hirschfeld: Miinchener med. Wochenschrift, 1904. “2Obersteiner: Op. cit. Abnormal Development of Toad Ova 9 nervous system to radium, most investigators are agreed concern- ing the marked action which radium and Roentgen rays have on the vascular system. Although the action of the rays on plants and ova shows conclusively that other than the vascular tissues may be directly affected by the rays, the changes in the blood ves- sels in young oradult mammalian tissues are among the most marked lesions found after exposure to the rays, so that the effect on other organs, as wellason the central nervoussystem, has been described by many as due to a secondary action on the tissues through a primary injury of the vascular system. While even the well-known lesions of the skin have been ascribed to action on the blood vessels or nerves, Oudin, Barthélmy and Darier® found in a study of Roent- gen ray alopecia in guinea pigs that the layers of the epidermis were affected, the follicles and glands were atrophied but no altera- tions in the blood vessels and nerves of the dermis or subcuta- neous were to be observed. Scholtz,‘ in an important contribution, concluded that both the nuclei and the cell protoplasm of the epi- thelial cells of the mammalian skin are injured by the rays, but that the effect on the connective tissues, elastic tissues, muscula- tureandcartilageisslight. “he skin on bothsides of a rabbit’s ear may be affected whenit is exposed toraysonone sideonly. Theeffect on the connective tissues he thinks due to a secondary inflam- matory reaction. Gassmann,* from a study of a deep Roentgen ray ulcer, concluded thatthe changes in the blood vessels formed the primary cause of the ulcer and its resistance to healing. Rudis- Jicinsky,* from an experimental study of X-ray burns in guinea pigs and rabbits, concluded that there is aninflammatory reaction to the X-rays followed by a development of fibrous tissue and a thickening of the walls of the blood vessels, and that degenerative changes followthe impaired blood supply. Baermann and Linser,* from a study of the action of X-rays on lupus, concluded that an endarteritis and contraction of the blood vessels with degenerative OQudin, Barthélmy and Darier: Monatsch. f. prakt. Dermat., xxv, 1897. “Scholtz: Archiv f. Dermatologie u. Syphilis, lix, pp. 87, 241, 419, 1902. ‘SGassmann: Fortschr. a. d. Geb. d. Roentgenbestr., 1899. *6Rudis-Jicinsky: New York Med. Jour., 1902. “Baermann and Linser: Miinchener med. Wochenschrift, 1904. IO Charles Russell Bardeen changes in the connective tissue 1s the primary result of the action of the rays, the epithelium being only secondarily affected. While there is little doubt but that the deepest layer of the epidermis is affected primarily by the Roentgen and radium rays, as shown by Scholtz,*8 the recent work of Scholtz,“ Halkin,® and others, has proved that exposure to radium rays, at least, also has a primary effect on the intima of the blood vessels. Whether the effect on the collagen of the connective tissues described by Unna,°* and others, is secondary to the vascular changes or is a primary effect of the rays cannot at present be conclusively answered. Exposure to Roentgen or radium rays has a marked effect not only on the blood vessels but also upon the blood and especially upon the blood-forming organs. Baermannand Linser® found no changes in the blood exposed to Roentgen rays. ‘The irradiation of serum caused, however, a decrease of haemolytic power toward blood corpuscles not exposed. Milchner and Mosse® found red blood corpuscles resistant to the Roentgen rays. London* noted a darkening of the arterial blood, and Grinbaum® hemolysis of the red corpuscles after exposure to the radium rays. In the spleen, lymphatic glands and bone marrow the results of irradiation are more marked than in the blood. Miulchner and Mosse® found in the bone marrow of irradiated rabbits degenera- tion of lymphoid and myeloid leucocytes but no destruction of the red corpuscles. Heinecke®™ exposed rabbits, white mice, guinea pigs and dogs to the Roentgen and radium rays and found an elective action on the lymphoid elements of the spleen, lymphatic glands and bone marrow. ‘These organs ultimately become much reduced in cell content. Warthin® began independently and “Scholtz: Op. cit. *8Scholtz: Deutsche med. Wochenschrift, xxx, s. 94, 1904. 50Halkin: Arch. f. Dermatologie u. Syphilis, xv, s. 201, 1903. 51Unna: Deutsche med. Zeitung, 1898. 52Baermann and Linser: Op. cit. °3Milchner and Mosse: Berliner klin. Wochenschrift, xli, s. 1267, 1904. 54London, Berliner klin. Wochenschrift, xl, s 523, 1903. 58Griinbaum. (Cited by Obersteiner.) 58Milchner and Mosse: Op. cit. 5’Heinecke: Deutsche med. Wochenschrift, 1903, 1904. 8Warthin: International Clinics, 1906. Abnormal Development of Toad Ova II carried farther than Heinecke an experimental study of the action of the Roentgen rays on the blood-forming organs. Warthin’s paper contains an extensive list of references to the literature relating to this subject, of which use has been made in preparing the present paper. Warthin employed white rats, rabbits, Belgian hares, and guinea pigs in his experiments and subjected them to brief and to prolonged exposures. Exposures of five hours or more caused death, usually in from two to five days. In allanimals exposed destructive changes were found in the lymphoid elements of the splenic pulp and follicles, marked by a degeneration of phagocytes, giant cells, and the epitheloid cells of the follicles. In all animals the effects lasted for some time and were not 1m- mediately followed by regeneration. A haemolytic action was indicated by the great increase of blood pigment in the tissues. Warthin suggests, however, that this may have been due in part to disturbances of splenic function. A fatty degeneration was noted in the lymphoid tissues. ‘The disintegration of lymphocytes was seen within 14 minutes after exposure and the cells continued to disintegrate for several days. ‘The greater part of the nuclear débris was quickly removed. The Roentgen rays caused also a destruction of the lymphoid cells of the lymphatic glands. ‘The small lymphocytes were destroyed before the other cells. Slight irradiation caused fatty degeneration; intense irradiation, nuclear degeneration. Regeneration sometimes took place after irra dia- tion was discontinued. ‘The effects inthe bone marrow were less intense than those inthe spleen andlymphatic glands. ‘The large lymphocytes and myelocytes were chiefly attacked, the small lymphocytes not showing the marked disintegration found in the spleen. No effect on the red cells was discerned. ‘There was an undoubted inhibition of white cell production in the marrow after irradiation. Pusey,®* Senn,” and a number of other American physicians, introduced the treatment of leuceemia by the use of the Roentgen rays. Pusey* attributes to Dr. A. J. Ochsner the first suggestion 5°Pusey: Jour. Amer. Med. Association, 1902. 8°Senn: New York Med. Journal, 1903. “Pusey: Jour. Amer. Med. Association, 1905. te) Charles Russell Bardeen of this treatment. A large number of papers on the subject have been reported. A list may be found in the paper of Warthin, men- tioned above. While nearly all of the leukemic cases treated have shown at first a marked improvement in the blood condition and general symptoms, there is some doubt as to the possibility of a permanent cure being brought about by irradiation with the Roentgen rays. Many of the earlier cases reported as cured have since relapsed or died. Warthin studied tissues derived from three patients who had been treated by X-ray irradiation for leukemia. In one there was a picture of an aleukemic lymphocytoma or lymphosarcoma with no lymphocytes in the vessels; in the second the tissues presented the picture of a myeloid leukemia without any changes attributable with certainty to the Roentgen ray treatment; in the third the immediate effect of the treatment was to cause in the diseased glands a fatty degeneration and necrosis of the atypical cells forming the glands. “This necrosis was usually followed by an apparent sarcomatous infiltration of the surrounding tissues. In the first and third cases renal lesions sug- gested toxemia. The symptoms leading to death Warthin thinks are due to an intoxication resulting from the disintegration of cell proteid. The Roentgen rays have proved of more certain value in a num- ber of other affections, especially in acne, rosacea, nevi, lupus vul- garis, cutaneous carcinoma, and some other superficially placed diseased conditions.” In some instances the benefit seems to be due to a slight inflammatory reaction set up by the rays; in others to the alterations produced in the tissues When the rays have direct access to the tissues, diseased cells, especially the abnormal cells which constitute carcinomata and sarcomata, may apparently be destroyed by Roentgen or radium rays before permanent injury is inflicted on the normal tissues. When filtered through normal tissues the rays seem largely to lose an elective action on the abnormal cells.® 82See Pusey, Jour. Amer. Med. Assoc., 1905. ®3For a list of references to the literature on the action of the Roentgen rays on tumors, see Warthin International Clinics, 1906. On the action of radium irradiation on tumors, see H. Rieder, Verhand- lung der Gesellschaft deutschen Naturforscher und Aerzte, 1903, p. 278; H. Apolant, Deutsche med. Wochenschrift, xxx, s. 554, 1126, 1904; and Neuberg, Zietschr. f. Krebsforschnung, ti, s. 171, 1904. Abnormal Development of Toad Ova 13 Normal or slightly diseased tissues may be rendered carcino- matous or sarcomatous. Carcinomata have resulted from the action of the Roentgen rays on the skin; and sarcomata have arisen apparently from the action of the X-rays on the lymphatic glands.” It is well known that sterility can be brought about in man and other mammals by sufficient exposure of the testicles to the Roentgen or radium rays. Several operators have thus been ren- dered sterile. Of those who have studied the action of the rays on spermatogenesis mention may be made of Albers-Schonberg,® Frieben,® Scholtz,” Sedlin® and Philipp.” ‘The exposure to the rays causes destruction of the spermatogonia and brings about aspermia. Exposure of the ovaries to the rays causes similar disturbances. According to Halberstaeder” the ovaries are more susceptible to the rays than are the skin and testicles. | In reviewing the literature on the effects of the Roentgen and radium rays on living organisms it becomes evident that both forms of irradiation have essentially similar, if not identical, phys- iological action. “The Roentgen rays lend themselves the more readily to exposures powerful in volume and intensity but limited in time, the radium rays are as a rule much less voluminous but can more readily be applied over considerable intervals of time. The difference in the phenomena observed after exposure to the two sources of radiant energy may perhaps be ascribed mainly to this, although it is probable that the less penetrating rays of radium modify the effects produced by the 7-rays. The latter are supposed to be identical with the more penetrating of the Roent- gen rays. Schaper” found the frog larve exposed to radium ema- nations apparently adversely affected, yet Levy” was unable to find ®Warthin: Op. cit, 85Albers-Schcenberg: Miinchener med. Wochenschrift, 1903. 86Frieben: Miinchener med. Wochenschrift, 1903. Scholtz: Deutsche med. Wochenschrift, 1904. 88Sedlin: Fortschr. a. d. Gebiete d. Roentgenbestr., vii, 1904. ®®Philipp; Fortschr. a. d. Gebiete d. Roentgenbestr., viii, 1905. 70Halberstaeder: Berliner klin. Wochenschrift, 1905. “Schaper: Anat. Anzeiger, xxv, p. 278, 1904. Levy: Archiv fiir Entwicklungsmechanik, xxi, p. 142, 1906. 14 Charles Russell Bardeen in the larve exposed to radium emanation and preserved by Schaper, any evidences of tissue degeneration such as follows direct irradiation. The Roentgen rays and the rays emitted by radio-active sub- stances are known to cause ionization of gases, to affect photo- graphic plates, phosphorescent substances and glass. It seems evident that they also modify the chemical nature of living bodies. It is certain that in some instances, at least, as proved by Perthes® Thies,” Koernicke,® and others, the cell nuclei are primarily affected by the rays. An injury to nuclei of cells sufficient to destroy the normal influence over metabolism would suffice to account for all phenomena which have been observed. ‘The special destructive influence on cells undergoing rapid assimilation, multi- plication and differentiation may be accounted for by the very important réle played by the nuclei in these processes and an apparently greater susceptibility of the nuclei at such periods. The recent important and suggestive paper of F. R. Lillie,”’ on the ele- mentary phenomena of embryonic development in Chztopterus, shows what an active part is performed by the nuclei of cells dur- ing the early stages of embryonic development in furnishing sub- stances to the cytoplasm and in turn assimilating substances from the cytoplasm. While irradiation does not as a rule directly stop the phenomena of mitosis it evidently severely alters the productive activities of the nuclei of the cells and thus affects metabolism. Irradiation may also possibly directly affect the cytoplasm and intercellular substances. Thies” from a careful study of the action of radium rays on the different mammalian tissues and organs comes to the conclu- sion that all tissues suffer, although the elastic tissues are relatively resistant. “The adenoid tissues are the most susceptible. Other very susceptible tissues are the epidermis, the intima of the blood vessels, the parenchyma of the sex glands, voluntary muscle, white ™Perthes: Archiv f. klin. Chirurgie, lxxi, 1903; Deutsche med. Wochenschrift, 1904. “Thies: Wirkung der Radiumstrahlen auf verschiedene Gewebe und Organe. Mitteil. aus den Grenzgebreden d. Medizin und Chirurgie, xiv, 1905. ™Koernicke: Berichte der deut. bot. Gesellschaft, 1904-05. Lillie: Jour. of Experimental Zodlogy, iii, pp. 154-263, 1906. “Thies: Op. cit. Abnormal Development of Toad Ova 15 fibrous tissues and cartilage. Neither Scholtz® nor Danysz”® place the connective tissues or voluntary muscle among the tissues especially susceptible to irradiation. The physiological chemical effects of irradiation are not yet clear. While some authors attribute much to the decomposition of lecithin, others deny that irradiation can decompose this sub- stance. There is some evidence that the action of ferments 1s influenced by irradiation.” EXPERIMENTS The following experiments were designed primarily to test whether nuclear alteration produced by exposure to the Roentgen rays would alone suffice to cause the tissue alterations character- istic of this exposure. Forthis purpose | exposed spermatozoa to the X-rays and then fertilized eggs with these spermatozoa. Since the chief portion of the spermatozo6n is the nucleus and since the mass of spermatozo6n is insignificant compared to that of the egg it seems fair to conclude that if the exposure of the spermatozoon to the rays influences the development of the ovum the action of the rays must be on those unknown substances in the nucleus, or the protoplasm most intimately associated with the nucleus, which control the morphogenetic activities of the cell. The toad was selected because it is comparatively easy to get both males and females at the height of sexual maturity and the spermatozoa will live for several hours in water after removal from the body. My procedure was to collect several pairs of toads, separate the males from the females and wash the latter for some time in running water. From the testicles and Wolfian ducts of two of the males a thick suspension of spermatozoa was obtained. This was slightly diluted and divided into two portions, one of 7Scholtz; Archiv f, Dermatologie u. Syphilis, lix, 1902; Deutsche med. Wochenschrift, xxx, 1904. ™Danysz: Comptes Rendus de l’Acad. des Sciences, Paris, 1903, 1904. 80Schwarz: Archiv f. die gesammte Physiologie, c. s. 532, 1903; Baermann and Linser: Miin- chener med. Wochensch., li, 1904; Danysz J,: Comptes Rendus de l’Acad. des Sciences, Paris, 1903, 1904; Henri and Mayer: Comptes Rendus, 1904; Lepine and Bonlud: Comptes Rendus, 1904; Neuberg: Zeitschr, f. Krebsforschung, ii, s. 171, 1904; Harry: Journal of Physiol., xxix, 1904; E. Benjamin and A.V. Reuss: Miinch. med. Wochenschr., liii, 1906, 16 Charles Russell Bardeen which was kept for control, the other was exposed for from one-half to two hours to powerful X-rays from a fairly hard tube. At the end of the exposure I opened several females to obtain if possible eggs immediately ready for laying. Several short pieces of strings of eggs were placed in the control dish and others in the dish which had been exposed to the rays. After fifteen to twenty minutes these strings were removed and placed in large dishes of water. ‘The development of the control and the experi- ment ova was then watched from day to day until it seemed likely that the ova fertilized by the exposed sperm could no longer survive. Specimens of the control and experiment embryos were then preserved for microscopic study. Several of the experiments proved of negative value because not even the spermatozoa of the control dish proved capable of fertilizing the eggs, owing either to too great a lapse of time between the removal of the sperm from the males or to a lack of sufficiently ripe ova. ‘The season was so short that only a few successful experiments could be carried out but these were convincingly posi- tive. All eggs fertilized by the control spermatozoa developed normally. One of these was finally attacked by some parasitic organisms, but even in this the tissues and organs showed micro- scopically perfectly normal relations. All eggs fertilized by the exposed spermatozoa, over fifty in number, developed abnormally, with a single exception. This ovum developed into an appar- ently normal tadpole but this tadpole died before any of its fellow control tadpoles. It died at a time when it could not be imme- diately preserved for microscopic examination. Outline of Experiments I, May 4. Owing, probably, to too great dilution of sperm and too great a length of time between removal from the body and the attempt at fertilization of the ova, neither the irradiated or the exposed sperm fertilized any ova. II, May 5. Sperm exposed for two hours to X-rays about four inches from the tube. A large number of the spermatozoa were active at the end of this period After fertilization the control and Abnormal Development of Toad Ova 17 experiment ova were divided into two lots, one of which was kept in lake water and one in the harder city water. May 6. Between 50 and 75 per cent of the eggs in each of the four lots were in the blastula stage. May 7. All fertilized eggs were in the gastrula stage. May 8. Allfertilized eggs were advanced to the stage in which the yolk plug appears small. May 9-22. The eggs in the city water during this period devel- oped much more slowly than the lake watereggs. “Ihe control eggs in the city water were apparently nearly all fertilized and developed apparently normally though slowly up to May 12 when, owing to the cover being left off the shallow dish in which they were con- tained, and consequent evaporation of water during the absence of the writer from the city, the whole lot was destroyed. Of the experiment ova kept in city water about 50 per cent developed. On May 11 seventeen unfertilized and partially developed eggs were removed, leaving eleven nearly normal larve with very short caudal processés. On May 13 only six of these were still alive. These six were then transferred to lake water. “wo could move about by movements of the body and tail. ‘The other four showed little power of motion. These four larve and one of the two more motile ones did not grow much more in size but became quite abnormal in external form similar to those pictured in Plates II to V. One of the two more motile larvae developed into a tapdole fairly normal in external form, although it died before any of the controls which were kept in the lake water. Of the control ova in the lake water four eggs were unfertilized and nine developed into perfectly normal tapdoles which were kept alive for several weeks. Of the experiment ova on May 11 eight unfertilized or undeveloped eggs were removed, leaving seventeen larvae with slightly developed heads and very short caudal processes. On May 13 the control larvae could swim for short distances. None of the experiment larve could do so. Externally the latter had begun to exhibit various abnormalities which became more marked from this period on. About a dozen of these larvae were preserved on May 14.. The others all died a few days later, 7. e., in from ten to fourteen days after the 18 Charles Russell Bardeen exposure of the sperm. ‘The external form of some of the larve preserved is shown on Plates II to V. III, May 14. Unsuccessful because ova proved unripe. Be- tween May 5 and May 13, owing to a spellof cool weather, no toads could be obtained. IV, May 14. The day was very warm. At the end of the hour and half during which spermatozoa were exposed to the rays both experiment and control spermatozoa were non-motile and proved incapable of fertilizing any ova. V,May15. Spermexposed an hourandahalf. A considerable proportion of the control ova were fertilized. Only a few of the experiment ova were fertilized. Of these only one developed past the early cleavage stages. On May 22, atthree o’clock, p. m., this larva, which was small, unsymmetrical and ill-developed, appeared about to die. It was, therefore, killed and preserved. Micro- scopically it presented the features characteristic of the tadpoles developed from ova fertilized by irradiated sperm. VI, May 16. Sperm exposed one and a quarter hour. At the end of this period neither irradiated nor control spermatozoa proved capable of motion or of fertilizing. The day was an exceedingly warm one. VII, May 16. Sperm exposed one-half hour to the rays. Of the control ova only about 20 per cent were fertilized. These developed normally. Of the experiment ova only three or four per cent were fertilized. None of these developed past the gas- trula stage. Structure of the Larve At the time of or soon after the hatching of the larve abnormal structural differentiation began to make an externally visible dis- tinction between those experiment larvae which developed best and the control larve. In the experiment larve the tails all showed more or less abnormality of form and the larve with a few exceptions proved incapable of swimming, although some were capable of irregular movements of the body and tail. In length the experiment larve did not grow much after this period. In width and thickness the growth was greater but was in large part Abnormal Development of Toad Ova 19 due to abnormal distension of various regions with fluid. The external form meanwhile became most irregular. ‘Thisirregularity of external form is more easily illustrated than described. See out- lines on Plates II to V. Atthe period when the sketches were made of the larve here illustrated the control larvae had the ex- ternal form shown in the outline on Plate I. The rudiments of the hind legs were beginning to be externally visible. The organs of the experiment larva show the same irregularity of form as that shown by the body as a whole. Wax model reconstructions of the organs of each individual larva which has been preserved and of a series of stages of the normal larve of the toad would be necessary for a thorough study of the abnormalities which exist in the experiment larve. It has not seemed worth while to make so elaborate a study because each experiment larva is affected in such an individual way that it seems improbable that broader generalizations could be drawn from such a study than from the simple inspection of the serial sections of the larve. Control Larva of the Size of the Experiment Larve The control larve at the period when the experiment larve began to show marked abnormalities in external form were about at the stage described by Marshall in his Vertebrate Embryology as characteristic of the tadpole of the frog shortly after hatching. ‘The organs may be briefly described as follows: Nervous System—The central nervous system is slightly more advanced than at the stage of development described by Marshall in the newly hatched tadpole of the frog. The forebrain extends anteriorly between the olfactory pits. The end ‘brain is begin- ning to be divided anterior to the olfactory pit into two hemi- spheres. In the groove between these is a vascular plexus. ‘The pineal body extends forward above the undivided portion of the forebrain. It 1s composed of cords of cells interlaced with capil- laries. The optic stalks are patent near the neural canal but toward the optic cupsare narrow and are apparently partly filled by newly formed fibers. The infundibulum is large, its lateral walls are thick. ‘The roof of the midbrain has begun to thicken. The floor of the fourth ventricle on each side is very thick. The 20 Charles Russell Bardeen spinal cord may be followed well into the tail. It gradually becomes narrow and more rounded as one passes posteriorly. ‘The neural canal is relatively large. Organs of Special Sense—The nasal pit is connected with the pharynx by a solid column of cells. “The thickened lateral wall of the brain projects against the olfactory pit and the olfactory nerve Is in process of formation. ‘The lens of each eye is in con- tact with the ectoderm. Although differentiation of the lens 1s well under way it still contains a vesicular cavity. The sensory layer of the retina is thick and closely applied to the pigment layer. As mentioned above, the optic stalk near the optic cup 1s small and seems to contain nerve fibers. Each auditory vesicle is a simple closed sac with a short blind diverticulum. ‘The auditory ganglion consists of a mass of cells situated between the anterior portion of the auditory sac and the medulla. Peripheral Nerves—The trigeminal ganglion is large. It 1s not in connection with the ectoderm. The ganglion of the facial nerve cannot be sharply distinguished from that of the auditory nerve. The glossopharyngeal and the vagus ganglia are well de- veloped but it is difficult to trace the chief branches of these nerves for any considerable distance. ‘The spinal ganglia consist of well marked groups of cells lying ventro-lateral to the spinal cord. The nerve fibers of the spinal nerves are beginning to form in the proximal spinal segments. Alimentary Canal—The shallow stomodzum opens into the pharynx. ‘The lips and jaws of the stomodzum are beginning to grow forward. The broad pharynx communicates by gill slits with the exterior The external gills are fairly well developed. The opercular fold is beginning to be formed. . The anterior part of the lumen of the cesophagus is blocked up by epithelial cells. “The pulmonary diverticula are small lateral out- growths which extend but a short distance posteriorly from the cesophagus. Passing back from the blocked region of the cesophagus the gut makes a curve to the left about the liver and pancreas, and then Abnormal Development of Toad Ova 21 terminates in a mass of cells filled with yolk and presenting no well marked lumen. ‘Toward the posterior end of the trunk a lumen appears in the dorsal region of this mass. This lumen may be followed to the anus. The sinusoidal circulation of the liver is well established. The bile duct is of some length and opens into the intestine slightly to the left of the mid-line of the body. The anlage of the gall bladder may be seen below the bile duct. The anlage of the pancreas is marked by a deeply staining mass of cells posterior to the anlage of the liver on the left side of the body. ‘The ducts at this stage are not clearly marked in this embryo. _ Heart and Blood Vessels—The heart is an S-shaped tube, the lumen of which is filled with blood. ‘There are no well-marked trabeculz in the ventricle. “The main arteries and veins are filled with blood and may be readily followed. In general they seem to correspond with the diagrams of the blood vessels for the recently hatched tadpole given by Marshall (Embroylogy, p.170, Figs. 77 and 78). Reproductive Organs—The coiled tubules of the head kidney are surrounded by a vascular plexus. Distally the two Wolfian ducts have a common opening into the cloaca. They are patent throughout. Skeleton—The notochord and a loose mesenchyme formed of anastomosing cells constitute the skeletal tissues of this embryo. Muscles—In the region of the head the muscles anlages are marked by dense masses of tissue, in some of which specific differen- tiation has begun. In the region of the spinal cord the myo- tomes form a well differentiated segmental musculature on each side of the chorda dorsalis, the spinal cord and the spinal ganglia. Control Larva of the Age of the Experiment Larve The control larve, Plate I, at the period when most of the experiment larva were preserved for examination were about at the stage of development described by Marshall for the 12 mm. tadpole of the frog. The organs may be briefly described as follows: 22 Charles Russell Bardeen Central Nervous System—The olfactory lobes extend between the nasal organs. ‘They are beginning to be fused in the region shown in section a. Anterior to this they are more clearly separ- ated and posterior to it the lobe on each side is continued into a well marked cerebral hemisphere. Section b passes through the brain slightly anterior to the junction of the two lateral with the third ventricle. ‘The anterior end of the choroid plexus is shown between the two lateral ventricles and above this the posterior portion of the pineal body. ‘The latter extends forward for a considerable dis- tance beneath the ectoderm. At this period it contains a small vesicular cavity. The lateral walls of the third ventricle are thick (sectionc). ‘The optic stalk has been converted into an optic nerve but has a short patent lumen near the third ventricle. ‘This does not appear at the level of section c. ‘The infundibulum is large. The opening into it from the third ventricle is small, but beyond here it rapidly expands (section d). The optic lobes project dorsally on each side of the midbrain (section d). The cerebellum is marked by a slight thickening of the anterior margin of the roof of the fourth ventricle. “The ventro-lateral walls of the medulla are very thick (section e). ‘The folds of the choroid plexus are beginning to appear in its thin roof. The spinal cord is well developed and extends far back in the tail (sections 7, g and hb). The Randschleier, in the region of the trunk, is as thick as the layer of cells surrounding the central canal. Organs of Special Sense—Large nasal fossz extend from the exterior to the pharynx. ‘The epithelium of the medial wall of each fossa is very thick (section a). Czecal outgrowths are taking place from the posterior dorsal portion of each fossa. Bundles of nerve fibers may be traced from the lateral side of each olfactory lobe to the thickened epithelium of the corresponding nasal fossa. The eyes (section c) are well developed. ‘The several layers of the sensory part of the retina may be distinguished. A thin sclerotic coat is present. The well formed lens is separated by a distinct interval from the ectoderm. The auditory vesicles (section ¢) are being divided by the ingrowth of septa into the various portions characteristic of the adult ear. ‘The auditory nerve is undergoing rapid development. Abnormal Development of Toad Ova 22 Peripheral Nerves—The ganglion of the trigeminal nerve 1s highly developed. In section da portion of it may be seen on each side of the midbrain. ‘The main branches of the nerve may be followed for some distance from the ganglion. ‘The ganglion of the facial nerve is still so close to that of the auditory nerve that no sharp line of division can be seen in thesections. ‘The glosso- pharyngeal and vagus ganglia are likewise still difficult to differen- tiate from one another in the sections. ‘The spinal ganglia are well differentiated. ‘The motor and sensory roots and the main trunks of the spinal nerves may be distinguished without difficulty. Alimentary Canal—The lips and beak are highly developed. Section a shows on each side of the oral opening a section of the lower jaw tipped by an epithelial tooth. ‘The operculum is attached to the ventral and right sides of the body posterior to the gills. Section e shows the opercular cavity near the anterior end of the heart. ‘The internal gills are of considerable size. Only traces of the external gills remain. Over the region of the heart the ventral wall of the pharynx gives rise to a medial projection which extends in a posterior direction in the pharynx. Near the tip of this the trachea arises as a solid column of cells from the ventral wall of the alimentary canal near the junction of the. pharynx and cesophagus. More distally two tubular pulmonary processes extend back one on each side from the trachea along the dorsal wall of the body cavity. ‘These do not reach so far as the anlage of the pancreas. ‘The cesophagusnarrows rapidly posterior to where the trachea is given off. It diverges at this period toward the right side of the body (see section 7, immediately in front of and at the left of the chorda dorsalis). It joins the stomach in front of the pancreas. In section 7, on the left side of the figure (right side of the body), the stomach is shown cut through in two places and between these two sections of the stomach the anterior por- tion of the pancreas shows on each side. ‘The two regions of the stomach here shown are joined together anterior to this section. The ventral portion is joined with the cesophagus posterior to the level of the section. The dorsal portion of the stomach shown in this section passes posteriorly into a much coiled intestine. Part of the coils of the intestine pass ventral to the heart, anterior to the 24 Charles Russell Bardeen liver. ‘The liver and gall bladder are highly differentiated. The posterior end of the liver and the bile duct are shown between the cesophagus and stomach in section 7. The gall bladder lies dor- sal to the liver somewhat anterior to the level of this section. The entrance of the bile and pancreatic ducts into the gut takes place considerably posterior to this level. ‘The anus in this embryo ‘passes out between the anlages of the posterior limbs (section g). Heart and Blood Vessels—The heart is much larger than that in the control embryo previously described. ‘The ventricle 1s crossed by large trabecule (section e). ‘The auricle is beginning to be divided into right and left halves. The blood vessels are dilated with blood. ‘The vessels seem to correspond with those described for the 12 mm. tadpole by Marshall (Embryology, Fig. 76, p. 166). Genito-Urinary Organs—The tubules of the head kidney are greatly coiled. ‘The tubules of the mesonephros are differentiated near the cloaca. More anteriorly the anlages of these tubules consist of dense masses of cells. ‘The genital folds are not promi- nent. Skeleton—The differentiation of the vertebre has scarcely begun in this embryo. ‘The chief cartilages of the skull have appeared (see Marshall, Embryology, p. 262, Figs. go, g1 and g2). A moderate amount of mesenchyme surrounds the differ- ent organs. Muscles—The chief muscles of the head are clearly differen- tiated. In the trunk the myotomes are larger than at the preced- ing stage and the muscle cells are more highly developed. The abnormalities exhibited by the experiment larvae may be illustrated by a few typical examples. Larve No. 1, Experiment II, Plate II External Form—This embryo exhibits a marked dorsal flexion of the body near the middle of the trunk, so that the long axis of the tail is at right angles to that of the anterior portion of the trunk and the head. ‘The bend seems due largely to an abnormal dilatation of the hind-gut. The head is quite irregular in shape, the deformity of outline being due to great abnormality in the Abnormal Development of Toad Ova 25 internal structures. Posterior to the mouth the sucker is well developed on each side (section }). Central Nervous System—The brain ends anteriorly in a hollow rounded protuberance from which a vesicular ventricle projects on each side (section a). ‘The brain does not extend anteriorly to the olfactory pits. The walls of the telencephalon and the hem1- spheres are irregular in thickness and the cells and fibers are very abnormally disposed. Many pigmented and appparently degener- ate cells lie in the walls of the third ventricle and free in the ventri- cle. The pineal gland lies between the roof of the third ventricle and the ectoderm in a region posterior to section a. It has a thin wall and a hollow central cavity containing scattered cells which show evidences of degeneration. As one passes back the neural canal becomes much dilated. In the region of the optic stalk it is thin-walled (section 6). The hypophysis is somewhat dilated but is relatively normal in structure (section c). ‘The roof of the midbrain is rather thin. In the region of the ears the left side of the brain is much less developed than the right side (section 2) and posterior to the ears for a considerable distance the left side of the neural tube consists of hardly more than an irregular mem- branous wall. Near the middle of the trunk the spinal cord 1s more symmetrical, but posteriorly it is once more undeveloped on the left side (section g). Organs of Special Sense—The nasal pits on each side are con- nected with the pharynx by columns of cells (section a) which show near the pharynx an imperfect lumen. ‘The eyes are irregu- lar in form (section 6). A cornea has been differentiated. ‘The sensory portion of the retina is poorly developed. ‘The lumen of the optic stalk is dilated and extends into a space between the two layers of the retina. There are no nerve fibers (section 0). The pigment cells of the pigment layer are irregularly disposed and project out into the neighboring tissues. “The auditory sac- cule consists of a rounded pouch with a dorsal diverticulum (see d). ‘The auditory ganglion is clearly marked but no nerve fibers can be distinguished. Peripheral Nerves—The cells of the ganglion of the trigeminal nerve are large and well differentiated. Nerve fibers cannot 26 Charles Russell Bardeen readily be followed. ‘The ganglia of the ninth and tenth nerves are present on the right side only and are rudimentary. Rudi- mentary spinal ganglia are present except where the left side of the spinal cord is defective. No nerve fibers can be distinguished. Alimentary Canal—Lips and beak are differentiated, though abnormal in form. Section a passes through the lower jaw and lower lip. The oral opening into the pharynxis large. ‘The inter- nal gills are irregularly developed and contain no blood vessels. The cesophagus is patent and extends to the left into a thick- walled stomach. In the latter part of its course the lumen is di- vided into three flues, all of which open into the stomach (sections eandf). ‘The tracheo-pulmonary process is short, does not branch and has a slight lumen. The liver is developed posterior to the region of the heart and ventral to the stomach (section f). A few blood corpuscles can be seen in the much dilated capillary spaces of the liver. The pancreas is developed on the left side of the body posterior to the stomach. It consists of columns of cells. From the stomach the gut passes into an irregular group of coiled intestines, situated well on the left side of the body in the anterior region of the body cavity (section e), but in the middle more pos- teriorly (sections 7 and g). ‘The hind gut is greatly distended. The anus seems to be occluded. ‘The epithelium over the anus extends outward with irregular branching processes. Circulatory System—The heart is occluded and contains no blood. The only blood vessels clearly distinguished are a por- tion of the dorsal aorta, the mesenteric artery, and a few venous sinuses in the liver. In these vessels a few blood corpuscles are scattered about. Genito-Urinary System—The tubules of the pronephros are much dilated. ‘There are a few glomerular tufts but these con- tain no blood vessels. The Wolfian tubules are abnormally dilated (sections e, 7 and g). Skeleton and Connective Tissue—The connective tissue is in most regions excessive In amount. Pigment cells are irregularly scattered about. Most of the cartilages of the head seem to be present but the tissue is not thoroughly differentiated and there are some abnormalities of form. ‘The auditory capsule is not Abnormal Development of Toad Ova 27 differentiated. ‘The chorda dorsalis is fairly regular in form except at the anterior end, where some of the cells seem completely to have disappeared, and in the region of the medulla, where it 1S asymmetrical in places. Musculature—A number of the muscles of the head are fairly well developed. The individual muscle cells are in some instances highly differentiated. ‘The myotomes are fairly normal except next the undeveloped region of the spinal cord on the left side. Skin—In several places there are villus-like outgrowths of epithelium. Larva No. 2, Experiments II, Plate III External Form—The caudal extremity of the embryo bends sharply in a dorsal direction. ‘The body cavity is enormously distended, although the alimentary canal is but slightly developed (see sections cto). In these sections the ventral wallof the body cavity has collapsed, owing to the action of the fixing fluids. The head is exceedingly irregular in shape, owing to the imperfect development of the organs of special sense and the abnormal accumulation of a loose mesenchyme. The mouth opensonthe back of the head anterior to the anterior nares. Apparently there is no sucker. Along the dorsal margin of the posterior end of the body and the tail, folds of tissue project (sections ) and 1). Central Nervous System—The ventral end of the central ner- vous system consists of a thin-walled dilated sac which does not extend as far forwardas the nasal fossz. Sections a,band c show no trace of it. Sections d, e and 7 show the abnormal condition of the walls of the neural tube. “There is no fibrilar framework (Randschleier). The cells are irregularly placed; many of them show evidences of degeneration and not a few lie free in the neural canal. The pineal gland is a small vesicular pouch, the walls of which are composed of cells which exhibit degeneration. ‘The infundibulum consists of a thin-walled projection from the ven- tral portion of the neural canal. Section d shows the entrance into it. The spinal cord consists of a round thin-walled tube in which no specific differentiation has taken place. 28 Charles Russell Bardeen Organs of Special Sense—The nasal fosse are patent but irregular in form (sections b and c). No olfactory nerve can be distinguished. ‘The eyes are very rudimentary (section d). The lens consists of a small round clump of cells. “The optic stalk 1s patent and the pigment and sensory layers of the retina are separ- ated by a space connected with the lumen of the eye stalk. ‘The sensory layer consists of a thin membrane of partly degenerate cells. The cells of the pigment layer are irregular in outline. ‘The audi- tory vesicles are small round sacs from each of which a short dorsal diverticulum extends. ‘The auditory nerve is not distinguishable. Peripheral Nerves—The ganglion of the trigeminal nerve is partly developed. ‘Those of the ninth and tenth nerves are not distinct. No spinal ganglia can be distinguished with the excep- tion of one small group of cells in the mid-thoracic region. Alimentary Canal—An irregularly shaped mouth extends from the dorsal surface of the head to the pharynx. Its margis are surrounded by folds of tissue (section a). ‘The pharynx is dilated near the mouth (section 4). Posteriorly it is flattened from front to back (sections cand d). ‘The gill clefts are patent. From the septa which partially close them irregular outgrowths arise (sec- tion d). The cesophagus passes posteriorly from the right side of the pharynx (section e). At the left of the origin of the ceaonhaeee there is a mass of cells, flat in cross section, which is continued for some distance along the dorsal wall of the body cavity (section /). This represents the tracheo-pulmonary anlage. ‘he cesoph- agus passes into the stomach at the left of the liver (section e). The gut curves in front of the pancreas (section 7), whence it is con- tinued into a straight gut which about the middle of the trunk ex- hibits a distinct lumen that is continued with some interruptions to the rectum (sections / and 7) and through the anus to the exterior. In addition to the main lumen several irregular tubular spaces occur in the mass of cells which compose the gut. ‘The greatly dilated body cavity extends far forward beneath the pharynx. The liver is partially differentiated, extends forward beneath the pharynx and posterior to the rudiment of the heart has a few blood sinuses containing blood corpuscles. The pancreas 1s partially differentiated posterior to the liver and dorsal to the stomach. In Abnormal Development of Toad Ova 29 section / it may be seen between the gut and the tracheo-pulmon- ary process. Heart and Blood Vessels—The heart apparently consists of a thin-walled tube which has been ruptured. The only blood vessels which can be made out in the embryo are irregular sinuses contain- ing a small number of blood corpuscles. Genito-Urinary Organs—The pronephric tubules are rudimen- tary and are irregularly drawn out in the distended wall of the body cavity (section 7, on the right side of the section). The Wolfian ducts cannot be followed uninterruptedly to the cloaca. They lie far out in the body wall on each side. In places they are dilated, in places apparently missing. Skeleton and Connective Tissues—The anterior end of the chorda is small and defective (section f). In the spinal region it is relatively normal (sections / and 1). ‘There is an excessive amount of loose mesenchyme throughout the body, especially in the region of the head. A few of the cartilages of the head are developed, though apparently not perfectly normal in form. Musculature—In the head a few groups of partly differentiated muscle cells indicate muscles. The outlines of the trunk myo- tomes are not distinct. [The muscle cells are-more scattered than normal, owing apparently to their being forced apart by the invasion of fluids and mesenchyme. Skin—There are numerous places in which the epithelium gives rise to finger-like projections. “The most marked of these 1s the dorsal margin of the trunk and the tail. Larva No. 3, Experment 11 External Form—The posterior half of the body, including the tail, bends sharply in a dorsal direction. ‘Thisis due to a great dilation of the hind gut. The head is rounded and swollen and it is difficult to make out clearly the specific characters of the tad- pole head. ‘The sucker is rather flat. Central Nervous System—The central nervous system ends anteriorly in two distinctly separated olfactory lobes into which the lateral ventricles extend but a short distance. The walls zre abnormally differentiated and contain both cells and fibrous tissue, 30 Charles Russell Bardeen (Randschleier). There is no pineal gland but the roof of the third ventricle projects inward in such a way as to suggest an inverted pineal gland. It is different from the inversion which accompanies the choroid plexus. ‘The ventral wall of the brain 1s © thick where the optic stalks arise. The infundibulum is large and contains many desquamated cells. ‘The roof of the midbrain is thin, although there are some evidences of the rudiments of the optic lobes. ‘The medulla is relatively normal. A considerable mass of tissue, resembling yolk cells, lies free in the fourth ventricle. The proximal part of the spinal cord is fairly well developed. Posteriorly in places the spinal cord 1s irregular in form. Organs of Special Sense and Peripheral Nerves—On the right side a nasal fossa extends from the nasal pit to the pharynx. On the left side a column of cells with an imperfect lumen does not quite reach the pharynx. Near the pharynx is gives off a lateral process. ‘The optic stalk is patent on each side and the cavity extends between the pigment and sensory layers of the retina. The sensory layer is not specially differentiated. ‘The cornea is closely applied to the front of the sensory layer and lies at some distance from the ectoderm. ‘The auditory vesicles are somewhat simple sacs with dorsal diverticula. ‘The auditory nerve may be fol- lowed into the brain. The sensory ganglia are for the most part fairly well developed. ‘The peripheral nerves cannot be followed. Alimentary Tract—There is a free opening into the pharynx. This is surrounded by imperfectly developed lips and jaws. On the lips and the jaws are numerous imperfect teeth. Internal gills are fairly well developed on the left side but on the right side forma dense mass fused with the operculum. ‘The opercular cavity does not extend ventral to the pericardium. ‘The cesophagus is nearly occluded by epithelial cells. From it a solid cord of cells passes posteriorly and represents the pulmonary anlage. As the cesophagus is continued into the stomach the alimentary canal swings to the left of the liver and pancreas. ‘The gut then passes to the right across the front of the body, then anteriorly and finally curves back and passes into a portion of the gut which is much distended and extends posteriorly for some distance, then takes Abnormal Development of Toad Ova 31 several coils forward and finally passes back into the rectum. This is greatly dilated and partly coiled. The anus is small and apparently partly stopped up by mucus. The ventral wall of the body cavity is strikingly thick, not only in front of the heart but also over the entire abdomen to the anus. Heart and Blood Vessels—The heart is thin-walled but fairly well formed. There are some trabecule in the ventricle. ‘The pericardial cavity is small, owing to the great amount of mesen- chyme in the body wall. Some, at least, of the chief blood vessels are present and contain a considerable amount of blood. Genito-Urinary Organs—The tubules of the head kindey are irregular in form and are much dilated in places. The Wolfian ducts may be followed to the cloaca. Near the cloaca the anlages of the tubules of the mesonephros may be seen. Skeleton and Connective Tissue—The chorda 1s relatively nor- mal. An excessive amount of mesenchyme is present, especially in the region of the head. In the head most of the cartilages are fairly well developed. Musculature—In the head the muscles are fairly well differen- tiated. The myotomes are fairly normal although the cells are somewhat more scattered than usual. Skin—In places irregular finger-like processes of the epithelium project from the body. Larva No. 4, Experiment II, Plate IV—A External Form—The tail isa mere rudiment. ‘The ventral part of the body is much swollen, owing to distention of the body cavity. The head is small and deformed. ‘The sucker is rather flat (sec- tion a). Central Nervous System—The neural tube ends anteriorly in two thick walled olfactory lobes. The right one is shown on the left side of section a. ‘The cells in the walls are irregularly placed and many exhibit signs of degeneration. “The Randschleier is not normally disposed. ‘There is a small pineal body with a narrow lumen and thick walls composed of cells not specifically differen- tiated. ‘The optic lobes are partially differentiated. The infun- dibulum has a thin wall but contains in its cavity a large mass 32 Charles Russell Bardeen of cells. The lateral walls of the midbrain are excessively thick and the lumen of the aqueduct is narrow. The ventral wall of the medulla is irregular in outline in places (section }). There are many desquamated cells in the fourth ventricle. The spinal cord is greatly deformed in most regions and in many cases shows no central canal (sections d and ¢). In the midthoracic region it is fairly normal in form. ‘The spinal cord does not extend into the rudimentary tail. Organs of Special Sense—The nasal fosse are patent. The medial wall of each fossa is thin (section a). The optic stalks are patent. ‘The pigment layer of the retina is separated from the sensory layer. The latter is not specifically differentiated. A lens rests against the sensory layer (section a). The auditory vesicles are simple sacs with dorsal diverticule. Peripheral Nerves—The sensory ganglia of the fifth, seventh and eighth, ninth and tenth nerves can be distinguished. The nerves can be followed but a short distance. Spinal ganglia are present in the midthoracic region where the spinal cord is fairly normal in form, but are not present elsewhere. Alimentary Canal—The mouth is patent. The jaws and lips are partially differentiated. ‘The gill slits are patent. ‘The in- ternal gills are rudimentary in form and contain no obvious blood vessels. ‘The operculum is only partially differentiated. There is no opercular cavity in front of the heart. The tracheo-pulmon- ary anlage is a short branched column of cells. The cesophagus is composed of dense tissue in which several irregular spaces sug- gest a coiled tube (section c). The stomach curves about the left side of the anlage of the liver (see right side of section d). The liver is highly differentiated (section c) and is connected by ventral and dorsal mesenteries to the walls of the body cavity. The large blood spaces in the liver contain traces of blood. The bile duct is large. ‘The pancreas is not specifically differentiated but its anlage is marked by a mass of yolk cells. The gut curves from left to right in front of the anlage of the pancreas and then turns distally. Beyond the region of the pancreas the gut passes nearly straight back to the anus. ‘The ventral portion is thick and filled with yolk cells as in young larve. The anus opens on Abnormal Development of Toad Ova 33 the right side of the body between the rudimentary tail and the body cavity (section f). The body cavity is greatly distended (sections d, e, f). Heart and Blood Vessels—The heart is a simple S-shaped tube. The ductus arteriosus is patent; the ventricle has very thick walls, a small lumen and no trabeculae. “The sinus venosus and the auricle are thin-walled. There are a few blood corpuscles in the lumen of the heart. Mere traces of blood vessels are visible. Genito-Urinary Organs—The tubules of the pronephros are much dilated. They lie in large spaces which here and there con- tain a few blood corpuscles (section d). “The Wolfian ducts are much distended (section e). Skeleton and Connective Tissues—The chorda is relatively nor- mal in structure. There is an excessive amount of connective tissue, especially in the region of the head. The cartilages of the head are partly differentiated. There are no auditory capsules. Musculature—The muscles of the head are partially differen- tiated. ‘The myotomes are relatively fairly. well developed (sec- tions d and e). Larva No. 5, Experiment IT External Form—The tail curves dorsalward. The abdomen is distended. The head is somewhat irregular in form. ‘The sucker consists of a short projection on the right side of the body. Central Nervous S ystem—Anteriorly there are two small olfac- torylobes. The lateral ventricles extend but a short distance into each. The pineal body is round and has a slender stalk. The cells composing the body are more or less scattered, although the outer wall of the body is fairly smooth. ‘The lateral walls of the third ventricle are thick and show a fairly normal differentiation into cells and Randschleier. “The hypophysis has a very thin wall. The optic lobes are not differentiated. The fourth ventricle is much distended. ‘The walls of the medulla contain many degen- erated pigment cells. Organs of Special Sense—The tissue of the medial wall of each nasal fossa contains many degenerated pigment cells. ‘The pig- ment layer of the retina is separated from the sensory layer. 34 Charles Russell Bardeen Neither optic stalk shows a lumen. ‘The sensory layer of the retina shows much degeneration. ‘Thelens rests againstit. The auditory vesicles are simple. Peripheral Nerves—The sensory ganglia of the head and the nerve branches may be followed better than in most of the experi- ment larvae. ‘The spinal ganglia and nerves are less definite. Alimentary Canal—The mouth is patent and is surrounded by partially differentiated jaws and lips. ‘The internal gills are im- perfectly developed. ‘The operculum extends but a short distance posteriorly. ‘The cesophagus is occluded with cells. “The tracheo- pulmonary process is short and branched. The cesophagus continues to the left into the stomach. ‘The gut curves ventrally to the right in front of the pancreas, then anteriorly on the right of the liver and then bends back in a posterior direction. ‘The liver is well developed and contains large sinusoidal spaces. A very few blood corpuscles appear to be contained in these spaces. There is a large gall bladder. The pancreas is also well developed. The gut, immediately posterior to the pancreas, is greatly distended and beyond the region of distention exhibits several partial coils. The anus is patent. In the body cavity many multi-nucleated cells can be seen. ‘The walls of the body cavity are excessively thick. Heart and Blood Vessels—The heart is S-shaped. The ductus arteriosus is solid. “There is a small lumen in the ventricle. The sinus venosus 1s thin-walled. In places blood vessels containing a small amount of blood may be seen, but the vascular system is imperfectly developed. Genito-Urinary Organs—The tubules «of the pronephros are greatly distended and lie in large spaces in which some blood cor- puscles may be seen. The Wolfian ducts are greatly distended near the pronephros but not much more distally. ‘That on the left side is much smaller than that on the right. Skeleton and Connective Tissues—The chorda dorsalis is moder- ately normal in structure. In the sections it is shrunken. ‘There is an excessive amount of connective tissue, especially in the region of the head and in the wall of the body cavity. The carti- lages of the head are fairly well developed. Abnormal Development of Toad Ova 35 Musculature—The muscles of the head are well differentiated. In places the spinal myotomes are fairly normal; in places the cells composing them are much scattered. Skin—The epithelium shows irregular projecting processes in many places. Larva No. 6, Experiment II, Plate V External Form—The long axis of this embryo is nearly straight. The head appears much crumpled. ‘The mouth is a large cavity bounded by irregular folds of tissue. ‘There is a small semi- circular sucker back of the oral opening. Central Nervous System—The brain is anteriorly much dilated (section a) and extends to the anterior extremity of the head. The neural wall is very thin and the central cavity is filled with degenerated cells. ‘The dorsal wall of the forebrain in places is fused to the ectoderm (section 5). ‘There are no visible traces of a pineal body. The ectoderm is irregularly thickened where it comes in contact with the brain and in some places is very thick. In one place a long column of ectoderm cells extends in between the ectoderm and the brain (right side of section b). ‘The dorsal wall of the brain on each side in the region of the eyes is greatly thickened. The midbrain is most irregular in form and gives rise to several vesicular processes of uncertain nature. ‘The walls of the midbrain are thin and the central canal is large. “The medulla is most abnormal in form (sections c and d). The dorsal part of the fourth ventricle is curiously dilated. ‘The spinal cord is more normal in form than the brain although it also is much deformed in places. Organs of Special Sense—The only trace of nasal epithelium is a collection of cells between the dilated pharynx and the ecto- derm in the anterior part of the head shown on the right side of section a. The eyes are very abnormal. ‘The optic stalk 1s dilated soas to make a direct opening from the neural canal to the back of the sensory layer of the retina. ‘There is no trace of specific differentiation in this layer (section b). Where the optic stalk approaches nearest to the ectoderm a lens has been differen- tiated. This is still in contact with the ectoderm and consists of 36 Charles Russell Bardeen a group of little differentiated ectoderm cells. The auditory vesicles are simple in form, thick-walled, and resemble those newly formed in the embryo. Peripheral Nerves—Mere traces of the sensory ganglia of the cranial and spinal nerves can be seen. Nerve fibers are not well marked. Alimentary Canal—The mouth is a large opening into the pharynx. ‘The opening is irregular in outline and its boundary presents mere traces of a beak and lips. The pharynx is dilated anteriorly (section a), but over the cardiac region is flat and bent to conform to the dilated pericardial cavity. The gill clefts nowhere open to the exterior but instead are laterally stopped up by masses of epithelial tissue (sectionc). No gills are specifically differ- entiated. The cesophagus is filled with epithelial cells. Thetracheo- pulmonary process isrudimentary. The cesophagus passes directly back into a primitive intestine. Neither liver nor pancreas seem specifically differentiated, although the anlages of each are repre- sented by masses of yolk cells (section ¢). The gut consists of a mass of cells which extends in a fairly straight direction from the cesophagus to the anus. Anteriorly it shows some tendency to form convolutions, and in the mid-body region it curves slightly toward the right side of the body. For the greater part of its course no lumen is present. FHeart and Blood Vessels—The pericardial cavity is very thick- walled anteriorly but is relatively large. More distally the ven- tral wall is thin (section c). Traces of a heart can be seen (sec- tions c and d). ‘This forms a slightly S-shaped structure with thin-walled ductus arteriosus and sinus venosus (section d) and a more solid ventricle (section c). “There is no blood in the heart cavity. No definite blood vessels can be made out in the embryo. Genito-Urinary Organs—The tubules of the pronephros are much dilated. “The Wolfian ducts can be traced only part of the way to the anlage of the cloaca. In the distal part of their course they are curious flat tubes (section 7, on each side of body). Skeleton and Connective Tissues—The chorda dorsalis is slightly asymmetrical in places. ‘The connective tissue is excessive, espe- cially in the head. ‘The cartilages of the head are not distinct. Abnormal Development of Toad Ova 37 Musculature—Mere traces of muscles are found in the region of the head. ‘The more anterior myotomes consist of rounded masses of cells on each side of the chorda but not in contact with this (section ¢). In the center of the trunk and in the tail they are somewhat more normal in form. Skin—Projections of epithelium may be seen on the dorsal margin of the tail (section g) and at the side of the ventral wall of the body cavity (section ¢). There are curious subcutaneous vesicles on each side of the head posterior to the sucker. Section } shows one of these at the left of the section. Larva No. 7, Experiment II, Plate IV—B External Form—The tail is short and stubby and curves dorsally. The heart is shrunken and irregular in outline. The sucker is apparently fairly normal. Central Nervous System—Two well separated olfactory lobes project forward as far as the nasal fossz and each is in contact with the medial wall of the corresponding nasal fossa (section a). The tissues of the olfactory lobes are partially degenerated. ‘The lateral ventricles are very small. The lateral walls of the third ventricle are thick and partially differentiated (section b). ‘The pineal gland projects above the third ventricle and its tissue 1s partly degenerated. ‘The infundibulum 1s small and thin-walled. The lateral walls of the mid-brain are thick and project inward so as to nearly obliterate the aqueduct. The hind-brain is rela- tively normal, although flattened from front to back (sections ¢c and d). ‘The spinal cord in places is fairly normal, in places the ' cells from the walls of the neural tube fill or nearly fill the central canal. Organs of Special Sense—TVhe nasal organs are fairly well dif- ferentiated. hey lie, relative to the brain, posterior to the nor- mal position. ‘The pigment layer of the retina is separated by a space from the sensory layer. ‘The latter is not well differentiated. The optic stalk is not patent but contains no nerve fibers. A lens is present. [he auditory vesicles are simple in form. Peripheral Nerves—Cranial and spinal ganglia are moderately well developed. ‘The nerve fibers cannot be readily traced. 38 Charles Russell Bardeen Alimentary Canal—The mouth is open. Lips and jaws are rudimentary. The pharynx is dilated and contains a sac-like protrusion through its floor from the pericardial cavity (section c). The gills are rudimentary and are contained within a cavity on each side which is formed by an opercular fold open behind (sec- tiond). “The cesophagus hasa lumen into which irregular vesicular spaces open. ‘The tracheo-pulmonary process is short. ‘The gut passes to the left of the liver and pancreas, then curves across the body in front and finally extends straight back. ‘The lumen is not distinct. Neither liver nor pancreas is well differentiated The body cavity is greatly dilated. Heart and Blood Vessels—The heart is an S-shaped tube. The ductus arteriosus and ventricle are thick-walled; auricle and sinus venosus are thin-walled. ‘There is a slight amount of blood in the lumen. The pericardial cavity projects into the mouth, pushing the floor of the pharynx ahead so that the heart comes to lie literally in the mouth. ‘The only definite blood vessels are two vessels which appear to be posterior cardinal veins. ‘These are dilated and anastomose with one another in several places. Genito-Urinary Organs—The tubules of the pronephros are greatly dilated. The Wolfian ducts are irregular in form. That on the right side appears to be missing in places. Skeleton and Connective T1ssues—The chorda dorsalis is appar- ently normal. ‘There is an excessive amount of connective tissue. Some of the cartilages of the head are fairly well differentiated. Musculature—Some of the muscles of the head are fairly dis- tinct. The myotomes are moderately normal, although in places the muscle cells are somewhat scattered. Skin—In the region of the head there are especially large masses of projecting epithelium (section a). Larva No. 8, Experiment II External Form—The long axis 1s fairly straight. ‘The general appearance is that of a normal embryo soon after the tail has grown out. The tail, however, is somewhat shrunken. ‘The sucker is fairly normal. Abnormal Development of Toad Ova 39 Central Nervous System—The olfactory lobes extend forward between the nasal fossa. ‘The tissue of the olfactory lobes is somewhat degenerated. ‘The pineal gland consists of a rounded hollow vesicle with a short much dilated stalk. The ventral wall of the third ventricle is abnormally thick. ‘The infundibulum is relatively normal. ‘The walls of the midbrain are also fairly nor- mal but contain some degenerated pigmented cells.. The optic lobes are beginning to be differentiated. “The hindbrain and spinal cord are relatively normal. Organs of Special Sense—Vhe medial walls of the nasal fossz are thick but contain many pigmented and degenerated cells. The eyes are very abnormal. ‘The optic stalks are greatly dis- tended. ‘The pigment layer of the retina is irregular. The sen- sory layer consists of a mass of degenerated cells. ‘The lens is differentiated. [he auditory vesicles are simple in form. Peripheral Nerves—The sensory ganglia are moderately well developed but the cells are many of them abnormal. Nerve fibers cannot be readily followed. Alimentary Canal—The mouth is open. Lips and jaws are clearly marked, although not highly developed. Gill slits are patent, but the gills are not well developed and seem to contain no blood vessels. Opercular folds extend over the anterior portion of the gill region on each side. The cesophagus has a lumen. The tracheo-pulmonary process is short and branched. The stomach lies at the left of the liver anlage. The gut extends straight back. Neither liver nor pancreas is well developed, although masses of cells indicate their anlages. No blood spaces are found in the liver. Heart and Blood Vessels—The heart consists of an S-shaped tube, but the walls are not normally differentiated. ‘There is some blood in the ductus arteriosus. ‘There are apparently a few blood vessels present but there is no well developed vascular system Genito-Urinary Organs—The tubules of the pronephros are slightly dilated. Skeleton and Connective T1issues—The chorda dorsalis is rela- tively normal. ‘The cartilages of the head are fairly well developed. 40 Charles Russell Bardeen There is a slight increase over the normal amount of connective tissue in the body. The pigment cells are abnormally scattered about. Musculature—Vhe muscles are fairly well differentiated in the head. ‘The more anterior of the myotomes are fairly normal but in the posterior half of the embryo they are not well developed. Skin—In many places the skin shows abnormal outgrowths especially about an irregular opening into the body cavity. SUMMARY AND CONCLUSIONS Toad spermatozoa removed from the body begin to lose both motility and fertility within half an hour. Both motility and fertility last much longer on cool than on warm days and some- what longer in unexposed than on spermatozoa exposed to: the Roentgen rays. On cool days the power of fertilizing lasts in some of the spermatozoa over two hours. When only ten or fifteen per cent of the control eggs are fertilized as a rule few or none of the eggs placed with the exposed sperm are fertilized. When only a few eggs are fertilized by the exposed sperm as a rule these eggs do not develop beyond the gastrula stage, but occasionally one may develop into an abnormal tadpole. When the spermatozoa have been well exposed to the rays and yet are still capable of fertilizing a considerable number of eggs, the eggs thus fertilized develop at first apparently normally or even better than the control, but beyond the gastrula stage the development begins to become retarded and at the time of hatch- ing, as the tail begins to grow out, marked deformities appear in the larvae. ‘These deformities are visible externally and are still more striking when the internal structure is examined. The illustrations given on Plates II to V illustrate these deformities more readily than they can be described in words. While they vary considerably there are certain features characteristic of most of the tadpoles. General Development Growth of the tadpole is inhibited beyond the stage which in- tervenes between hatching and the time when it should begin to Abnormal Development of Toad Ova 41 swim. Thus when the control tadpoles of the same age as the experiment tadpoles are equivalent in general form to the 12 mm. tadpole of the frog described at some length in Marshall’s well- known text-book, the development of the control tadpole 1s, as a rule, more nearly similar to the newly-hatched tadpole of the frog described by Marshall. External Form The head is usually abnormal in shape, the anterior end appear- ing shrunken. ‘The coelom isin many of the tadpoles abnormally distended. ‘The tail is usually short, more or less deformed and is often bent in a dorsal direction. Internal Structure The vascular system is little developed in any of the experiment embryos. The heart usually is S-shaped but 1s rudimentary in form and may have no continuous lumen. In some embryos the wall of the ventricle is thickened by muscle cells but in none are there strong trabeculz in the ventricle. The chief arteries seem in none of the embryos to be completely developed, although in some there are here and there traces of them. The chief veins are likewise in none of the embryos completely developed although in one embryo the cardinal veins are large. In the liver the capil- laries are sometimes well, sometimes but slightly developed. There are relatively a very few blood corpuscles in any of the em- bryos. These lie in some of the scattered vascular anlages. It is uncertain whether the blood had circulated in any of the embryos, but in some of them it is fairly certain that no circulation was established. In all of the embryos the spaces in the tissues indi- cate a considerable amount of lymph either free in the tissues or confined in lymph vessels. Of the central nervous system the brain is the part most con- stantly and deeply affected, but the spinal cord in many of the em- bryos is markedly deformed. The abnormalities consist partly of failure of development or tissue differentiation, partly of irregular growth of tissue, pigmentary degeneration of nerve cells and the 42 Charles Russell Bardeen filling of the central canal with partially degenerated cells. In one embryo the hind brain and anterior part of the spinal cord are exceedingly rudimentary on one side. Of the organs of special sense the eye exhibits the greatest deformities. [he nose and ear are as a rule rather rudimentary than markedly deformed. ‘The eye, however, usually shows a patent optic stalk connecting with a space between the pigment and sensory layers of the retina, a lack of differentiation in the sensory layer, and a more or less highly differentiated lens resting against the sensory layer. The abnormalities in the alimentary canal are exceedingly variable and may affect any or all parts. “The mouth 1s in all instances patent, the lips and jaws rudimentary. The pharynx and gills vary much in structure in the different embryos. As a rule there are traces of internal gills and of the opercular folds but the gills, owing to lack of development of the vascular system, are rudi- mentary. - The cesophagus is patent in some, closed in other of the embryos; the stomach as a rule lies at the left of the anlages of the liver and pancreas. The latter structures are seldom - highly developed. ‘The rudiments of the lungs are slightly devel- oped. The intestines may be more or less coiled, but are in none of the embryos highly developed and in some are very rudimen- tary. In many embryos the abdominal cavity is greatly distended while the gut 1s rudimentary. The pronephric tubules are usually greatly swollen in places and this dilatation is also frequently found in the Wolfian ducts. There are seldom distinct traces of the metanephric tubules. The myotomes, when not well developed, usually consist of muscle cells somewhat scattered about in the surrounding mesen- chyme. The muscles of the head are usually more or less dif- ferentiated. The mesenchyme of the embryos is considerably greater in amount than in normal tadpoles. The cells seem to be spread apart by fluids in the tissues. ‘The cartilages of the head and the chorda dorsalis are relatively normal. The ectoderm in most of the tadpoles shows in places outgrowths of an irregular nature. ‘These may be extensive villus-like pro- Abnormal Development of Toad Ova oe cesses. In one instance marked ingrowth of processes from the ectoderm occurred. The cells of the tissues appear for the most part clear in outline. Many of the cells of the central nervous system seem to have under- gone a pigmentary degeneration. Numerous cells in most of the tissues show mitotic figures. I have been unable satisfactorily to determine whether or not there are abnormalities in these figures. ‘There is an abnormal number of cells with two or more nuclei. A striking feature of the experiment embryos is the irregular dis- tribution of the pigment cells. “They are much more irregularly distributed through the tissues than in the normal embryos. There is a striking resemblance between tadpoles which de- velop from ova fertilized by sperm exposed to the Roentgen rays and the tadpoles exposed directly to radium irradiation by Schaper™ This shows clearly that injuries produced in nuclei may be carried through many generations of cells in an individual and finally give rise to deformities corresponding with those due to direct irradiation. Bohn* found that the rays of radium rapidly enfeeble or kill the sperm of strongylocentrotus lividus, but that the eggs appear more fertile after exposure. He doesnot describe the effect of exposure of the germ cells on subsequent development. Herbst® found by treating the sperm of sea-urchins with fresh water, alkalies and potassium-free salt water and then fertilizing ova of a different species with the sperm thus injured that the ova sometimes developed as if injured but that there was no evidence that the specific hereditary factors transmitted by the spermatozoa were altered. [Further studies are necessary to determine if the hereditary factors carried by the sperm may be specifically influ- enced by irradiation. D. T. Macdougal* has shown that in some plants mutations may be produced by injecting radium prepara- tions, sugar solutions and solutions of calcium nitrate and of zinc sulphate into the ovaries. This most important work sug- 8\Schaper: Anat. Anzeiger, xxv, p. 298, 1904. Levy: Archiv f. Entwicklungsmechanik, xxi, p. 130, 1906. 8G. Bohn; Comptes Rendus de I’Acad. des Sciences, Paris, cxxxv, p. 1012, 1085, 1903. 88Herbst: Archiv fiir Entwicklunsmechanik der Organismen, xxi, P- 293, 1906. ®‘MacDougal; The Popular Science Monthly, September, 1906, p. 16. 44 Charles Russell Bardeen gests that the hereditary factors contained in spermatozoa might be so altered as to produce specific variation in the individuals springing from ova which they fertilize. The effects of altering the normal course of development of vertebrates by electrical, magnetic, chemical or mechanical agents applied to the whole organism, have been shown by Dareste,® Féré,* Roux,’’ and many others, to be seldom confined in a specific way to an organ or group of organs, although some organs, like those composing the nervous system, are especially sensitive to all such factors. ‘he experiments with irradiation show that although some tissues are much more susceptible to the rays than others, there are wide differences in the effects of the rays on different individuals. In conclusion, I desire to express my thanks to my colleague, Professor B. W. Snow, for the use of the Roentgen ray apparatus belonging to the Department of Physics, and for his aid in con- ducting the exposures. EXPLANATION OF PLATES I to V. On these plates there are represented outlines of the external forms and transverse sections through the body of one control and five experiment larve. Descriptions of the larve represented are to be found in the text as follows: Plate I, p. 21 Plate IV—A, p. 31 Plate II, p. 24 Plate DV—B, p. 37 Plate ITI, p. 27 Plate V, p- 35 85Dareste: Recherches sur la production artificielle des monstruosités. Paris 1891. “SFéré: Comptes Rendus de la Société de Biologie, 1893-1905 S7Roux: Gesammelte Abhandlungen tiber Entwicklungsmechanik der Organismen, 1895. ABNORMAL DEVELOPMENT OF TOAD OVA PLATE; Cuartes Russert BARDEEN Tue Journat or ExperiMENTAL ZOOLOGY, VOL, 1V, No. I pe. “ < ’ a Le a Vie Swed al Bus 1eaS ste te ee Se - 7 , | ABNORMAL DEVELOPMENT OF TOAD OVA PLATE II CuHartes Russet, Bardeen Tue JourNat or ExperIMENTAL ZOOLOGY, VOL. IV, No. I ABNORMAL DEVELOPMENT OF TOAD OVA PLATE III CuHarces Russert BarDEEN THE JourNaAt or ExperRIMENTAL ZOOLOGY. VOL. 1v, NO. I ABNORMAL DEVELOPMENT OF TOAD OVA PLATE IV Cuartes Russet. BarpEEN Tue Journat or ExPERIMENTAL ZOOLOGY, VOL. IV, NO. I ABNORMAL DEVELOPMENT OF TOAD OVA PLATE V Cuartes Russert BARDEEN Tue JourNAL or ExPpERIMENTAL ZOOLOGY, VOL. IV, NO. I ? - PANO UTES Wea x ; ee aan a sg AN ECOLOGICAL AND EXPERIMENTAL STUDY OF DARCOPRMAGIDA WITH RELATION TO; EAKE BEACH DEBRIS? BY WILLIAM B. HERMS Wirtu SEvEN Figures CONTENTS it Sitageehienteel Spooenne yanosoehpo aban cguebeemaLosmodcO sens HtancpocbSauonpodaod cutee 45 II Habits and life histories of Lucilia cesar, the common green flesh fly; Compsomyia macellaria, the screw-worm fly; Sarcophaga sarracenie, a common large gray flesh fly; and Sarcophaga aasiduasraccmiallmoray tleshifly =Ss)at.z Re eT AE MN oka A AST AAS ORC fe) Bassiand perch 25- sade sche soil ag ee arene Otee Renita alae eer 15 Maan eye 25:0). Ke oe eerie Suse cean ene kee Ee eee ere 2 Cait SRR US RE of oto eres te ea sO te rp ROL Ons MOE erie ote 3 Org piney 8 ara Pa eee tn ten Pye APS Ave Pete aay oD SE I Aa 1 a ere EINER RET aN ree Fai WA pment ime Ne Ac 45 Total weight, 4.65 kilograms. Ecological and Experimental Study of Sarcophagide 47 This gives an average of 103.3 grams. The result on the whole was disappointing, fewer fish having been cast up than during similar surfs. A second weighing trip over the same ground was made about two weeks later, conditions being similar. ‘The results of this trip are as follows: Perehtane simalu-mouth Dass, “..c.05.5''s, SA fe. ae ae Rey Ae 441 SLUGS) St ee ee ee niin AMEE MARL. Ur iterei SRO ale 18 PVMNSIIOLEY Sr ewtets ca ti Yod'cs een ctice SG. eseat-cie aid Gee eee ten eee eee 50 [DSSG! Sao nn Be neha NRE REE eee SPER eee cite eae tele aN 12 Oba Sree che sb dlghe Be ats Wind wae aoe Set cose at eke 10 IAT Sicie aT OTS Ay ais jade NR mn ase. yobs tre ce eee ence ene een 7 TRG Res ies ee ae eae ee Mn Mere Pane Neem ety 8 al disk eid, dues 538 Total weight, 20.38 kilograms. The average 37.9 grams is low, but this is due to the fact that practically all of the perch and small-mouth bass were young, ranging in weight from 14 to 42 grams. ‘The lake storms appear at quite regular intervals during the summer, which allows a period between each high surf for a complete cleaning up by the scavenger insects. For this reason it is possible to get only those fish on record which are newly washed up. It would make an interesting and no doubt profitable investi- gation to ascertain the cause of death to the large number of young perch and bass which are cast up at times during the high surfs. With a beach so well scattered with dead fish as to show 538 to a mile, and every reason to believe that this was notan exceptional stretch of beach on this particular morning, one has some con- ception of the work to be done by Nature’s scavengers. A walk along this same mile of beach about three days after a storm can but impress one with the effective work of the scavenger insects. An examination will show that the many fish carcasses have been reduced to mere shells, which are comparatively odorless. Below is a list of the scavengers most active in the removal of organic beach débris. ‘This list is not intended to be a complete one, there are other species of both Diptera and Coleoptera and also other insects active in the work though in a very small degree. 48 William B. Herms Diptera Coleoptera Fam. Sarcophagidz Fam. Staphylinidz Lucilia caesar Linné Creophilus villosus Grav. Compsomyia macellaria Fabr. Fam. Scarabzidz Sarcophaga sarraceniz Riley ‘Trox scabrosus Beauv. Sarcophaga assidua Walker Fam. Silphida Silpha americana Linné Necrophorus orbicollis Say. Necrophorus vespilloides Hbst. Necrophorus tomentosus Web. Fam. Dermestide Dermestes vulpinus Fab. Dermestes caninus Germ. Fam. Histeridze Saprinus pennsylvanicus Payk. Saprinus lugens Er. Fam. Calandridzx?® Sphenophorus ochreus Lec. Phytonomus punctatus Of the above, the flies are the chief agency, and it is a brief study of these that we shall enter upon. II HABITS AND LIFE HISTORIES OF LUCILIA CAESAR, THE COMMON GREEN FLESH FLY; COMPSOMYIA MACELLARIA, THE SCREW- WORM FLY; SARCOPHAGA SARRACENIAE, THE COMMON LARGE GRAY FLESH FLY; AND SARCOPHAGA ASSIDUA, A SMALL GRAY FLESH FLY Habits of Adults On emergence from the pupal cases, after the wings are sufh- ciently dry, the first impulse is to seek food. It seems to be the object of each individual to seek its own nourishment, since each fly takes flight alone. ‘These insects are never seen flying about in ageregation in quest of food. Sarcophaga sarracenie is rarely found in large numbers about a carcass, while Lucilia caesar and Compsomyia may be very numerous. "Diptera kindly identified by Prof. J. S, Hine, of Ohio State University, 3Accidental. Ecological and Experimental Study of Sarcophagide 49 The presence of food is detected in a remarkably short time, and this can only be accounted for on the assumption of a very acute sense of smell. Comparatively fresh fish were exposed where no flies were to be seen, and in ten or fifteen minutes many flies were hovering about the food and some eggs had already been deposited. That the compound eyes so prominent in the Sarcophagidz are of importance in orientation we are reasonably certain. If these insects were deprived of their eyesight, food would probably be found with difficulty. In several cases the eyes of Sarcophaga sarraceniz were painted with india ink, affecting the flies in a manner similar to that of animals whose semi-circular canals are disturbed. Orientation was almost completely lost for a time. On placing the individuals on their backs, they were barely able to right themselves after frantically using both legs and wings. They crawled about on the table in an aimless manner, or on the writer’s fingers. After a few minutes they flew slowly away, buzzing noisily, passing over several pieces of fish placed on a table. Their flight was directly toward a window, which they struck with a thud. From this it would seem that the light was not perfectly excluded. No doubt, much of the disturbance above mentioned was due to the penetration of the india ink. It appears that the adults prefer the fresher food; fresh fish or fish newly cast up being attacked more readily than those having been allowed to dry. ‘This is readily explained because of the greater abundance of liquid food on the bodies of the fresher specimens, and also because of the more pronounced (‘“‘fishy’’) smellof such specimens. On several occasions a fish that had been allowed to dry for a day or two was laid outside, and each time no eggs nor larva were deposited thereon, and the fish dried up in the sun. Under natural conditions this would probably not occur. . Sarcophaga sarraceniz is rarely found in large numbers about a carcass, while the screw-worm fly is most abundant nearer the water and on larger carcasses. Lucilia casar is found in the majority on large or smaller carcasses farther away from the water and on the small ones near the water. 50 William B. Herms Egg Deposition Lucilia caesar deposits eggs in irregular masses on the softer portions of the fish, e. g., around the eyes, around the anus, be- tween the gills, on an abrasion, or on the underside of the carcass. This is due to the presence of much liquid food at these particu- lar portions, which the adults suck up while depositing eggs. Whether the deposition of eggs is associated with the stimulation of the food within the alimentary canal of the female is discussed in Chapter VI. ‘The deposition of eggs on these softer portions is, however, an adaptation favorable tothe larve, since they can thus immediately gain easy access to the body cavity. The gill slits offer good receptacles for eggs, and into these they are pushed by means of the protruded abdominal segments, the open mouth and opercles of the fish affording a good entrance. Eggs are also commonly deposited on the upper side of the carcass while the fly excitedly flits about sucking the juice. In this way great masses accumulate on the large fish, sometimes one mass growing to the size of a walnut. Several times the mouth and gills were masked with a cloth to watch the effect. This asain in eggs being laid on the cloth and even on the loose ends of the string with which the mask was tied. One female was observed to remain in the same position about six minutes, with abdominal segments pushed into a fold of the cloth mask just mentioned. At the end of this time between ninety and a hundred eggs were laid. From among a number of flies taken on a dead fish, a large female (72 mg. in weight) was dis- sected to ascertain the number of eggs contained within the ovaries. ‘Two hundred and forty-seven fully developed eggs were taken out and the dissected ovaries showed no trace of immature ova. ‘The great weight and distended condition of the female would indi- cate that few if any eggs had been deposited. How the above number compares with the normal number of ova produced by a single female of Lucilia czsar, the writer cannot say. This matter should have further investigation. Sarcophaga sarraceniz deposits living young and deposits them anywhere on the carcass or even near it, compelling the young Ecological and Experimental Study of Sarcophagide 51 larve to find a suitable place of entrance. Eighty-two living larvze were taken from one female. Compsomyia macellaria, the screw-worm fly, deposits very minute living young, but is careless about placing them on the fish. This habit of bringing forth living young seems to be exceptional in this region for the species. From thousands of eggs promiscu- ously collected, not a single screw-worm fly was reared; all were Lucilia czsar, and all observations in the field resulted in seeing living young extruded. Prof. J. S. Hine, of Ohio State University, reports that he has seen Compsomyia deposit eggs and that he has also reared them from eggs at Cedar Point. Larval Habits The young larvae, when hatched or extruded, at once eat into the softer parts, attacking the viscera and later consuming the muscular portions. ‘The fish is eaten clean to skin and bone, the skin remaining as a mere shell; this, too, would be eaten to the scales were the entire surface sufficiently moist. ‘This is evident because the portion of skin nearest the earth, where it is moist, is invariably eaten away, leaving a hole in the under side, which incidentally allows a concealed means of escape during migration. Migration depends wholly on the food supply. If the fish 1s large enough and the number of larvz is not too great, migration takes place in from two and a half to three days, during which time the larve have reached their full growth. If the number of larvz is large in proportion to the fish, migration takes place earlier. This phase of the subject is treated below under the head of “Over and Underfeeding.”’ On leaving the remains, the larve immediately burrow into the sand below or close by the fish. The great majority burrow just beneath, going down two to six inches into the sand and remaining there temporarily. ‘This migration may take place any time during the day or night, though the tactics vary for these periods. Bur- rowing temporarily just beneath the fish carcass during the day not only affords protection from the intense heat of the sun but also from birds. On cloudy days when migration sometimes takes place away from the fish, the sandpipers, in numbers, feed on the 52 William B. Herms plump migrating larva. Since these birds are quite numerous along the beach one can readily see that this would be a potent factor in extermination, and some means of protection is very advantageous. During the night, or when the sand is cooled, migration from beneath the remains takes place, and it is then that the larve travel a greater distance—fifteen, twenty feet and over, and then again burrow. Larve that were kept indoors in boxes were observed to repeat this performance several nights in suc- cession, each time burrowing for the day. The sand in the laboratory was not heated by the sun, yet the larve followed their normal habit and were characteristically active by night. Ants are a minor source of destruction to the larve that migrate from fish which have been dragged a distance away from the water. Live larve, wriggling frantically, are carried away by these little marauders. At noon of August 10, 1905, several dozen larve were found lying dead upon the sand, within a radius of about five feet from a fish. ‘The sand was extremely hot, about 140° just below the surface, and the larve had been literally baked. What induced this attempt to migrate from the fish at this time is a question. Pupation The interval between migration and pupation varies. With individuals reared indoors this interval varies with the degree of moisture—extreme moisture retarding pupation, as also does extreme dryness. A certain amount of moisture is necessary for pupation; therefore, a small amount of water was added to the sand in which the indoor individuals were kept. “Temperature probably also affects this stage. All observations were made during the summer months, consequently what happens in spring and autumn later than the middle of September is still open for investigation. ‘[he deep loose sand of the beach affords a unt- form condition for the burrowing larva and for pupation. ‘This condition, together with the food supply, also comparatively unt- form, would naturally cause less variation in the life histories of species here than in parts remote from the beach. Ecological and Experimental Study of Sarcophagide 53 The pupal period is quite regular as far as observed; the dura- tion of this stage varies, however, with the species. The emergence of the imagines from their pupal cases is inter- esting. With the great blister-like frontal sac, not unlike a tiny balloon attached to their heads, the case is burst and gradually the body is withdrawn, much as a person might extricate himself from a closely fitting tube. All the while the send particles are Fig. 1 Cut showing the manner in which screw-worm flies cling to the beach grasses (Ammophila and Panicum) after emerging from the pupa cases. Notice that the heads are mostly directed downward. thrown aside by the rhythmically inflated sac. Slowly, pull after pull, the imago passes upward through the sand, and emerges at the surface. After a moment of rest, it starts for the nearest grass stem; up this it crawls in apparent haste, and there it remains to unfold its wings. The accompanying plate shows the manner in which the screw-worm fly clings to the stems and blades of the beach grasses 54 William B. Herms (Ammophila and Panicum), while spreading and drying its wings. When a high surf washes up a big carp into the tall grass, the larve after feeding on the carcass, migrate nearby and pupate. When the imagines emerge the grasses are immediately resorted to. This accounts for the presence of veritable swarms of flies in a restricted area with no carcass near. Closer investigation will almost invariably result in finding the bones and scales of some large fish in close proximity. Most imagines reared in the laboratory emerged early in the morning at the first break of day. In one instance a dozen pupz were kept in one vial and out of these, seven adults emerged within two minutes. Life Histories Lucilia caesar Linné (the common green flesh fly): Eggs of this species are cylindrical, rounded at both ends and slightly curved, smooth and white. The average weight of one egg is about 0.1 mg.‘ Young larve hatch in from eight to eighteen hours, depending on the time of day the eggs are deposited. If deposited toward noon the time will be lengthened, since observations show that in such cases larve do not emerge until the following morning together with larve coming from eggs deposited any time during the afternoon. ‘This period also undoubtedly varies with the age of the ova at extrusion; 7. e., depends on the length of time the ova are retained within the female. The actual feeding period of the larvae varies as already indi- cated from two to two and a half days and over. ‘The interval between migration and pupation varies from two to four days and over, but the actual period of pupation is more constant— about eight days. In the region studied all of the above periods are generally quite regular, so that we may consider the period of development from egg to imago as covering about fifteen days, varying a day either way. Sarcophaga sarraceniz Riley: “This common large gray flesh fly ‘This weight was obtained by weighing two sets of fifty eggs, four or five hours old. The weight of each set was 5 mg., and by throwing the two sets together ro mg. resulted. Later five sets of twenty eggs were treated in the same manner, resulting in a like average. Ecological and Experimental Study of Sarcophagide 55 deposits living young. ‘The average weight of newly extruded larvee, before any food has been taken, is about 0.2 mg.’ ‘The erowth of the larve is very rapid, and the feeding period 1s about the same as Lucilia cesar (two to two and a half days), but pupation takes place more regularly. ‘The interval between migration and pupation is about three days, and the pupal period covers about thirteen days. ‘Thus the period of development for this species is from ezghteen to nineteen days. Compsomyia macellaria Fabr.: The screw-worm fly is very abundant along the beach, far outnumbering Lucilia cesar. “Ordinarily living young are deposited. ‘These can be distinguished from Sarcophaga sarraceniz larvae by their small size and lack of the prominent dark coloration of the head. ‘The complete larval period requires about five days, the pupal period about four days, thus giving a very short period of development, namely, about nine days. Sarcophaga assidua Walker: Of this species only two adults were reared. ‘This small gray flesh fly, strongly resembling the house fly in size and appearance, has a larval period of five and six days for the two reared, and a pupal period of seven days, giving a total of from twelve to thirteen days for development. III NORMAL GROWTH OF FLIES THROUGH LARVAL AND PUPAL STAGES In taking up the study of growth of the species below named, it was decided that weight is the most convenient and most readily applied method of measurement for these forms. As weight has been the basis of measurement for observations made by other investigators, it also forms a readier means for comparison. Growth is best represented graphically by means of curves, and below is thus shown the course of growth of Lucilia caesar and Sarcophagasarracenie in terms of weight in milligrams. The eggs of Lucilia for this experiment were collected and weighed on the afternoon of July 12, 1905. The larvae emerged eatly July 13, when the larve of Sarcophaga sarraceniz were 5This weight was obtained by weighing two sets of ten on August 31, 1905, which gave a total of 2 mg.; throwing the two sets together gave 4 mg, for the twenty larve. 56 William B. Herms started. It was not part of the original plan to trace the course of growth of this latter species, but when one of the females de- posited a quantity of larvz on several pieces of unprotected flesh, TABLE I Showing larval growth in six sets of Lucilia Caesar, including maximum weight at migration. A copy of a portion of the sheet used for tabulation of data I 2 3 1905 | 1905 | 1905 July | mgs.| July mgs. || July mgs | (10) (1) (10) (1) (10) (1) 12 | 7.30p.m |zoeggs | 2 12 | 7.30 p.m.|20 eggs 2 12) 7.30 p.m./20 eggs 2 | (ac) (io) | @) |] | 13 | 9.45 a.m.|/20larve | 4 13 |10.00 a.m.|20 larve 4 == | - |(10) (1.3) (10) (1.3) 13 |11.45 a.m.|15larve | 2 13, |12.00 a.m.|15 larve 2 13) 12.30 p.m.|10 larve 2 13 | 1.45 p.m.|10 larvee | 5 13 | 2.00 p.m.|r1olarve Sli Ss = = = 13 | 3.45 p.m.|1olarve | 6 13 | 4.00 p.m.|1o larve 5 13, 5.30 p.m, 10 larvae | 3 13 | 5.45 p.m.|1olarve | 9 13 | 6.00 p.m.|1o larve 8 13 | 7-45 p.m.|rolarve | 13 13 | 8.00 p-m.|1olarve | 10 | =| = = = 13 | 9.45 p.m.|10 larve | 17], 13 10.00p.m.jrolarve | 18 || = 3g) 111.45 p-m.|1olarve | 26 13 |12.00 p.m.jtolarve | 23 || | = 14 | 1.45 a.m.j1olarve | 35 14 | 2.00 a.m.jrolarve | 35 || —| et || = = 14 | 3-45 a.m.jrolarve | 49 |) 14 | 4.00a.m.J1olarve | 38 || 14) 3-30 a.m. 10 larvae 64 14 | 5.45 a.m.|1olarve | 61 14 | 6.00 a.m.Jrolarve | 52 || | = 14 | 7.45 a.m.j1olarve | 73 14 | 8.00 asm.Jrolarve | 65 || — = | = => 14 | 9.45 a.m./1olarve | 89 || 14 |10.00a.m.|1olarve | 82 14 | 1.45 p.m.|rolarve | 145 || 14 | 2.00 p-m.jrolarve | 173 || 414) 1.30p.m.1olarve | 178 14 | 5.45 p.m.|1olarve | 230 14 | 6.00 p.m.|rolarve | 208 | | | | 14 | 9.45 p.m.|1olarve | 248 | 14 |10.00 p.m.|1olarve | 270 | 14) 11.30p.m.|1olarve | 310 15 | 1.45 a.m.|1olarve | 297 || 15 | 2.00 a.m.|1olarve | 325 i eee ea ans = 15 | 5.45 a.m.|1olarve | 335 || 15 | 6.00 a.m.|rolarve | 330 | = = <= = 15 | 9.45 a.m.|rolarve | 359 || 15 |10.00 a.m.|rolarve | 371 || 15] 9.30 a.m.|1olarvae |*417 | (10) —|(386)|] | (10) —_|(404)/ (10) |(304) 15 | 3-45 a.m. Slarve | 193 || 15 | 4.00 p.m.| slarve | 202 || 15, 7.30p.m.| 5larve | 197 *Error (?) these were taken for the purpose and eventually furnished an excellent comparison. Ten sets of larvae were started at the same time, twenty in a set for several hours until they were larger when the number was Ecological and Experimental Study of Sarcophagide 57 reduced to ten larve in a set. Six of these sets were Lucilia and the remaining four Sarcophaga. Weighing began shortly after the larvae emerged from the eggs, July 13, and continued at regu- lar intervals during the entire period of development. Eggs and larva were first weighed at extrusion. TABLE I—Continued | | 4 | 5 | 6 LyOS ‘ || 1905 | 1905 July mgs. || July mgs | July | | mgs. Bes x = = u =| | % (10) (1) (10) (x) || | (10) (1) 12 | 7.30 p.m.}20 eggs 2 | 12 | 7.30 p-m.|20 eggs 2) ile sazel| 7-30 p-m.|20 eggs 2 (10) | (2) (eo) | (3), 13 |10.30 a.m.|20 larve 4 13 |10.45 a.m.|15 larve 2 Sel aie (ro) |@z) || (10) | (3) 13 |12.30 p.m.|10 larve 2 || 13 |12.45 p.m.j15 larve 4 || 13 | 1.00 p.m./15 larve 5 13 | 2.30 p.m.|rolarve 5 13 | 2.45 p.m.|1o larvee 5 | — 13 | 4.30 p.m.|r1o larve 6 13 | 4.45 p.m.|1o larvee 5 || = 13 | 6.30p.m.J1olarve | 14 || 13 | 6.45 p.m.|1olarve 7 || 13 | 6.00 p.m.|10 larve 8 13 | 8.30p.m./1olarve | 19 | 13 | 8.45 p.m.|10 larve TIE a oe = a 13 |10.30 p.m.|1olarve | 31 13 |10.45 p.m.|1olarve | 10 | 13 |11.00p.m.|1olarve | 23 14 |12.30 a.m.|1olarve | 35 || 14 |12.45 a.m.jrolarve | 15 | 14 | 2.30a.m.|1olarve | 39 | 14 2.45 a.m.|1olarve | 17 | = = = = 14 | 4.30a.m |rolarve | 46 14 | 4.45 a.m.j1olarve | 25 | 14 | 4.00 a.m.|1olarve | 45 14 | 6.30 a.m.J1olarve | 65 || 14 | 6.45 a.m.|1olarve | 35 || 14 | 8.30a.m.jrolarve | 73 14 | 8.45 a.m.|1olarve | 48 14 | 9.00a.m.|1olarve | 84 14 |10.30 a.m.\1olarve | 80 14 |10.45 a.m.|1olarve | 50 =|] = == = 14 | 2.30p.m.jall dead | — 14 | 2.45 p.m.|rolarve | 69 14 | 2.00 p.m.|rolarve | 151 = = = =| 6.45 p-m.|1olarve | 125 | 14 | 7.00 p.m. tolarve | 223 = — — — || 14 [10.45 p.-m.|1olarve | 158 | 14 |12.00 p.m.|1olarve | 284 = = = — || 15 | 2-45 a.m. 1olarve | 200 | = = ss —= — || 15 | 6.45 a.m.jr1olarve | 255 || 15 | 5.00 a.m.|1olarve | 350 = — — — | 15 |10.45 a.m.|1olarve | 314 | 15 |10.00 a.m |1olarve | 374 (10) |(360) | t = — = rh 15 | 4.45 p.m.| 5larve | 180 | 15 | 3.00 p.m.|1olarve | 390 Four sets of Lucilia were weighed at two-hour intervals for twenty-four hours and the interval increased two hours for each succeeding twenty-four hours until migration. Another set of the same species was weighed at five-hour intervals throughout the larval period to migration, and another set at ten-hour intervals. 58 William B. Herms TABLE II* Showing the average change in weight of Lucilia cesar Linné, from the egg to the adult flesh fly (Set No. 2) I 2 3 4 5 GU aa mags 9 10 II = x y be Ss o b| = a = = = o = w | o M ge] oo» (eles) § |eela 2 ele gale 6:3) ae |e mara Ss ae HBlsa| 2 |S S/S Asses ems) oss a = Be es a oO je |4 aoe ae = ms 12 | 7.30p.m. | eggs 2 20 | I co) = = eggs 13 |10.00 a.m. g 4 aio || ay) 2a) |e 20bis 050 50. larve 13 |12.00a.m. 5 2 D5 |, | iS), == -ep 2 bre || -sogcuslae tyes 13 | 2.00 p.m. | 7 5 10 a5 agp |p 2 lines +185 142.30 13 | 4.00p.m. | 9 5 10 5
    6, 3> 6, 35 Q; 12, 6; 6, Q; 6, 6, 2540: 1905-9, 125) 45.3, 35 33'2> 85 4, 05.3, G, 0510310, 650, 35 40 It can readily be seen that 3 is again the prominent factor. Fish are only cast up in quantities by a surf, and a surf is alone caused by a prevailing wind from the northeast or east. These fish are practically the sole food for the flies along the beach and this is especially true of those individuals living on the narrow strip of sand called Cedar Point, upon which the laboratory is situated. Were adults to emerge from their pupa cases at a time when no fish or very few fish were present on the beach the proba- Ecological and Experimental Study of Sarcophagide 75 bilities are that such individuals would suffer starvation. If this were often repeated the tendency would be to impair the vigor of the species, especially by interfering with the normal egg-laying habit. This latter would certainly be the case if the usual number of adults were to emerge with an undersupply of food present upon which the eggs could be deposited. ‘The large number of larvae for the short supply of food would result in producing smaller indi- viduals, which has been proven by experiments. ‘That the sarco- phagids given in the list are normal, as compared with individuals of the same species breeding elsewhere, is evident to the most super- ficial observer, and they are certainly not less numerous. Considering the above facts and also bearing in mind that the food supply is influenced by the comparative regularity of the surfs, there seems then to have been somewhere in the past an adaptation to the surf-producing storms. When the adult fly emerges from the pupa case it is likely to find available food on the beach, or has but a very short time to wait for it. ‘Then since egg deposition and food supply are so intimately connected, eggs are deposited and the cycle begins anew. As soon as the liquids have been sucked from the accessible parts of the fish, egg-laying ceases and the remainder of the work is left to the larva. ‘The presence of juices would then seem to be a gauge for regulating the number of eggs and young larve deposited on one fish by the females. ‘This will recall the state- ment made above that eggs are seldom if ever deposited on fish that have become dry, which fact should also be borne in mind in connection with.the adaptation to the surfs. If fish were to lie around for any length of time before the flies emerge, the juices would be dried up by the sun, and the fish would become unfit for food. However, it is very probable that the adults would after all deposit eggs on the dry fish. Lack of food for the adults would necessarily be a serious menace to the species. Under conditions as they now exist a drying out of a fish by the sun would not likely ° occur, since the flies would not permit a single fish to dry out thus. The assertions relating to this are based on laboratory experi- ments, 7. ¢., drying out a fish in the laboratory and then placing it outside in reach of flies. 76 William B. Herms That which is of principal interest in regard to this correlation of life histories to the surf-producing storms is the brief interval between the storms, represented by the factor 3 or 6, and this with its relation to the days required for development with each species, viz: Compsomyia about 9 days, Sarcophaga assidua about 12 days, Lucilia about 15 days, and Sarcophaga sarraceniz about 18 days. Further, it must be remembered that the life history of each species for this locality covers a comparatively definite period, which is a necessary consideration in this matter of correlation. When eggs or larvae were collected, very little chance was involved in predicting the date on which the imagines would appear. ‘The writer made use of this factor in his experiments with the three most abundant species. It would also be useless to speak of a correlation to the surf producing storms if the life histories of the species studied here corresponded to the life histories of the same species in localities remote from a beach. From the literature consulted the following data was secured relative to the latter. Compsomyia macellaria: Morgan (’90) gives (August 18 to August 29-30) rz to 12 days, Francis (’90) larval stage about a week and pupal stage from g to r4 days, a total of from 16 to 21 days. Sarcophaga assidua: Howard (’00) gives (July 3 to July 25) 22 days, also (July 9 to July 18-26) 9 to 17 days. Lucilia caesar: Howard (’00) gives (May 12 to 29) 17 days. Sarcophaga sarracenie: Howard (’00) gives (May 12-30) 18 days; (July 2-29) 27 days; (June 6-17) 11 days, (June 13-26) 13 days; (July 7-21) 14 days; (July 9-22) 13 days, (July 24 to August 9-11) 16 to 18 days; Kellogg ('05) gives 10 to 12 days, Howard (’02) 10 days. One can readily see from the above citations that there is a marked variation in each species, and that these periods do not coincide very closely (excepting the first period in the last-named species) with the results secured in these studies. It must, how- ever, be admitted that more extensive and systematic work should be done relating to the question under discussion. Conditions as stated above may lead to the impression that egg deposition is a direct result of the presence of food within the Ecological and Experimental Study of Sarcophagide 77 alimentarycanal. But this is evidently not a necessity, as is shown by the following observations. In carrying on experiments it was always a matter of concern from the first to guard against outside larva. On several occasions fish were covered by a screen of netting to keep out flies, but the females of Sarcophaga sarracenie invariably deposited their young on the netting and these then found the fish without much difficulty. While carrying on experiments indoors the flesh was kept in Petri dishes and covered with like dishes. Several times it happened that a female of Sarcophaga sarracenie gained en- trance to the room through the door and deposited larvz on the outside of the dishes. Not being able to get at the flesh the larva perished. Furthermore, in such cases where the head of the fish was hooded with cloth, the females of Lucilia czsar deposited eggs very freely on the cloth and also on the loose ends of the string used to tie the hood. ‘These observations led the writer to be- lieve that it is not necessarily the presence of food within the ali- mentary canal that stimulates egg deposition. In this connection, however, it might be interesting to note, that no eggs were secured from individuals of Lucilia caesar kept under confinement with plenty of accessible food. ‘The flies crawled about on the fish apparently sucking the juices, but all died in a short time. Con- finement very probably was the cause. ‘This evidently agrees with experiments on the house fly cited by Howard (’00), viz: “T am inclined to believe that what may be termed the psycho- logical influence of confinement, even in so large an enclosure as the one used in the 1898 experiments, alarmed the flies, caused their early death, and prevented them from obeying their natural instincts and performing their natural functions.” VII EXPERIMENTS UPON THE TROPISMS OF FLY LARVAE The following experiments and observations are not intended to cover the topic of tropisms and their relation to fly larve with any degree of thoroughness. The object of this final chapter to the general paper is to present a statement of experiments made on movements in reaching the food and in migration, including a preliminary discussion. 78 William B. Herms Chemotaxis ‘Two experiments were tried with reference to chemotaxis. First Ex periment— Thirty-six larve between four and five hours old, were placed in a small vial 4.3 cm. in length, and a piece of fish weighing about one gram was placed 1.5 cm. from the bottom of another vial 12.7 cm. in length. While this was being done, care was taken that the flesh did not come in contact with the sides of the long vial. After the larvae were shaken to the bottom of the smaller vial, the two were put mouth to mouth horizontally ona table. ‘This took place at 9.05 o’clock, a.m., June 30, 1905. In three or four minutes there was a decided movement toward the mouth, but because of the unevenness of this region, the move- ments became scattered. At this juncture a short piece of paper was placed like a bridge inside the vials connecting them. The following table shows the results of the experiment. Un- fortunately the data taken were not sufficient to make a complete table. Second Experiment—A small piece of fish weighing about two grams was placed in the center of a large sheet of heavy white paper, then young larva were put at different distances from the flesh, after dipping them partly (posteriors) in glycerine so that a trail would be left in crawling. The first larva was placed with head toward the meat at 9 cm. distance, and reached the food in four minutes after taking a some- what winding course. The wind was favorable in this case. A second larva was placed at a distance of 11 cm. with its head away from the flesh and the wind at right angles. ‘This larva started at 10.32 o’clock, a. m. and after a very circuitous route, circling frequently though always drawing nearer and never going beyond, reached the food at 10.52 0’clock. ‘Time, twenty minutes. ‘The course of this larva took it considerably to one side of the flesh almost to the starting place of the first larva from which point the two paths to the food were almost parallel. A third larva was placed at a distance of 11 cm. on the wind- ward side of the food. After a great deal of traveling, making many circles and stopping frequently like the other two, it reached Ecological and Experimental Study of Sarcophagide 79. Time Started TABLE VI Showing result of the experiment above indicated ae June 30, a.m., Time Arrived Larve mes = 9.05 vials placed I 9.20 9-30 2 9.27 pio 3 9.28 9.31 4 930 9-37 5 O32 O38 e 6 9-32 II .0O-12.00 7 Wei 9°39 8 no record 9-39 9 no record 9.40 ite) no record 9.42 II no record 10.00 12 no record 9-52 x3 Doris) Jebu Mer e535) B)o UU ey 15 9.46 9.51-10.15 16 9-47 9. 51-10.15 17 9-47 g.51-10.15 18 no record g.51-10.15 19 no record g.51-10.15 20 no record Wao). 105 21 no record 9-51-10.15 22 no record 10.15-10.30 23 no record 10.15§-10.30 24 no record 10.15-10.30 25 no record 10.1§--10.30 26 no record 10.15-10.30 27 no record 10.30-I1.00 28 no record 10.30-I1.00 29 no record 10.30-11.00 30 10.40 I0.30-I1.00 31 no record II .00-12.00 32 no record II .0O-12.00 33 no record 10,00-12.00 34 no record II .0O0-12.00 35 no record II .00-12.00 36 no record II .00-12.00 Time Required in Minutes eee ol 4 3 6 2 got 7 betw. 6-7 betw. 6-8 betw. 6-10 betw. 6-28 betw. 6-20 6 betw. 6-30 betw. 7-30 betw. 8-30 betw. 8-30 22 larve arrived by 10.15 27 larve arrived by 10.30 io 30 larve arrived by 11.00 Observations interrupted by lecture at 11.00 36 larve (all) arrived by 12M Note—When the larve started toward the food, they hastened on, stopping once in a while to sway the head about in the air for the purpose of orientation. If the flesh was not reached by means of a direct route, as for instance along the upper side of the vial, the larve crawled to the end, then down and 80 William B. Herms a point farther away from the flesh but on a line with it, and the starting point of the firstlarva. ‘[his process required an hour and four minutes and the larva died at this place apparently from exhaustion. A fourth larva placed 34 cm. to one side of the starting point of the first also failed to find the food. Its course led it farther away and finally off the paper. Discussion—From the above two experiments it will be seen that there are two factors involved in finding the food. First, the primary stimulation of the larve by meansofthefood. Whateverthenature of this stimulation may be, and whatsoever the internal mechanism involved, the process which underlies the turning of the larva in an effort to draw nearer to the food, may be termed chemotaxis. The second factor is the swaying of the head from side to side or in anarcofacircle. This the larva does for the purpose of orienta- ° tion, and the process may be termed, according to Holmes (’05), “Selectionof Random Movements,” or,accordingto Jennings (’04), “Trial and Error Movements.”’ Both processes cover the case equally well. The writer has been unable to detect any dissimi- larity between the two theories, as applied to the behavior of fly larve. The larve stop frequently in their course, sway the head as above indicated, also circle frequently while crawling. ‘The same course may be pursued again or there may be a change in direction which is generally the case after a pause. It 1s clear that an over- production of random movements is involved; that a selection is made from these, depending on the force of the stimulation, and that the larve are thus guided on their way. On the other hand, it may be said that the larva reach the food successfully because they pause frequently in the course and sway the head about in order to try the conditions, then when they change the course, it back to the food. Comparatively few found it necessary to do this, since the more direct route was naturally along the lower side. The smallest larve seemed to have the most trouble in reaching the food. One very small larva (No. 6) remained within a distance of 2 cm. from the flesh for over an hour and a half. No larva left the flesh to return to the smal!er vial, though once in a while one started away, but always to return in afew seconds. The larve were under observation all day and all evening. Ecological and Experimental Study of Sarcophagide SI is evident that an error has taken place. ‘Therefore, it does equally well to apply the “Trial and Error Theory.” The following quotation from Jennings (’05), p. 475, apparently, makes little distinction, if any, between the two theories just men- tioned: “We perform movements which subject us to various con- ditions, till one is found that relieves the difficulty. We call the process searching, testing, trial, and the like. In the lowest and highest organisms the injurious condition acts as a stimulus to produce many movements, subjecting the organisms to various conditions, one of which is selected.” The use of the terms “selected” and “ selection” which frequently recur in the paper above quoted should not be overlooked with- out a thought as to their significance. ‘The first impression is that these terms imply intelligent choice on the part of the organism, but the author (Jennings) undoubtedly expects the broader inter-. pretation, such as expressed by the term selection when applied to a magnet. This also holds equally well for the theory of the “Selection of random movements.” ‘There is in reality no intel- ligent choice involved; the organism responds reflexly to the stim- ulus, either positively or negatively. Each of the three theories of animal behavior evidently ex- plains much, but the writer believes, at least in reference to his own experiments cited above and others cited below, that the tropism theory is not sufficient without either the second or third theory, and vice versa. It is still largely a matter of theory whether animal behavior can be so readily explained. Even in fly larve we have to deal with what seems to be a death feint, and that in itself leaves much to be explained. Phototaxts Observations were made on the same larvz used in the first experiment. In the evening of the same day (June 30, 1905,) on lighting the lamp, the larvz were noticed to leave the flesh at once and hasten toward the side of the vial nearest the light. ‘This ‘took them 4.5 cm. away from the food. Changing the angle between the vial and the light or rolling the tube over always resulted in a readjustment on the part of the larve. Moving the 82 William B. Herms lamp from one side of the tube to the other resulted likewise. Gradually increasing the distance between the vial and the light resulted in a return to the food when a maximum of thirteen feet was reached. On decreasing the distance again, the larve once more left the food when a distance of nine feet was reached. ‘This experiment was repeated several times with like results. The lamp used was an ordinary oil lamp with small (No. 1) wick turned up fairly well. “he adjustment to what was appar- ently the exact point of greatest photic stimulation was very remarkable, as was the almost frantic effort to gain this point when the angle was changed. Here we have an example of positive phototaxis overcoming the action of positive chemotaxis which is surely not a useful reaction. Stereotaxts The larvze when placed in a receptacle which was ridged, preferred to crawl in the grooves. In one instance larve were kept in a bottle which had a convex bottom; on examination later, all were found in a circle wedged in close together around the margin of the bottom, with heads down and posteriors extended. On several occasions larvze were found crawling in the crevices of the floor, and some of these were wedged in so tightly that it was a task to extricate them without injury to the larve. Positive stereotaxis is a prevailing phenomenon in the lower orders and fly larve are no exception. Geotaxis Fly larve are positively geotactic, the burrowing habit ( ?) being very marked. On the other hand, imagines when first emerging from the pupa cases crawl out of the sand and up nearby grasses, remaining there until the wings are spread and dry. The cut showing this also illustrates how the flies cling to the grasses with head downward. Note—The experiments and observations relating to this paper were conducted at the Ohio State University Lake Laboratory at Sandusky, Ohio, chiefly during nine weeks of the summer, 1905, and a portion of the summers, 1903 and 1904. The writer is indebted to Prof. Herbert Osborn, Director of the Lake Laboratory and Asso- ciates, Profs. F. L. Landacre and J. S. Hine; also to Dr. W. E. Kellicott, Barnard College, for the kind assistance rendered and suggestions offered during the course of these studies. Ecological and Experimental Study of Sarcophagide 83 LITERATURE CrbeD Davenport, C. B., ’04—Statistical Methods with Special Reference to Biological Variation. New York. Francis, M.,’90—The Screw-worm. Bull. No. 12, Texas Agric. Exp. Sta., Sept. Hing, J. S., ’04—A Note on Insects as Scavengers, etc. Jour. Cols. Hort. Soc., vol xix, sec.;/pp; 123-128. Homes, S. J.,’05—The Selection of Random Movements as a Factor in Phototaxis Jour. Comp. Neurology and Psychology, vol. xv, Mch., pp. 98-112. Howarp, LELAND O., ’00—A Contribution to the Study of the Insect Fauna of Human Excrement. Proc. Wash. Acad. of Sciences, vol. 11, Dec. 28. ’o2—The Insect Book. pp- 429. New York. Jennincs, HERBERT S., ’04.—Contributions to the Behavior of Lower Organisms. 7th paper. The Method of Trial and Error in the Behavior of Lower Organisms. pp. 235-252. Carnegie Institution, Washington. °05— The Method of Regulation in Behavior and in other Fields. Jour. of Experimental Zodlogy, vol. 11, No. 4. KELLOGG, VERNON L., ’05—American Insects. pp. 674. New York. Minot, CHARLES SEDGWICK, ’9I—Senescence and Rejuvenation (Plates IJ, III, IV). First Paper: On the Weight of Guinea Pigs. Jour. of Physiol., vol. x11, pp. 97-153. Morean, H. A., ’90—Texas Screw-worm. Bull. No. 2, Second Series La. Agric. Exp. Sta. NeepuaM, J. G., ’oo.—Insect Drift on the Shore of Lake Michigan. Occasional Memoirs of the Chicago Entomological Society, vol. i, No. 1. REJUVENESCENCE AS THE RESULT OF CONJUGATION BY SARA WHITE CULL Thirty years ago Bitschli proposed the view, since confirmed by Maupas and others, that the life histories of infusoria run in cycles, and that a period characterized by binary fission 1s fol- lowed by another in which conjugation takes place; this latter process resulting in a thorough reorganization of the excon- jugants and a Verjtingung or rejuvenescence, which shows itself in a higher rate of cell division and, generally speaking, in renewed life activities. If conjugation does not take place nor an equivalent stimulus be given the organisms they will eventually die of what has been termed “protoplasmic old age.” Hitherto it has been supposed that both cells in conjugation were benefited by the process, a mutual fertilization taking place; - but in a series of experiments made by Calkins on Paramecium caudatum, the fact was noted that when both exconjugants live, in some cases one is far more vigorous than the other, as demon- strated by the greater number of offspring in one case than in the other.t. Dr. Calkins suggested that I should examine this point and carry out some other observations that he had already made. The work was done in the zodlogical laboratory of Columbia University in the fall and winter of 1905-06. Butschli has pointed out the striking analogy which exists be- tween conjugation and fertilization as it is seen among higher organisms and among those protozoa which show sexual dimorph- ism. In many of these forms such as the peritrichous ciliates or the coccidiida, there is a marked sex-differentiation in the size and activity of the gametes. Here in fertilization, a more or less passive individual of normal or more than normal size, a macro- 1 Studies on the Life History of Protozoa. 1 Arch. f. Entwk., Bd. xv, 1 02. Tue JourNAL oF ExPeRIMENTAL ZOOLOGY, VOL. IV, NO. I. 86 Sara White Cull gamete, completely fuses with a smaller cell of greater activity, a microgamete. [he complete union of two cells alongwith differences in size and activity are characters which distinguish the process of fertilization as usually understood, from the process of conju- gation, as seen in forms like Paramecium. Both processes agree in having the same essential feature, the union of the nuclei of the two cells. In the different classes of protozoa all steps may be found from conjugation in a general sense to a process exactly similar to fertilization used in a strict sense. Even the maturation phenomena which play so important a role in the history of metazoan germ cells are represented in some sort by processes which have been observed in a few protozoa. In isogamous union, such as that which takes place in Para- mecium caudatum, two individuals of the same size and approxi- mately equal activities unite for a short time, and the ectoplasm around the mouths of the two organisms fuses to form a sort of bridge over which the nuclei pass. During the maturation phases, previous to this nuclear exchange, the micronucleus of each organism gives rise by division to four or more pronuclei. Two of these are destined to be functional and the others, cor- puscles de rebut, as Maupas calls them, disintegrate. One of the two functional pronuclei passes into the other organism where it fuses with the stationary pronucleus of that cell, forming one single reorganization nucleus. From this, by repeated division arise the micronucleus and macronucleus of the rejuvenated protozoan. These organisms then proceed to reproduce by ordinary fission. The species used for the experiments described here was Para- mecium caudatum and the material was what is known as “‘wild.”’ Each conjugating pair was taken up in a fine pipette and put into a hollow slide containing some drops of the culture liquid—hay infusion—free from all other protozoa. ‘These slides were then put into moist chambers. In all cases an examination was made after the isolation of the conjugating pairs to see that they had not been separated in the process of handling, forif this precaution were not taken, one could not be sure of dealing with the results of conjuga- tion. On the day following isolation, when, in most cases, the exconjugants were swimming freely through the water, each one Rejuvenescence as a Result of Conjugation 87 was put into a small glass vial containing liquid similar to that from which they had been taken, and these vials were marked in such a way as to indicate the connection between the various individuals. ‘These vials were examined and the animals counted every few days fora month, anda fresh but not a new food medium was given them each time, the same being used for all the organisms. Ninety-three pairs of these wild conjugants were isolated at different times and of that number at the end of one month repre- sentatives of sixty-five pairs, or seventy per cent, were alive. At least one of the original conjugants remained of each pair, in the majority of cases both had given rise to offspring. Forty pairs of conjugants from long-continued cultures living in the laboratory on hay infusion were isolated by Calkins (loc. cit.) and examined from time to time. Only six pairs, or twelve per cent of these paramecia were represented by living forms at the end of a month. A comparison of-these observations with those now made on the ‘“‘wild” material would seem to indicate that the fertility of conjugation is dependent upon the condition of vitality in the individuals pairing, for, in both cases, the medium was the same. On the other hand, the explanation may lie in the fact that both conjugants had lived in the medium for many months, so that their chemical composition was too similar to pro- duce a new compound by fusion of their nuclei, this new com- pound being, perhaps, the source of energy for reorganization. A study of the mortality of these paramecia showed that at the end of one week the strains of both conjugants had died out en- tirely in six percent of the original ninety-three pairs. After three weeks had passed thirteen pairs, or approximately thirteen per cent, had died. ‘These facts confirm Calkins’ observation that conjugation is by no means always successful in producing reju- venescence. The point which interested me chiefly in these experiments was that of double or reciprocal fertility—do both conjugants possess new power and ability to carry on the activities of life, or is but one of them fertilized as is the case among higher organisms? The statistics which were gathered with this in mind show that, at the 88 Sara White Cull end of the month, of the sixty-five pairs then represented by living cells, in twenty-seven pairs, or forty-one per cent, one of the excon- jugants only or the offspring from it were alive; in fifteen pairs, or twenty-three per cent, the progeny of one exconjugant was three times as large as that of the other; in six pairs, the descendants of the one were twice as numerous as those of the other organism; and in only five cases had both conjugants given rise to the same number of offspring. The twelve remaining pairs showed a wide disparity in the number of paramecia produced by any two con- jugants. ‘The following table shows these results in summarized form: ; ProGEeNny oF ONE ProGeny or Boru PROGENY oF ONE ProGeny or ONE ConjJuGANT TwIcr CoNnJUGANTS Conyucant Derap, ConyuGANT SHows Drab OF THE OTHER ALIVE gees eka GREATER VIGOR OrHer* Number of Per- |Number of Per- |Number of Per- Number Per- Pairs centage Pairs centage Pairs centage | of Pairs centage After 7 | | days 6 GSH 20 | 22 2y 33 = = After 20 | days 13 ey Sas | 31 28 35 = = After 30 | | days 28 aq) | 27 | 41 21 32 48 74 * Exclusive of cases where the progeny of one exconjugant had died. It may be broadly stated that of the sixty-five pairs which I have’ observed one conjugant either died or left a weak strain in which the descendants were half as numerous and much less vigorous than those of the stronger exconjugant. ‘This striking difference in the restored vitality of the conjugants and their descendants gives strong grounds for the belief that conjugation as seen among these infusoria is really incipient fertilization as seen among the higher forms of life. Here we have indications that one gamete gives up its vitality to and loses its individuality in the other just as the spermatozo6n loses its identity in the egg where its presence forms a stimulus to development analogous to the rajewnissement and greater activity in cell division which follows conjugation. There is little reason to doubt that a physiological and perhaps a Rejuvenescence as a Result of Conjugation 89 physical difference exists between the two unicellular organisms which unite in conjugation and a difference of the same nature as that expressed morphologically in the case of Adelea ovata, where the male gamete does not fuse with the female but dies after delivering one of its four pronuclei. Baltimore August, 1906 A f ; ; i. , ihe, "7 j ‘. i € = 4 ey » d a + , As q A 2a) A ' } . Ua. we ~ ; ‘ i . at ARTIFICIAL PARTHENOGENESIS IN THALASSEMA MEBEIEA BY GEORGE LEFEVRE Wir Six Prates i Deal aa gota NYE Cie laditacl caPAenn Gro CIE aeRO RICE ae OHSS E My SAAR Aipicte oO orccinnerka ccm isio ccna g! Pie Axttieaibarthenopgenesrsm Anmelidsera.rtu0 ctr claire ateistteie tats arerenee a echyte alee yet ie eet ale 93 MW? Mixterialiand) Methods) Sapvs crete cle =xcinre sn w cecloln nn ace onus ae eiete eee oe Mimo snes acinar ks 97 I\iig Jeb (o@otaGalPUGanhie a oan os 6 ome Aen OU OOnOUMAB ABTA OSORSOgcasucOopugbouatopoe sac 98 TeeAciisiasiartnenorenecicia Cen tsmey sci fa ae icit tater ter tard ter tted tee rete te act eee 99 z Artificial Membrane-Formation and Parthenogenetic Development ................. 104 View @ bservahotson tie ivimedWatertall tances stepietoisrtcte te iterate ier tee tetelaaree t-te aan 109 it ADS Wineeautsntieel ss be oh ad oeBounns Job au bd rads HennosebanosU GoM soonaG one 109 2) Bormationvof PolariBodies yi... acct e sce else seve det tebe parts aois arsietenol yer aeays IIo 3), KOIGENETS btecd cb SdacnclongdapebAnsobnanoGb oe ooUbnbob ooo DabRAtpo chee Ee acn as 112 45, Bormation of Water Embryoran dara seers sold itil le rere ee 114 Gee Albnormalitiesioy Cleavare andy Beli aviory sme acssit ler tesere ttyl tetas eat 116 Wiis @Obsevrations onjthe PresenvediMaterialie ren .csaa ne ars rivera ee eee a eee 119 Tene. Oocyte andthe Maturation) Divisions nyariyrii iste reer eradicate ete 120 Ze OripimohtueCleavyape Nucleus andisunplita ster weer lsat llereteetepel etcetera irre 124 Seine Cleavage:Stavestacne sr eyiiee crt rate orn etyererack cy ee Re Rereaces eer eel cree eee 125 4 Gastrulation and the Formation of the Trochophore .................-...--+------ 127 Be RudimentaryGells, .tyeccie titan seks) easter tibiae taste isicee ree tery ere 129 SeeabnormallMaturatronsbhenomenare ae menace aie teeters 130 aavabsenceolthe Second Polaribodyae. ade eee soca ae cee eee teres 130 be Absence'of Both Polar Bodies 22) 4-2 2542-4644o eee Ee RA ohn tent nec Iara 132 PaO NDHOGLIM AN MILOSES? ters =: 5 Io m MOM Soea le laoOe 1100: COMSCA=WALEN oF ai.cis poo 5 a phates ante ess OIE ee aes 8 20 m . . 2 Cee Oxalic acids-}-'88) Gen 'SGa-Water tc. cles sce yee 8 20 m . . WGC Gree ACECIC. VACIG <=}. Oy CEs SEA-WaAteh na 2 wueicttess aca eee he 5 fe) In the case of CO,, the gas was passed from a generator into sea-water for ten minutes and the eggs immersed in the charged water for one hour, after which they were transferred to pure sez- water.? The result was very satisfactory and usually about 50 per cent of swimming larva were obtained by this method. Although a wide range of solutions and exposures were tested in the case of each acid, in the table on page 12 are placed a few results which are selected from a great many experiments and which will serve as characteristic illustrations. After determining by experiment the optimum solution and exposure in the case of each acid, satisfactory results were usually obtained by adhering more or less closely to such conditions as experience had proved to be the best, but an examination of the following table will show that the expectation was not always fulfilled. For example, in Nos. 3 and 4, when the same solution of HNO: and the same exposure were employed, one experiment yielded 40 per cent of swimming larve, while the other gave only 5 per cent; and again, in Nos. 5 and 6, 60 per cent and 25 per cent were obtained, respectively, from an equal exposure to the same HCl solution. It is difficult to assign causes to this seemingly capricious difference in the relative proportions of developing eggs in experiments carried on under conditions as nearly identical as possible. In addition to the variability of the results obtained in different experiments, where the same solutions and exposures were used, 2A “sparklet”’ apparatus was not available at the time my experiments were made. 102 George’ Lefevre Wanetar PHOS No. Solutions Employed Exporure of Swimming Trochophores m . I 17, (cc; = EIN @s i Say ccenSe We 5 minutes 55 ie) m : 5 2 18 cc. —HNO; + 82 cc. S.W. 3 minutes U7 10 m . 3 18 cc. —HNO3 + 82 cc. S.W. 4 minutes 40 10 m ‘ 4 18 cc. —HNOs + 82 cc. S.W. 4 minutes 5 10 m ; | 5 fi Ges le (Cl oe ig ees SMe 5 minutes | 60 10 m . 6 rity (ee adn (OM oe ly ees SNA, 5 minutes 25 10 m . 7 160ccs — HCl 8aNcc.S.W- 5 minutes 14 10 m . 8 TS 9 CG. ELC] 4-8 2ces.SeWVi- 4 minutes 4 10 m 9 10 cc. —H»SOx4 + go cc. S.W. 8 minutes 35 20 m 10 12 cc. — HeSO,4 + 88 cc. S.W. 8 minutes 10 20 m II 15 (cc. —IHSO,4 > 8giccss.W. 5 minutes 6 20 m 12 12 cc. — Oxalic + 88 cc. S.W. 8 minutes 50 20 m 13 15 cGa— 1 Oxalict|-1 6) CG SW 6 minutes 45 20 m 14 TONCC.—ACELC: on GOlCCase Ws 7 minutes 30 10 m 15 WG CCe —WAGELIG Sots SiS CG. Ss We 5 minutes 60 10 m 16 ity (le = aN aint St (eles SENIG 6 minutes 60 10 17 COz passed into water for 10 minutes. 1 hour 50 18 CO, passed into water for 20 minutes. y hour oo Artificial Parthenogenests in Thalassema Mellita 103 I was greatly struck with the marked difference in results observed when the strength of the solution or the duration of the immersion was varied bya very slight degree. For example, ina given experi- ment 60 per cent of the eggs developed into actively swimming trochophores, which could not be distinguished from normal larve, after five minutes’ exposure to the following solution: 15 cc. = HC1 +85 cc. sea-water. Another lot of eggs from the same female, treated with the same solution, but for 6 minutes instead of 5, yielded only about 5 per cent that underwent any develop- ment at all, while in none of the eggs did this proceed beyond the early cleavage stages. Here a difference of but one minute in the time of exposure gave rise to a profound difference in the result, in the one case the solution being adequate to initiate the developmental processes in a majority of the eggs, which then pro- duced apparently normal larve, while in the other case only an abortive early development was induced in a very few eggs. Such differences, however, in the relative proportion of larve were by no means constant; in Nos. 15 and 16 of the table it is.seen that a difference of one minute in the exposure to the same solution of acetic acid had no effect upon the percentage of larve obtained. The following table illustrates a similar variability in cases where the duration of immersion was constant, but the solutions differed very slightly in the degree of concentration: | ; | Percentage No, Solutions Employed | Tie | of Swimming | Exposure Pa ae h | | | rochophores | m | | 1 | 17 cc.—HNOs + 83 cc. S. W. 5 minutes | 40 10 | | ters | 20) 18ce- HINO: > 182 cc. S. W. | 5 minutes | 3 | Io | | eth | | Bn I4icc. Ace 86) cc1S. We _ 6 minutes ° 10 m | } A |) TS (CINE Se CS le Ss Nive 6 minutes 55 | 10 | 104 George Lefevre In each of the two cases cited above, the eggs were taken from the same females and placed in the solutions at the same time, the only difference being that the second solution was stronger than the first by I cc. of the dilute acid. 2 Artificial Membrane Formation and Parthenogenettc Development The unfertilized eggs of Thalassema after transference from the acid solution to normal sea-water, throw off a membrane iden- tical with that which is formed upon entrance of the spermatozo6n. The artificial production of a membrane has been observed by former experimenters. O. and R. Hertwig (’87) first discovered that, by the addition of chloroform to the sea-water, the unfertilized eggs of the sea-urchin may be caused to form a fertilization mem- brane which is entirely normal in appearance. Herbst (’93) later confirmed the result obtained by the Hertwigs, and found that not only chloroform but several other substances, namely, clove oil, creosote, xylol, toluol and benzol, act in a similar manner, the best results being given by benzol. More recently Herbst (’04) has obtained a normal membrane formation by the use of silver salts. Loeb (’o5d, ’o5e) tested the action of hydrocarbons in this respect and found that the ripe eggs of Strongylocentrotus and Asterina, when put into 50 cc. of sea-water which has been shaken with 1 cc. of benzol or amylene, immediately form membranes which are identical in appearance with the normal fertilization membrane. By subjecting unfertilized eggs of Strongylocentrotus purpuratus to a 24 to 14 7 NaCl solution or to a 24 1 cane sugar solution, he also succeeded in causing a mem- brane formation, but in these experiments the osmotic pressure was so high that the eggs were greatly injured and underwent cytolysis without subsequent development. Solutions of lower osmotic pressure caused development, but not membrane forma- tion (’04a, p. 79). It should be mentioned that Wilson (’oI, p. 533) states for ‘Toxopneustes that “some of the magnesium eggs showed a faint ragged membrane, but others were absolutely devoid of a membrane,” although he gives no details of his observations on Artificial Parthenogenesis in T halassema Mellita 105 this point. Hunter (04, p. 214) also records the presence of a membrane surrounding unfertilized eggs of Arbacia after treat- ment with MgCl.,. Loeb (’05), in a series of recent papers, has published the results of experiments which have confirmed my observations on the formation of a membrane after exposure of unfertilized eggs to acid solutions. Although our observations agree as to the power of acids to call forth a membrane formation, certain marked differences occur in our results, and it may be well to compare his experiments and my own in this place. By the use of an improved method, Loeb has succeeded in closely imitating the process of normal development in the unfertilized eggs of Strongy- locentrotus purpuratus. If the eggs are treated with hypertonic sea-water alone, no membrane is formed, and only a small per- centage undergo any development at all. ‘The rate of develop- ment of these is much slower than in the case of fertilized eggs and the larve arising from them do not rise to the top but swim at the bottom of the dish. By first exposing the unfertilized : n eggs, however, to 50 cc. of sea-water to which 3 cc. — of a fatty 10 acid, e. g., formic, acetic, propionic, butyric, valerianic or caproic . acid, are added, for from 4 to 1} minutes, they forma character- istic fertilization membrane when put back into normal sea-water. The membrane was not produced as long as the eggs were left in the acidulated water, nor was it formed when they were taken out a little too early or too late. Eggs treated with the acid alone do not develop, but in a few hours begin to disintegrate, and 584 diameters, Fig. 11 A section of the blastoderm of an egg thirty-one hours old that had spent the first twenty- three hours after fertilization in a 74; m MnCl2 solution. The central periblast, cpb shows much thickened, with many large nuclei accumulated in this region. c, Cavity inthe blastoderm. x 58} dia- meters, seen in living eggs. The embryos are always much dwarfed and pale. The heart never contracts although the embryo may remain alive for as long a period as two weeks. The pericardium is often puffed out and is unusually prominent, as also occurs in some other solutions, as shown in Fig. 12, pe for an embryo from a mixture of MgCl, and NaCl. 184 Charles R. Stockard Eggs that remained as long as thirty hours in 4 m distilled water solutions of KCI would recover if placed in sea-water. Other eggs were left for three days in KCl 3 m distilled water solutions, and afterward recovered, the heart beginning to beat, etc., when returned to sea-water. Normal embryos several days old were very readily killed if subjected to even weak solutions of KCI; their heart’s action being stopped. It thus seems as though an embryo may live and develop without its heart ever having con- tracted, but if it had once begun to contract any cause that may stop this contraction proves fatal. My results then in a general way agree with Loeb’s observations though I should take exception to his statement that the circulatory system develops normally even though the blood does not circu- late. ‘The major parts of the system do seem to develop but by no means nor- mally, the heart being small and weak and it is often only a straight tube with the balloon-like pericardium surround- ing it. Many clots of red corpuscles are noted in several of the sinuses. The above facts are also of interest in connection with Howell’s analysis of Fig. 12 An embryo from a mixed the inhibitory action of the vagus nerve ae Ae aes *™ on the heart beat as being due to the Yoenree hours om Sowin& liberation of K-ions about the nerve endings. A mixture of a } m KCl and a molecular NaCl solution was prepared with 60 cc. of the former to 10 cc. of the latter. This mixture showed the same general effect on the eggs as the simple KCI solution; the blastoderm bulged up slightly and the yolks were shrunken. Many embryos, however, seemed stronger and better developed with more pigmentation and with larger red blood clots. Some of these embryos were placed in sea-water when seven days old but failed to recover.. A CaCl, + m solution in distilled water proved highly toxic. The blastoderms flattened down, the cells apparently spreading unusually far apart. The eggs died within about twenty-four ewollen, pc, pericardium. The Influence of External Factors on Development 185 hours. Eggs that were subjected to { m CaCl, one hour after fertilization were almost all dead within four hours, the living ones were abnormal, and all died after twenty-four hours. ‘This result further illustrates the readiness with which the egg membrane is penetrated during the first few hours after fertilization. A mixture of 60 cc. $ m CaCl, and 10 cc. 1.0 m NaCl proved equally as fatal as the CaCl, alone had done, neither of the cations seem to exert an anti-toxic action toward the other. Solutions of NH,Cl in distilled water of concentrations ;'5 m, +m, +m, 4m, 4m and 4m were used; a molecular solution of NH,Cl is equivalent to about a 5.05 per cent solution. Sea-water solutions of } m, $ m and ? m were also employed. ‘These solu- tions seemed to cause the yolk to shrink slightly, the blastoderm to thicken so that on examining the living eggs one would think that the segmentation cavity was abnormally large as it is in the lithium embryos. On studying sections of these, it was found that the cells are loosely connected, making the blastoderm unusually thick so that it projects down into the yolk. The segmentation cavity is, therefore, not abnormally large as in the lithium embryos. Many of the eggs die at various stages. The rate of development is retarded and the blastopore is slow to close. Many of the embryos are short with their tail ending abruptly. In some embryos the heart beats slowly and the circulation is sluggish; in others there is no pulsation at all, and still others in the same solution may show a very good circulation. Embryos lived as long as eighteen days in such solutions but failed to hatch. Such short embryos as those above described seem to result from any cause that retards development and prevents the nor- mal down-growth of the germ-ririg. When such embryos were removed from 4 m NH,CI solutions when forty-three hours old and placed in sea-water, they recovered in one day and hatched when fourteen days old. The embryos in NH,Cl are always dwarfed, with poor circulation, lightly colored blood, and sparse pigmentation, having a pale appearance. ‘The sea-water solutions of NH,Cl were much less toxic than the distilled water ones. Mixtures of NH,Cl + m + MnCl, + m, NH, Cl 4m + -MnCl, 7s m and NH or +m + MnCl, #5 m were tried. The 186 Charles R. Stockard eges lived better in these solutions than in either the NH,Cl or the MnCl,. The first twenty-four hours of development is almost normal though some eggs die, the rate of development after this period becomes retarded, the embryos have swollen pericardia in some cases, and are pale and small. Some of them continued to live in these solutions for fifteen days but were far from the hatching stage at this time. The weaker toxicity of these mixtures when compared with the action of the salts used singly may be due as Loeb (’02) has claimed, to an anti-toxic effect of one ion on the other. The mixtures with less NH,C] were always less active. Loeb found the bivalent cations to show an anti-toxic action toward the monovalent ones. MnCl, solutions were used of the strengths =; m, 3's m, and 7m in distilled water; and 3, m, 'z m, and § m in sea-water. Eggs that were subjected to the action of such solutions responded in the following way: ‘Those in the distilled water solutions go normally for several hours, then when about eighteen or twenty hours old the blastoderm shows a dark central portion when viewed from above, while in side view it shows that the dark area protrudes downward into the yolk. A section of such a blastoderm is seen in Fig. 11. The dark area is shown to result from an unusual thickening in the center of the central periblast, cp, and the accumulation at this point of a number of the large periblast nuclei. A slight cavity, c, is not uncommon near the surface of the blasto- derm. This unusual thickening of the periblast seems to render difficult the subsequent descent of the germ-ring, and development is thus slightly retarded. When about forty hours old many embryos have their germ-rings only one-half over the yolks, and a short embryo is outlined on the embryonic shield. Many of the eggs died in these solutions. The embryos have a feeble pulse and the blood is often clotted in some of the larger vessels. When fifteen days old embryos hatched in the s'; m solutions but swam abnormally; one embryo was seen to hatch in a 75 m MnCl, solu- tion but it was entirely unable to swim. The solutions of MnCl, in sea-water formed slight precipitates and the results are thus no doubt vitiated to some extent, neverthe- less the dark central portion of the blastoderm always showed. The Influence of External Factors on Development 187 Short embryos with open blastopores were formed in many cases. The heart was weak and tubular with feeble contractions and was surrounded by a swollen pericardium. One embryo when eleven and one-half days old hatched in a 3'; m solution, but was unable to swim. In many of these embryos no heart beat could be detected and the yolks were badly shrunken. In one case a one-eyed embryo was noted, this is mentioned on account of the tendency of magnesium salts to produce such a condition, but the eye structures of chs embryo were very imperfect and no lens was present, this condition will be found to differ entirely from that described below as caused by the action of MgCl.,,. ges were eu iaaed to MgCl, solutions of the following aa ee to m, } m, 4m, and im in distilled water, and 0.238 m, 0.25 m, 0.286 m, 0.33 mand 0.5 m in sea-water; a molecular solution of MgCl,.6H.O being equivalent to about a 20.3 per cent solution. The early development in all of these solutions 1s strikingly normal considering the large death rate which occurs during these stages. The salt seems especially toxic to the early embryo. At seventy- four hours someembryos are well formed, though behind the con- trol in their development, and the blood eeculceal is slow in some while others have a quick heart action. When ten days old all are weak and smaller than the control, the blood flow is slow and spas- modic; in some embryos the circulation has ceasedand the blood is collected in the sinus and heart and appears as a red streak in front of the head. Many of the livelier embryos wave their pectoral fins. In the } m and $ m distilled water solutions many embryos hatch when about fifteen days old, though they swim abnormally on account of their bodies being twisted. ‘The sea-water solutions cause the yolks to shrink and in these the embryos are also small with sluggish circulations. Although kept alive for twenty-four days none of the eggs in the sea-water solutions would hatch. The conditions cited above are general and occurred also in a number of different salt solutions, but the condition which may now be considered seems peculiarly characteristic of the Mg salt. In the §m sea-water solutions one-eyed embryos occurred with sur- prising regularity in 50 per cent of the eggs. This experiment 188 Charles R. Stockard was repeated three times and each time it so happened that exactly one-half of the embryos had only one eye. ‘These cyclopean fish were rather abnormally shaped though they were able to twist about and wave their pectoral fins vigorously. “The other embryos were apparently normal in all particulars, the magnesium seeming not to have affected them. In sections the one-eyed condition was found to result from the union or fusion of the Anlagen of the two optic vesicles. Cases were found illustrating various degrees in this fusion, it seemed as though the optic vesicles were formed too far forward and ventral and thus their antero-ventro-median surfaces fused. This condi- tion results in one large optic vesicle which in all cases gives more or less evidence of its fused or double nature. As a rule but a single lens is formed, the size of which depends upon the size of the optic cup or more exactly upon the size of the ectodermal area influenced by the optic cup toforma lens. Thislens formation is interesting in connection with the results of the experi- mental work of Lewis (’04) and otherson the lens development in Amphibians. I (’07) have entered into a more detailed discussion of this subject elsewhere. The lens was found to show a double or fused structure in one case out of the ten embryos that were sectioned; the other portions of this eye were also more distinctly double than was usually the case. ‘This condition represents the last step in the fusion of the two eyes, slightly greater fusion would result in a single eye. With no other solution has such a condition as the above been procured, and its abundant occurrence in sea-water solutions of MgCl, strongly indicates that this one-eyed condition is character- istic of the action of such solutions on the developing Fundulus embryo. Solutions of MgCl, 75 m + NaCl} m in distilled water, and MgCl, 4 m + NaCl 4m in sea-water were tried on the eggs with the results following: The distilled water mixture produced no effect on the develop- ment, nor do such strengths of the two salts employed separately. The sea-water mixture contained twice as much MgCl, as the distilled water one. ‘The results are instructive. When eighteen T he Influence of External Factors on Development 189 hours old the blastoderms were raised up prominently on the yolks. Many eggs died during early cleavage, and altogether the eggs are decidedly abnormal. Neither of these salts acting alone would give such an effect. When forty-two hours old the yolks are shrunken and all of the embryos have a balloon-like pericardium in front of the head, Fig. 12, pc. Later, the circulation often becomes feeble. ‘This occurs also in simple MgCl, solutions. When nine days old the embryos are small and the yolks shrunken. All steps of the fusion of the two eyes into one are shown. This condition makes it certain that the magnesium of the mixture has acted upon the embryos. After fifteen days the eggs are still alive, though small and pale. Thus this double solution is more active than a simple MgCl, solution and produces magnesium effects with really less magnesium present than is necessary to give a like result when MgCl, acts alone in sea-water. It was stated above that a strength of } m MgCl, in sea-water was the weakest solution that caused the one-eyed embryo. ‘The fact that in the mixture a } m MgCl, sea-water solution gives a like effect may be due to the additional osmotic pressure exerted by the NaCl present as has been suggested by Morgan (’06), to explain similar phenomena in the action of salt solutions on frog eggs. It may also be suggested that the Mg ions act against the Ca ions of the sea-water and thus permit the Na ions to become more active, but this explanation will certainly not apply here, since the embryos show characteristic magnesium effects. Eggs were subjected to distilled water solutions of NaCl 5 m, 4m and2m. During the first day of development many died in most of these solutions. When the eggs were forty-eight hours old the 2 m solution contained many dead eggs, although the few still alive were almost normal in appearance. ‘This solution contains only 2.19 per cent NaCl which is less than the amount in normal sea-water yet it is obviously toxic to these eggs. It is evident that other salts present in the sea-water counteract this toxic effect of NaCl. When fourteen days old all of the living embryos appear normal. The 2 m solution contained one hatched embryo which had a slow pulse and feeble fin movements, it lay at rest on one side but moved if pricked with a needle. In the } m solution of NaCl 190 Charles R. Stockard more embryos had hatched than in the control though all of these fish swim with a jerky motion often moving in a spiral course or even turning somersaults in the water. ‘The salt seems to act either upon the nerves or muscle fibers of the embryo causing the nervous twitching or jumping movements. ‘The pectoral fins seem to lack their usual coordination. “This condition is not induced by the absence of some constituent of the sea-water since embryos hatched in distilled water swim normally. ‘The result is then undoubtedly due to the action of the NaCl. The embryos die within one or two days after hatching with their bodies pecul- iarly curled or twisted. Jenkinson (06) has lately recorded a simi- lar twisting and inability to swim for newly hatched tadpoles in NaCl solutions. Sea-water solutions of 2 m, 2 m and molecular concentrations of NaCl showed only a tendency to shrink the yolk. ‘The develop- ment progressed almost normally and only a few eggs died. On their shrunken yolks the embryos when six days old were small and behind the control in their development. At fourteen days the embryos hatched in the m and 3 m solutions, those in the weaker solution swam Rosia while These in the stronger showed the same jerky motions described above. On comparing these effects with those in the distilled water solutions it is reasonable to suppose that some constituent of the sea-water is capable of counteracting the effect of NaCl up to a given point’ but when an excessive amount of the salt is present its action is not entirely checked. Eggs lived for twenty-four days in a healthy condition in the molecular NaCl solution although none of them hatched. Loeb kept eggs as long as five weeks in a NaCl solution in sea-water without hatching. Embryos three days old were subjected to a double molecular 7In 1902 Loeb found that Fundulus embryos would not develop in a solution of NaCl in distilled water equivalent to the concentration of NaCl in the sea; he then added a trace of calcium salt and found development to be normal. After a number of experiments the conclusion was reached that the salts of monovalent cations with monovalent anions exert a toxic effect at certain concentrations. This toxic effect could be apnihilated through the addition of a small amount of a salt having a bivalent cation or by a still smaller amount of one having a trivalent cation. In other words, the antitorxic effects of cations vary directly as the valence of the elements. It was also found that mono-, bi-, or trivalent anions were all unable tu produce a like effect. The Influence of External Factors on Development 191 solution of NaCl in sea-water and they continued to develop in an apparently normal fashion but with their yolks shrunken. Loeb (94) had found that embryos three or four days old might be placed into a 27 per cent sea-water solution of NaCl and continue nor- mal development. None of these embryos, however, will hatch. In several of the NaCl solutions I found embryos that lacked all skin pigmentation thus appearing almost white, these were not true albinos, however, since their eyes showed pigment. Such pale embryos hatched when returned to sea-water. After a consideration of the foregoing results one must admit, it seems to me, as probable that some of the elements exert a specific stimulus on the fish embryo and cause it to develop in a character- istic manner. LiCl, KCl, MnCl, and MgCl, seem to induce rather constant and definite effects or types of embryos. ‘The form of the embryo seems to be influenced by external factors in development as well as by internal ones; in other words, the chemi- cal environment of an egg is important in determining the final resultant of the factors in inheritance. It may be suggested as a probability that every element that forms a chemical union with the germ substance produces on the developing egg through its action definite anatomical and physio- logical effects, which of course will vary in different kinds of eggs. Thus since the normal form of an animal may be altered in a def- nite way by certain chemical actions of the elements, we may assume that the specific nature of any animal is a product of the chemical composition of the egg cell from which it sprang. THE ACTION OF MIXTURES OF SALTS IN SOLUTION: THE CHEMICAL VERSUS THE OSMOTIC EFFECTS The following experiments were conducted in order to deter- mine whether or not by increasing the osmotic pressure of the solu- tion through the addition of a chemically indifferent substance, such as sugar, the chemical action of salts might be augmented. In other words, will eggs become more susceptible to the chemical action of a weak salt solution if the osmotic pressure of this solution ke increased? Morgan (’06) has performed similar experiments 192 Charles R. Stockard with frog’s eggs and concludes that in order to be effective the two solutions together must exert a higher pressure than the one pro- ducing its effect at the lower limit but less than for the other that produces its effect at a higher pressure. ‘These osmotic pressure effects are somewhat contr: dictory as I have above pointed outin mentioning Morgan’s results in which he finds the upper limit of NaCl to be about 2 per cent with a pressure of 13.61 atmospheres, while a like fatal limit for sugar was found to exert a pressure of only 8.376 atmospheres. It is also recalled that I described above a like contradiction in comparing the pressures of fatal sea-water solutions of sugar with similar solutions of MgCl,. As there stated, this contradiction is possibly due to the fact that the cane sugar in solution becomes inverted and thus the actual pressure is really double that calculated. In working with Fundulus eggs, as has been already pointed out, the experimenter has the advantage of being able to keep them alive in solutions which exert pressures both above and below that to which the eggs are normally accustomed. ‘This fact has been of especial value in analyzing the results of the following experiments. To anticipate what is to follow it may be stated that on adding certain percentages of sugar to a distilled water salt-solution, the action of the salt was increased although the total pressure of the solution was less than the osmotic pressure of ordinary sea-water. Such a result may probably be due to the action which would take place if the sugar became inverted in the solution. The following distilled water solutions of LiCl+ sugar were employed in one experiment, LiCl 0.128 m, 0.096 m, 0.064 m and 0.032 m with 0.44 m of cane sugar in each. All of these solu- tions exert an osmotic pressure less than that of sea-water, except possibly the first which has an almost equal pressure. After nine- teen hours the eggs in LiCl 0.128 m + 0.44 m sugar had polar caps with “bubbles” beneath and many were dead, those in LiCl 0.096 m + 0.44 m sugar were in about the same condition. LiCl 0.0641m + 0.44 m sugar hadalso produced polar caps and no germ- rings were formed, LiCl 0.032 m + 0.44 m sugar had caused half of the embryos in it to die, while the living ones had formed abnor- mal germ-rings. Eggs in a solution of LiCl 0.032 m are scarcely The Influence of External Factors on Development 193 if at all affected at this time, and those in 0.44 m sugar are normal. The eggs continue to show these graded abnormalities in the dif- ferent solutions and when sixty-eight hours old were as follows: All were dead in the three stronger mixtures, and a few short embryos had been formedand were still alive in the LiCl 0.032 m + 0.44 m sugar. In LiCl 0.128 m at sixty-eight hours many were dead but a good number of short embryos were present; in the LiCl 0.032 m 20 per cent of the embryos were almost normal. In the 0.44 m sugar solution the embryos were normal. The result shows that sugar augments the action of the LiCl although the pressure of the mixed solution 1s less than that in which the eggs usually live. This conclusion seems to me correct for now I realize the improbability that the sugar may have inverted which would thus have exerted twice the pressure supposed; if this were true then all of the solutions would have a pressure higher than that of the sea-water, though still not high enough in themselves to cause any of the above effects as will be readily seen by comparing the pres- sures of sea-water solutions in which the eggs develop normally. A reverse experiment was conducted in which the amount of LiCl present in the solution was constant while varying amounts of sugar were added. LiCl 0.032 m was mixed with 0.293 m, 0.44 m, 0.586 m and 0.88 m sugar, and LiCl 0.016 m with 0.293 m, 0.44 m, 0.586 m, 0.88 m and 1.253 m sugar. ‘The results of these experiments showed as one would expect from the above that the injurious action of the solutions increased with the amount of sugar present, and moreover the activity of the mixture was always stronger than that of either constituent when used alone. The last point is well illustrated by eggs of forty-eight hours in the solution of LiCl 0.032 m + sugar 0.586 m. All the eggs in this solution have the blastoderm in the form of a ball on the upper pole, only a few are still alive and in these the large bubble-like segmen- tation cavity is present. “The osmotic pressure of this mixture 1s lower than that of sea-water provided that the sugar has not inverted. At this time, forty-eight hours, eggs in 0.586 m sugar solution are all normal, and those in 0.032 m LiCl almost all have their germ-rings three-quarters of the way over the yolks with short embryos formed; some, however, have the germ-ring only one- quarter or one-third of the way down. 194 Charles R. Stockard Mixed solutions of LiCl and sugar were also prepared in sea- water. A 0.293 m solution of sugar was added to 0.336 m, 0.256 m and 0.192 m solutions of LiCl. ‘The general results agree with those described above for the distilled water solutions, although the contrast between the simple LiCl solutions, and the mixtures was not so sharp. Figs. 13 to 17 of eggs when twenty hours old serve to indicate very well the conditions caused by the solutions at this period. Fig. 13 shows the appearance of the majority of eggs in LiCl 0.256 m + sugar 0.293m. Fig. 14 shows the egg nearest normal in the same solution. Fig. 15 indicates the stage that the large majority of eggs in LiCl 0.256 without the sugar have reached at this time. A marked difference exists between this embryo and those in the mixture. Fig. 16 is the most abnor- mal one in the LiCl 0.256 m and Fig. 17 shows a control egg at this age. Eggs were subjected to the following distilled water mixtures of NH,Cl and sugar, NH,Cl 4 m, + m, and 75 m + 0.44 m sugar. The development of the eggs in these different strength mixtures was as we would expect from the result shown above. ‘Those in the NH,Cl 3 m + sugar 0.44 m were all dead within nineteen hours with their blastoderms in the form of balls of cells on ie upper pole of the egg. At this time some of those in NH,Cl + +sugar 0.44 m had the germ-ring one-quarter way down the es the majority, however, showed the blastoderms as polar balls which had not flattened down; many were dead. ‘The weakest solution produced fewer abnormalities. ‘The eggs in the 0.44 m solution of vee were normal at this time, nineteen hours, and those in NH,Cl } m had thirteen normal and seven dead. When forty-three hours old all of those in NH,Cl +m + sugar 0.44 m were dead, no embryos having been formed. ‘Those in the weak solution NH, Cl3'5 m + sugar 0.44 m were also dead at this time. Both of these solutions exert an osmotic pressure less than that of sea-water. After forty-three hours eggs in NH,Cl + m were almost normal, and their condition in NH,Cl 75 m was the same, while those in the 0.44 m sugar were well up with the control. ‘Thus again we see that the mixture exerts a far greater influence on development than either constituent acting alone is T he Influence of External Factors on Development 195 capable of producing. It appears in this instance rather illogical to state that the extra pressure induced by the addition of sugar to the solutions of NH,Cl caused this salt’s action to become more pronounced upon the eggs, for as mentioned before the pressure of these mixtures is often below the usual pressure in which the eggs live, and from the experiment cited below we shall find that the Fig. 13 An embryo when twenty hours old in LiCl 0.256 m + sugar 0.293 m, the majority of eggs in this solution are in a similar condition. sc, segmentation cavity. Fig. 14 The least affected egg in the above solution. Fig.15 The majority of the eggs in simple LiCl 0.256 m solution show this condition. Fig. 16 The most abnormal egg in the LiCl 0.256 m solution at this time. Fig. 17 Accontrol egg when twenty hours old. All X 17} diameters. addition of sugar to sea-water solutions of NH,Cl, which are of course hypertonic, furnish rather indifferent results. One might argue on the other hand that salts of the sea counteract the effects of the NH, ion, but even if this does occur the high pressure does not particularly injure the eggs, and we are still in the dark con- cerning the question why the distilled water solutions of NH,Cl 196 Charles R. Stockard act more violently in the presence of sugar unless it be due to some action which might take place when the sugar molecules split if they become inverted in the solutions. Eggs are necessarily very delicate chemical indicators and it may be that an action hitherto undetected might be shown by them. Furthermore, Fundulus eggs are exceptionally adapted to the study of such questions as they are not necessarily subjected to abnormally high pressure in experimentation. Sea-water mixtures of NH,Cl 4m, 4 m, and 7; m + sugar 0.293 m were used with rather indifferent results. In each of the three mixtures the yolks were slightly shrunken and in the two stronger a small per cent of the eggs always died during the first day of development, but from this time until nine or ten days old they developed in a normal manner though somewhat slower than the control. When fifteen daysold in NH,Cl } m + sugar 0.293 m 95 per cent of the eggs were dead and the few embryos alive were small with feeble pulse. “They appeared as embryos should when seven or eight days old. In NH,Cl 4m + sugar 0.293 m 50 per cent were dead and the others were small and otherwise like those described above. The eggs in the NH,Cl 73 + sugar 0.293 m were all normal excépt for the small size of the yolks. None had hatched. In the 0.293 m sugar solution all had a nor- mal development; and in the sea-water solutions of NH,Cl + m and 75 m development was almost normal except for the contrac- tion of the yolks. The embryos were a little retarded in develop- ment and none of them hatched. ‘These results lead also to the same general conclusion, that the mixture acts more violently than would either constituent acting alone, although the difference in action here is not great. SUMMARY AND CONCLUSIONS 1 The membrane of the eggs of Fundulus heteroclitus is readily permeable to salts in solution as is shown in embryos a few days old by the fact that KCI will stop their heart action within a few moments. During the early stages the membrane is also easily penetrated since eggs subjected to the action of strong solutions of LiCl for The Influence of External Factors on Development 197 one or two hours do not recover from the effects of this treatment after being returned to sea-water. Many other facts go to show the readiness with which this membrane is permeated. 2 Fundulus eggs develop normally, although at a somewhat faster rate, when kept on moist plates entirely out of water. The embryos developed out of water are unable to hatch while on the moist plates, but if at any time after the control has begun hatching some of the eggs are immersed in sea-water they will soon begin hatching, commencing usually in about ten minutes after pene in the water and all coming out very promptly. On hatching the embryos show a positively heliotropic and a negatively geotro- pic reaction. Embryos were kept for thirty-three days, or twenty days after the control had begun hatching, on these moist plates without beginning to hatch. ‘The fish within the egg membrane grows in length and absorbs its yolk at about the same rate as hatched ones do. They finally die of starvation after having assimilated all of their yolk, being still confined within the egg membrane. 3 Fundulus eggs are not entirely immune to osmotic effects though it has often been stated that they are. In weak cane sugar solutions the yolks were observed to swell, this has never been seen even in eggs developing in distilled water and may probably be due to some change taking place in the sugar after it has permeated the egg membrane. In concentrated sugar solutions the yolk shrinks in a somewhat definite manner. A 1.53 m distilled water solution of cane sugar killed the eggs within twenty-three hours. ‘The osmo- tic pressure of such a solution is calculated to be 34.278 atmos- pheres or about twelve atmospheres more than that of sea-water. Some salt solutions which exert even a greater pressure do not kill the eggs. ‘This contradiction might be explained if the cane sugar becomes inverted in the solutions but from the evidence at hand this interpretation seems improbable. ‘There may possibly be an action of the new substances resulting from the inversion of the cane sugar molecule which is also injurious to the eggs. Eggs hatch in 0.166 msolutions of sugar in sea-water. On com- paring the effects of sea-water solutions of sugar with distilled water solutions it was found that a pressure more than double as 198 Charles R. Stockard high in sea-water produced a much less marked effect. Such o' servations seem to indicate that the eggs were less resistant to chemicals when treated in fresh water, due possibly to a slightly weakened condition when out of their usual medium. ‘The fresh water solutions showed a strong tendency to become acid and a fungus-like growth was often present (see footnote, p. 178). It will also be recalled that the acid condition of the medium would in itself be injurious to the eggs. 4 Eggs that were subjected to the action of LiCl, LiNO, and Li,SO, were all affected in a similar manner, seeming to indicate that the cation common to the three salts was the active principle concerned. Of the large number of other salt solutions employed none of the metallic ions gave the same constant abnormalities which lithium induced. ‘The lithium larva of Fundulus is as definite and well marked as those recorded by Gurwitsch and Mor- gan for the frog. 5 a It was found, as Loeb had already shown, that this egg will develop in solutions of KCl and live for several weeks without developing a heart beat. Loeb’s statement that the circulatory system develops normally is incorrect, since the heart itself is abnor- mal, the pericardium is often greatly swollen, and other portions of the system are defective. Although eggs will liveand develop in these solutions if placed in them soon after fertilization an embryo several days old will be killed in a few moments if treated in a like manner. Thus when the heart’s action has once become estab- lished the embryo can no longer withstand the action of KCI. b The effects of NH,Cl on these eggs were rather general, development was retarded, the blastopore was slow closing and many short embryos resulted. ‘The circulation was poor. Some lived in these solutions for eighteen days though none hatched. In mixtures of NH,Cl with MnCl, eggs were less affected than in solutions of either of these salts used singly. ‘This fact may be due to the antitoxic action of one cation on another as Loeb has claimed to take place. c MnCl, solutions prepared in fresh water caused a thickening or concentration of the central periblast in early stages, develop- ment was retarded, and the embryo had a feeble pulse. Some of The Influence of External Factors on Development 199 the embryos in the weaker solutions hatched butswam abnormally. Solutions of MnCl, in sea-water induced similar effects. d Sea-water solutions of MgCl, caused the embryos to form one large single and almost terminal eye. ‘This single eye results from an early fusion of the two optic vesicles. ‘The optic cup 1s, therefore, abnormally large and the size of the lens in such eyes varies directly with the size of the optic cup. ‘This condition is to be compared with that known in human monsters as Cyclopia. MgCl, when mixed with NaCl also caused this abnormality. e Eggs that were treated with NaCl solutions showed no abnormalities during their early development. In the weaker solutions many embryos hatched but were unable to swim in a normal fashion. ‘The NaCl affects either the nerve or muscle substance of these fish causing them to swim with jerky motions, and to fall on one side when at rest. “The embryos would live for many weeks without hatching in very strong NaC] solutions. 6 Mixed solutions of salts and sugar act more violently on these eggs than either constituent would if used alone. Very small doses of a salt will give the effect of a much stronger dose, provided that sugar has been added to the solution. ‘The presence of the sugar thus seems to augment the activity of the salt. This may be due to the additional osmotic pressure that the sugar exerts, but such an explanation is not entirely satisfactory. Pathological Laboratory Cornell University Medical College New York City, December 1, 1906 LITERATURE CITED Brown, O. H., ’03—The Immunity of Fundulus Eggs and Embryos to Electrical Stimulation. Am. Jour. Physiol., ix, pp. 111-115. ‘05. The Permeability of the Membrane of the Egg of Fundulus Hetero- clitus. Am. Journ. Physiol., xiv, pp. 354-358. Garrey, W. E., ’05—The Osmotic Pressure of Sea-water and of the Blood of Marine Animals. Biol. Bull., vii, pp. 257-270. Gurwitscu, A., ’95—Ueber die Einwirkung des Lithionchlorids auf die Entwick- elung des Frosch und Kréteneier (Rana fusca und Bufo vulg.). Anat. Anz., xi, pp. 65-70. 200 Charles R. Stockard Gurwitscu, A., ’96—Ueber die formative Wirkung des veranderten chemischen Mediums auf die embryolane Entwickelung. Arch. f. Entw.-Mech., ili, pp. 219-260. Hersst, C., ’92—Experimentelle Untersuchungen tiber den Einfluss der verander- ten chemischen Zisammensetzung des umgebenden Mediums auf die Entwickelung der Thiere. I. Theil. Zeitsh. f. wissensch. Zool., iv, 3 pp. 446-518. ’93—Experimentelle Untersuchungen. II. Theil. Mittheil. aus der Zool. Station zu Neapel, xi, pp. 136-220. *96—Experimentelle Untersuchungen. IIJ,IV,Vund VI. Theil. Arch. f. Entw.-Mech., ii, pp. 455-516. Howe 1, W. H., ’06—Vagus Inhibition of the Heart in its Relation to the Inorganic Salts of the Blood. Am. Jour. Physiol. xv, pp. 280-294. Jenkinson, J. W., ’06—On the Effect of Certain Solutions upon the Development of the Frog’s Egg. Arch. f. Entw.-Mech. xxi, pp. 367-460. Lewis, W. H., ’04—Experimental Studies on the Development of the Eye in Amphibia. I. On the Origen of the Lens. Rana palustris. Am. Jour. Anat., ili, pp. 505-536. Lors, J., °92—Investigations in Physiological Moephotsee III. Experiments on Cleavage. Jour. Morph., vii, pp. 253-262. 93Ueber die Entwicklung von Fischembryonen ohne Kreislauf. Pfliiger’s Archiv, liv, pp. 525-531. °94—Ueber die relative Empfindlichkeit von Fischembryonen gegen Sauerstoffmangel und Wasserentziehung in verschiedenen Entwick- lungsstadien. Pfliiger’s Archiv, lv, pp. 530-541. ’95— Untersuchungen iiber die physiologischen Wirkungen des Sauerstoff- mangels. Pfluger’s Archiv, Ixii, pp. 249-294. *oo—On Ion-proteid Compounds and Their Réle in the Mechanics of Life Phenomena. I. The Poisonous Character of a Pure NaCl Solution. Am. Jour. Physiol., ii, pp. 327-338. °o2—The Toxic and the Antitoxic Effects of Ions as a Function of their Valency and Possibly their Electrical Charge. Am. Journ Physiol., Vi, pp. 411. ’o5—Studies in General Physiology. Univ. of Chicago Press. Matuews, A. P., ’04—The Relation between Solution Tension, Atomic Volume, and the Physiological Action of the Elements. Am. Jour. Physiol., X, pp. 290-323. Moreau, T. H., ’03—The Relation between Normal and Abnormal Development of the Embryo of the Frog, as Determined by the Effects of Lithium Chlorid in Solution. Arch. f. Entw.-Mech., xvi, pp. 691-712. °06—Experiments with Frog’s Eggs. Biol. Bull., xi, pp. 71-92. a The Influence of External Factors on Development 201 RonpeaAu-Luzeau, ’02—Action des Chlorures en Dissolution sur le Développement des ceufs de Batraciens. Théses prés. Faculté des Sci. de Paris Univ. STocKarD, C. R., ’06—The Development of Fundulus Heteroclitus in Solutions of Lithium Chlorid, with Appendix on its Development in Fresh Water. Jour. Exper. Zool., 111, pp. 99-120. ’°07—The Artificial Production of a Single Median Cyclopean Eye in the Fish Embryo by Means of Sea Water Solutions of Magnesium Chlorid. Arch. f. Entw.-Mech. xxiii, pp. 249-258. Sumner, F. B., ’06—The Physiological Effects upon Fishes of Changes in the Den- sity and Salinity of Water. Bull. U. S. Bureau Fisheries, xxv, pp. ; 53-108. MOVEMENT AND PROBLEM SOLVING IN OPHIURA BREVISPINA? BY ©. C. GLASER Witu Five Ficures INTRODUCTION The observations and experiments which I shall describe and discuss in the following pages were made in the Marine Biological Laboratory at Wood’s Hole, for the purpose of testing Preyer’s conclusion that ophiurans are intelligent animals. In spite of the fact that there is still much difference of opinion as to what we mean by intelligence, all will agree, I think, that it involves at least the ability to learn and to modify behavior in accordance with experience. Jennings (06, p. 291) has formulated in the law of the resolution of physiological states, the way in which behavior is modified in experience: “The resolution of one physiological state into another becomes easier and more rapid after it has taken place a number of times.”” I have attacked the problem of intelligence in Ophiura brevispina from the point of view afforded by this law of resolution. PROGRESSION Progression in ophiurans has been described by a number of observers, including Romanes (’85), Preyer (86), von Uexkull (05) and Grave (’00). “These writers agree as regards the general method of locomotion in ophiurans, but they have not described all of the movements which these animals perform. All of these authors have noticed two types of progression, the first of which may be visualized by the aid of Fig. 1, in which the arms are numbered, and so distinguished by heaviness of line, that the most active is the widest, the least active the narrowest. } Contributions from the Zodlogical Laboratory, University of Michigan, No. 107. THe Journat or ExPERIMENTAL ZOOLOGY, VOL. IV, NO. 2. 204 O. C. Glaser In movements conforming to type J, Fig. 1, 4, the two arms 1 and 3 are used as a pair, whose strong backward stroke drives the animal in the direction indicated by the arrow. Arm 2, which projects forward rather stiffly, serves only the function of guiding, this being also the effect of 4 and 5, which are dragged behind. A slight modification of type J, 4, 1s found in type J, B, in which the distal end of arm 2 waves from side to side, and in this manner adds to the propelling force furnished by z and 3. Type J, C,is a further modification of J, 4, in which arm 2 instead of bending only distally makes a stroke as effective as either that of z or 3, and bends either to the right or to the left, so that the animal is if Fig, I propelled by two arms on one side and one on the other. The course is zigzag if regular alternations in the direction of the stroke of arm 2 occur but if this always falls on the same side the course is circular. The movements that fall within this type are variable to an extent which has not been pointed out. J, 4, represents in its pure forms one of the two types which all previous writers have noticed, though Grave (’00) has also observed the modification B, of which C is the extreme case. Von Uexkiill (’05), who calls this type of movement Typus Unpaar voran, says: ‘Beim Bewe- _gungstypus Unpaar voran, zeigt sich welch grosse Unterschied in der Bewegungsamplitude des ersten und zweiten Gangpaares besteht. Letzteres verhalt sich beinahe passiv. Doch kann es gele- Movement and Problem Solving in Ophiura 205 gentlich auch stirker in Aktion treten. ” Both Preyer (’86) and Grave ('00) state that the “posterior” pair is dragged behind, and I have never observed more than insignificant movements in it. Type II (Fig. 2), observed by all of the writers mentioned, and called ‘Typus Unpaar hinten by von Uexkiill, may be described as two pairs of arms working synchronously, or alternately, the anterior pair initiating movement at one time, the posterior at another, or the movement may be begun by arms 2 and 4; by 1 and 3; by 2 and 1; or by 3 and 4; the only constant factor is the behavior of arm 5 which is invariably dragged behind. A third type of movement, Fig. 2, //J, not previously recorded, involves the activity of all the arms in such a manner that the animal is forced forward by three arms on one side and a pair on. the other. ‘This type may be thought of as a modification of J, C, IT Biesi2 in which arms 4 and 5 have become active, or as J, in which arm 5 has become active. Type III, is really I, C, plus an additional pair, and as in J, C, the course is zigzag if arm 2 alternates regu- larly from side to side, circular if the stroke falls always in the same direction. It is not necessary to describe the finer variations to which these types of movements are subject; to point out, as has been done in von Uexkill’s excellent paper (’05), how one may pass over into another, or how the course is affected by differences either constant or variable in the rate and strength of stroke of particular arms or particular combinations of arms. With the exception of type 206 O. C. Glaser : Unpaar voran, in which according to von Uexkill effective move- ments occur in the two arms which are usually dragged passively behind, I have observed that Ophiura brevispina moves in practic- ally all the ways in which it is possible for a pentaradiate animal of its construction to move. INDIVIDUALITY The movements described are directly dependent upon the pentaradiate symmetry, but this symmetry does not exhaust the possibilities of behavior. A little observation shows that each animal is unique at any given time and that while its movements fall within the system of classification proposed, they have pecu- liarities that distinguish them from other movements of the same type. In general the movements may be either rapid or slow, and certain individuals seem on first acquaintance to be distinctly active or distinctly sluggish. More careful study shows, however, that very sudden sina of behavior occur, and that an active, rapidly moving animal may unexpectedly enter into a state of sluggishness that sometimes lasts for hours. 1 do not understand these sudden changes. They are not due to the conditions in the aquaria; they occur with great suddenness and not in all of the animals; they are not due to either gentleness or roughness in handling because either may or may not be followed by a change in the behavior of the same individual in successive trials. Possi- bly any sortof handling may, in certain physiological states, cause a change of behavior, but what the physiological state in which this occurs is, is hard to ascertain. In certain experiments in which I encumbered the arms with rubber tubes, after the manner of Preyer (86), I frequently encountered the same sudden change from activity to passivity, and arms which were flexible and easily encumbered, would suddenly bend at their tips and stiffen, so that it was impossible to slip the tube overthem. This stiffening might take place at the first trial, or some other one, and never again, or it might reoccur upon every attempt to encumber the arm. Eeuiodie changes from activity to sluggishness also occur. Movement and Problem Solving in Ophiura 207 Thus, in June of the present year, more than half of the animals I studied were very active and quickly responsive to stimull during sometime of my acquaintance with them, but by August the whole race had changed. Perfectly fresh material brought into the laboratory in excellent condition and kept in large tanks of running sea-water, was so sluggish that I was forced to give up the experiments which I had planned for that month. None of the stimuli employed in June elicited reaction, and acids sufh- ciently concentrated to attack the skeleton, as well as the electrical current, resulted in nothing but a few spasmodic contractions with no attempt at progression or escape. What the reason for this change was is not certain. A sluggish individual almost always has very large bursal openings; in fact, it is possible to predict with considerable certainty the behavior of an individual by , examining its ventral surface. The enlarged bursal openings may be consequences of the spawning process, and the periodic change of behavior of the breeding activities. O. brevispina begins breeding in June and ends in August. Late in June many individuals have spawned, and many have the enlarged bursal open- ings; by the middle of August all have spawned (Grave "00) and most of the individuals have the enlarged bursal openings. As the genital ducts lead into the bursae—which in some species are used as brood-pouches—their enlargement may very well be due to sexual activity, which is a drain upon the animals, and undoubtedly leaves them in a state of physiological depression. If this view is correct, the enlarged bursal openings are the indices of a lethargic state following the breeding season. Rapidity and sluggishness of movement have consequences of great importance in problem solving. Sluggish animals not only make fewer movements and take more time to perform them than active individuals, but they use in general fewer arms; their move- ments are less varied, and the arms very rarely come into contact with one another or cross. All this is very different with active individuals; their movements are quick and varied; they use relatively more arms, often move these through greater arcs than the sluggish animals; and, in addition, the arms touch and cross with great frequency. How “contacts” and “crosses”’ are related 208 O. C. Glaser to activity and sluggishness 1 is easy to see. An active individual using four arms in progression has a much greater opportunity to make ‘‘crosses”’ and “contacts” than if fewer arms were used. Very often when the animals move by means of two pairs of arms, the anterior pair is crossed by the posterior regularly. “The same frequently happens when only three arms are used. Contacts and crosses also depend on the length of the arms, as the chances that they will occur in long armed individuals are ereater than in short. How important arm length is, is indicated in the following table in which are summarized observations on three individuals which were active, but differed in the lengths of their arms and also in the manner of using them. ‘The effect of the latter factor emphasizes that of the former. The longest armed individual 4 used the “two pair of arms” stroke, only once in a total of 141 effective backward strokes, whereas the shorter armed individuals B and C, used this stroke eight times in 129 and four times in 126, respectively. TABLE I | | No. of arms moved Individual |No. of Movements ee ee | Contacts Crosses | | BSW ZOOS 2 desl e§ A 141 | 13 119 1 | 26 | aS || i | o | times moved | | B 129 Il 16 o| 29 | 13] 8] o | times moved Cc 126 6 2, | o | 16 | 26| 4 | o | times moved RIGHTING MOVEMENTS Two types of righting movements were observed, only the first of which has been described by von Uexkiull (05) in an excellent paper illustrated by means of kinetoscope photographs, and by Grave ('00), who says: “Two adjacent arms straighten out so that together they form a straight line. On these arms as an axis the body revolves, being pushed over by the three remaining arms, but mostly by the median one of the three.” This description, correct as far as it goes, is incomplete. At the bases of the straightened arms, and in the interradial portion of Movement and Problem Solving in Ophiura 209 the disc between them, movements occur whose effect is to bend the ventral surface in the direction indicated by the arrows. When this process, by which a small portion of the ventral surface is brought into the normal position (Fig. 3, 4), has proceeded far enough, the animal is righted suddenly by its own weight, since while the process de- scribed has been going on, arms 3, 4 and 5 have so ele- vated the dorsal surface of the disc that this falls into the normal position. In the second type of right- ing movement, Fig. 4, arm 2 curves near its base, and bends under the disc which, as in the previous case, is elevated by the otherarms, particularly by 4 opposite 2. ‘The disc thus rotates on the base of 2 as a pivot, and after it has been sufficiently elevated, the animal falls into the righted position of its own weight. The length of time required to execute the righting reaction was measured on eight individuals. I have summarized these results in Table II, in which are given the average time for each individual, as well as the maximum and mini- mum consumed. (See Table IL.) These averages of course do not show the differences between the suc- cessive individual rightings of any of the animals used. These differences Were in some instances very large, and have had a great effect on the averages. (See able LL.) ‘These measurements show that the variations from the mean may be very great; that because an individual has Fie. 4 righted itself very quickly a number of times is no reason for believing that it will continue to do so. in spite of those cases in which righting took place slowly, the 1c) (CH O. C. Glaser 2® Hg. = he | «@ atl w& gb | ge 68 | ue 4881 | uv eM, | ie a | We of "UN, “XBL | UT, “xe | UNL “XPIAL | UNIAL “XPT | UNA “xe | UNA,“ XE YL | UAL “XI “UNIT XP, : 6°L hte or o1 67 09 IZ 09'S og ft | Ol “sy +9 of ob'+ o1'S of or o1'S Outer || ob'9 09'S o1'S or *sIy QI gz yog yooh Woxaaey yooh o1'h | 09°01 og" t eroea ol “say & Lt wrk | o0$ Sz Lew | ogee | 8 AS lz | ary | jos ¥1 ayer | 429° t 8 P) Lzaunf{ H 3) d conse aie ye 3) a Vv sje, “ON | S[eAquE | aIeG sjputup yyd1a fo sauiij-BunYysis adviap Il @TaVL QUT Movement and Problem Solving in Ophiura TABLE III Individual righting-times of eight animals in many successful trials mt oma Sorter wn 19 19 a on owreoe + + 5 3 5 i 4 14 minto t+ + Oy cy ANMN NAA FH int +m oO on p fe) a) 31 12 35 17 been aaa arto t+ ub 9 ub) 20 2) De st NO) SO ce) -_ wm 12 27 41 39 212 O. C. Glaser records when averaged show that these animals, on the whole, may be expected to right themselves in less than 45 seconds. One fact of considerable interest 1s clearly demonstrated by the averages as well as by the individual records—there 1s no reduction in the amount of time required to perform the righting act; in other words, under normal conditions, these animals do not im- prove by practice in the execution of their righting movements. PROBLEM SOLVING The expression “problem solving” is almost self-explanatory. Under this heading, I have placed such behavior as an ophiuran exhibited when stimulated by interference more or less unusual, and from which it was able sooner or later to escape. What I did was to observe the way in which the escape was made—the prob- lem solyed—and how much time was consumed in doing tt. The problem—the same asthat employed by Preyer—was to rid one or more arms of the small pieces of loosely fitting rubber tub- ing with which I encumberedthem. Inthe selection of individuals for experiment, my choice was guided by two considerations: whether all the arms were approxi- mately equal in length, neither broken, nor recently regenerated; and whether the individuals were not too active to make the obser- vations easy to record. When encumbered in the man- ner represented in Fig. 5, an ophiuran does many things, some of which are recorded in von Uexkiull’s photographs. At first it may pass through a brief latent period, during which it lies motion- less on the bottom of the dish, and then it may crawl, dragging the encumbered arm behind it. Often the animal moves at an angle to the encumbered arm, orin rare cases inthe direction of it. The / Fic. 5 Movement and Problem Solving in Ophiura 212 progression may be of a very violent character involving many contacts and much crossing of arms, or the animal may simply writhe, without changing its location. If it does not move about, it usually waves one of its arms, especially the encumbered, in a horizontal plane, though the movements may also occur in a vertical plane and in circles. The encumbered arm is moved in a vertical plane oftener than the unencumbered ones; is frequently rubbed against the disc; against the adjacent arms; against the sides of the dish; and even against itself. Sometimes the encum- bered arm is waved over the disc, much as a man waves a long whip, and then is “‘cracked,’’ so that the encumbering tube moves nearer the distal end, and often slides off. When relieved the animal usually does not remain quiet, but continues its move- ments for a short time and makes several strokes that remove it from the place where the tube was gotten rid of. If at the instant of riddance the animal was not progressing, a short journey is begun at the moment of relief. When encumbered on more than one arm, the latent period is longer than when only one arm is encumbered; the first move- ments are not through as great arcs, nor are they so long continued in any direction. One movement is succeeded rapidly, not by its duplicate, but by another in a different direction, and this by still another. The behavior changes constantly. If all of the arms are encumbered, the above changes in behavior cease very soon, and an entirely different kind of action is begun. Instead of movements in the usual sweeping manner, the arms quiver andtremble. In one case, one arm (the first to be rid of its rubber tube) in particular attracted my attention by quivering when the rest of the animal was perfectly quiet. “These quivering movements occur in a horizontal plane, and are so rapid, and many of them so slight, that it is impossible to record them accu- rately without special apparatus. Of all these movements, several are more effective than many others in bringing about riddance. The most effective are the “whip movement,” the “stripping movements;” certain of the “wavings,” and violent progression which involves a number of different movements. Of these the whip movement is the rarest; 214 O. C. Glaser the violent progression next, whereas the strippings and the wav- ings are the commonest of all. These observations open two ways in which the problem of resolution may be attacked; by studying the time taken to solve the problem and by noting the relative frequency of the most effec- tive movements. ‘The time and the frequency might both remain constant, or might change, or only one might change. As a reduc- tion in the amount of time taken to solve the problem need not necessarily be due to an increase in the relative frequency of the most effective strokes, these two must be considered separately, although an increased frequency of strokes best fitted to solve the problem would involve a reduction in the amount of time. Ifa reduction in the amount of time required does occur, it means that the physiological state produced by the rubber tube has been resolved into the normal state more rapidly than it was resolved the first time. In other words, the animal has learned by experience. The following Table IV contains my measurements of problem solving time. In every case the animal was given the same prob- lem consecutively, viz: the same rubber tube was placed on the same arm, under the same conditions. As little time as possible was lost between trials. TABLE IV* Trials Individual....... I 2 3 | 4 5 6 7 A Bu AS 6) | 25400" eal 20a, | 3/ 00” 6’ co” B o/ 3Q” 4030 |) oo 1 4/ 00” c o 45” ee sh Me | Tid(croyt 1’ 30” | D 2/ 00” 3/ 00” 7o20 ain NI aaoos 3/ 00” Zale | Nos! E Enger eae HSS Hl GX PARR OTP AS, ESra Pel area *These measurements include the latent periods. The number of trials recorded in Table IV is small. I was prevented from eollecting more data by the sudden changes of behavior before alluded to. Other animals were tried but failed to react regularly even five times. The results as they stand, how- ever, are worthy of confidence; they are representative of the whole behavior which is varied and uneven; like the measurements of Movement and Problem Solving in Ophiura 215 righting time they neither increase nor decrease—the apparent increase being due to the failure to respond, for had this failure occurred sooner, some of the last measurements would have been smaller than the first. Fatigue played no part in the result, as the figures are too uneven. The objection might be advanced that these cases which I have called “problems,” were not such; that there was no reason why the animals should modify their behavior, and that what they did under the conditions of the experiment was nothing that they would not have done under normal conditions. ‘This objection 1s met satisfactorily I believe by the following experiments. A given arm was stimulated by encumbering it with a rubber tube, or by painting it with strong or dilute formalin or hydro- chloric acid of different strengths. ‘These trials, of which I made a great many, yielded very definite results. In only one case did an animal progress in the direction of the stimulated arm; in a few cases at an angle to it, using it as one of the propellers, whereas in the vast majority of cases it moved in the direction diametrically opposite the stimulated arm. If the stimulus was strong, the movements were very violent, but no difference in direction was noted in the case of weak and strong stimuli. Under ordinary circumstances it is impossible to predict the direction in which an ophiuran, all of whose arms are of the same size, will move, but if one of the arms be encumbered the prediction that the animal will move away from the stimulus will be verified in the vast majority of cases. I think it is justifiable to assert that the direction of pro- gression has been determined in these cases, and if this is true there is a determining cause—a problem. My second line of inquiry—whether encumbered animals showed a noticeable increase in the number of movements best adapted to solve the particular problem given, was begun by find- ing the percentage of crosses and contacts in the same animals under the two conditions stated. The results are summarized in Table V. As contacts and crosses usually result from wavings | counted these in animal JJ unencumbered and with one arm encumbered. The results are summarized in Table VI 216 O. C. Glaser The general conclusion to be drawn from these experiments is that there is neither a decrease in the amount of time taken to solve the problem, nor an increase in the relative frequency of movements best fitted to solve it. In other words, the animals did not modify their behavior in accordance with the law of resolution, and consequently, so far as is objectively recognizable, learned nothing. TABLE V Animal | Arms Encumbered | Movements Per cent Contacts |Per cent erase Problems oe | | ) I ° | 231 4.0 2.0 | ° I I 202 3-9 1.9 2 II ° | 267 13.8 10.8 ° II 5 | 553 | 9.2 ite 2 5 TABLE VI cea ae oe nesare ee ees = : 22 Animal | Arms Encumbered Movements Per cent Wavings | 7.) ae II ° 117 | 26.4 II I 192 | 19.3 = ee is DISCUSSION The facts which I have brought forward in the foregoing pages agree with those of Preyer and von Uexkill in showing that in problem solving the animal repeatedly changes its behavior, not persisting in a certain reaction when that is unsuccessful. If I venture to take issue with Preyer, and to assert that the behavior which both he and I observed does not warrant the conclusion that ophiurans are intelligent, I must rest my claim upon the validity of my interpretation of the facts, and this validity I shall now attempt to establish. The behaviorof Ophiura brevispina may be summarized by say- ing that this animal under normal conditions performs practically all the movements possible to a creature constructed as it is; that except for this limitation, its ordinary behavior is not predictable, Movement and Problem Solving in Ophiura Dig and that even the righting movements, because of their variety occupy a place between the ordinary behavior and reflex behavior, for though more definite than the former, they are less precise than those highly perfected types of response which gave us our first idea of reflex action. Regarding the manner in which ophiurans rid their arms of encumbrances, Preyer (’86, p. 125) says: ‘‘Aus den beschriebenen und ahnlich leicht zu variirenden Versuchen ergiebt sich zunachst, dass Ophiuren in 5-fach. verschiedener Weise sich gegen die beim Tasten und kriechen thnen sehr hinderliche Bekleidung mit einem Schlauche vertheidigen: (1) streifen sie ihn ab durch Reibung am Boden wenn er locker ist, (2) schleudern sie ihn fort durch geissel. formiges Hin und Herwerfen, (3) drucken sie ihn fest gegen den Boden mit dem freien Nachbararm, und ziehen den Arm aus dem dadurch fixirten Rohre heraus, (4) stemmen sie abwechselnd beide Nachbararme mit deren Zahnchen unten gegen dasselbe und schieben ihn ruckweise ab, (5) brechen sie durch Selbst- amputation den Arm mit der unbequemen Bekleidung ab. Hilft dass eine Verfahren nicht, dann wird das andere angewendet. Sehe ich hier von dem letzten, der Autotomie, ab, von der noch die Rede sein wird, so beweist schon die 4-fache Art der Abwehr bei einem und demselben Individuum unter denselben atisseren Verhaltnissen, dass hier kein einfacher Reflex vorliegt. Vielmehr besitzen die Ophiuren die Fahigkeit sich ganz neuen, von ihnen noch niemals erlebten Situationen schnell anzupassen.”’ “Wenn Intelligenz auf dem Vermogen beruht, Erfahrungen zu machen, d. h. zu lernen, und das Erlernte in neuer Weise zweck- massig zu verwerthen, so miissen also die Ophiuren sehr intelligent : 3) sein. Preyer’s reasoning seems to be this: When encumbered on its arms the animal moves in different ways; failing to free its arms by these movements, it moves in other ways, and continues to change its movements until the encumbrances have been removed. The animal thus exhibits the process of discovery by elimination, learning in other words, and is therefore intelligent. If this indeed be learning, then all movements which any organism may under any circumstances execute are outward signs 218 O. C. Glaser of the process, for movements are never without cause, and the stimulus is aggravated, alleviated, or unchanged by them. What- ever be the result of the movement, the animal “learns”? what has been the effect upon the stimulus, the cause of the movement. Two criticisms may be made of this point of view: In the first place, in behavior such as that of an ophiuran, movements which fail to solve a specific problem, or to contribute anything whatever to its solution, are often repeated immediately. If the animal learned anything from them, it forgot what it learned at the instant of learning, for the intervals between two successive move- ments which fail for the same reason mzy be less than one second; to forget as rapidly as to learn, can be objectively recognized as neither. In the second place, in ophiurans at least, it is the excep- tion for an animal to perform only one movement at a time. Usually a considerable number, four, five, or six distinct move- ments are performed synchronously. All of these, on the assump- tion I am criticising, result in learning, but the knowledge which they give may be of two sorts; some of the movements may tell the animal how to solve the problem, the others, how it cannot be solved. It is impossible for me to believe, without striking evi- dence to the contrary, that an ophiuran can learn at the same instant half a dozen facts, belonging some to one, some to the other of two distinct categories. If the idea that mere movement in various directions is a sign of learning, involves the serious difficulties which it seems to me to involve, we have nothing but behavior more or less permanently modified as the result of experience to fall back upon. I have shown that under ordinary circumstances Ophiura brevispina does not improve with practice, in its righting behavior, and in problem solving it shows no greater aptitude. Iam, therefore, forced to the conclusion that neither intelligence nor even learning have as yet been demonstrated in this animal. My experience with ophiurans also leads me to the conclusion that resolution will be very difficult to demonstrate, not only because of those sudden changes in behavior for which it is difficult to assign causes, but also because of the remarkable “action sys- tem” exhibted by these animals. This action system shows better Movement and Problem Solving in Ophiura 219 than many others, that behavior is structure in motion, and that complexity of behavior depends on the complexity of that which behaves. An act performed by one arm may also be performed by any of the others. “The arms may all do the same thing at the same time; some may do one thing and others another; and finally a single arm may execute different movements at different levels. As the disc itself may also execute varied movements, the number of possibilities is enormous. With this marked versatility to contend with, it is not surprising that resolution, demonstrated , according to Jennings (’04, ’05, ’06) for Protozoa, Ccelenterates, and other forms lower in the scale of complexity than echinoderms, or as low, should remain undemonstrated for ophiurans. The number of movements possible to an ophiuran is immense; if the animal only acts, the chances that it will perform movements fitted to relieve a certain physiological state are better than the chances that such will be the case in most other animals. If one of the many movements that will serve is not performed, another will be, and we should not expect to find resolution, unless the fit things to be done are few. Any of the problems presented might have been solved in a variety of ways. One or more of these ways were superior to any of the others, but all served the purpose. Where the variety of solutions to a problem is great, there 1s no need of resolution, and it does not occur. I have profited much by the elaborate criticism which Professor Jennings made of an earlier draft of this paper, and I take this occasion to thank him for his kindness. University of Michigan Ann Arbor, Mich. February 1, 1907 220 O. C. Glaser LITERATURE “CITED: Romanes, G. J., ’85—Jelly-Fish, Star-Fish and Sea Urchins. Kegan Paul, Trench & Co., London. 1885. Preyer, W., ’86—Ueber die Bewegungen der Seesterne. Mitth. a.d. Zool. Stat. z. Neapel. Bd. vii. von UEXKULL, J., ’°05—Studien uber den Tonus II. Zeitsch. f. Biologie. Bd. xlvi. Grave, C., °0o—Ophiura brevispina. Mem. Biol. Lab., Johns Hopkins Univer- sity, iv. 5. Jenninos, H. S., ’04—Contributions to the Study of the Behavior of Lower Organ- isms. Carnegie Institution, Publication 16. °05—Modifability in Behavior. Journ. Exp. Zodl., vol. 11. °06—Behavior of the Lower Organisms. Columbia Univ., Biol. Series x. OCCURRENCE OF A SPORT IN MELASOMA (LINA) SCRIPEA AND ITS BEHAVIOR IN HEREDITY BY ISABEL McCRACKEN Laboratory of Entomology and Bionomics, Stanford University Wirth ONE PLate During the year 1904, early in the breeding season of the chryso- melid beetle, Melasoma scripta,! about 1000 pupz, and larve, in advanced stage, were collected from willows in the neighbor- hood of an artificial lake near Stanford University. Such of these as were not parasitized matured during the latter part of April and early May. ‘The adults represented the dichro- matic extremes of the species, the elytra being either spotted- brown (referred to in this, as in previous papers, as ““S’’), or black (referred to as “B”’), the thorax in each case having a central black area widely emarginated with brick red. In the center of each red area and nearly adjacent to the central black region (sometimes approximating it) is a small black spot representing a single punctation. (Figs. 1 and 2.) During the course of breeding through four generations from this collected material there occurred a number (four or five) wholly black individuals (Fig. 3), thorax as well as wing cov- ers being totally black (referred to in this paper as ““AB”’). Since during the casual outdoor observations made throughout that TA description of this beetle is given in the Journal of Experimental Zodlogy, 1905, vol. ii, pp. 117, 136, and vol. iii, pp. 320-336, where it is called Lina lapponica. It seems that this identification made for me is not correct. The beetle is evidently the one figured by Riley under the name Plagiodera scripta (Fabr_) Ann. Rept. Agric. for 1884, pp. 336-340, pl. viii, Figs. 1 and 2; by Lintner, under the name Lina scripta (Fabr.), Rept. N. Y. State Entomologist for 1895, pp. 181-189; and by Felt as Melasoma scripta (Fabr.) in N. Y. State Museum, Memorr 8, vol. i, pp. 317-322, Pl. 16, Figs. 16-20. Tue Journat or ExperiMeNTAL ZOOLOGY, VOL. IV, No. 2. 293 Tsabel McCracken season no such freaks were found, and those bred in the laboratory failed to mate, they were looked upon as representing possibly a pathological condition. However, in .1905 outdoor scriptas were kept under constant surveillance throughout the breeding season. At stated periods, four or five weeks apart, several hours were spent in the field, at which time several hundred individuals passed under inspection with the following result: First inspection, March 4. Thousands of beetles in a limited area feeding and beginning to breed. (These were in all proba- bility the hibernated individuals from the previous year.) No “all black”? (AB) individuals were observed. Several hundred individuals, representing each of the dichromatic extremes, were collected at this time for indoor controlled breeding. April 12: Many thousands of beetles observed; two AB females collected. May 14: Many thousands observed, two AB females collected. June 21: Many thousands observed, two AB males collected. July 28: Individuals in this particular feeding ground becom- ing noticeably fewer. Many hundreds of beetles observed, one AB female collected. August 21: Many hundreds of beetles observed, one AB male collected. Hence a total of five females and three males were collected in this locality during the five months the locality was under observa- tion and covering the breeding season of the beetle. “Three similar sports were collected during this time from poplar trees a half mile or so distant from this locality. During the progress of these outdoor inspections, indoor breed- ing was in progress from the collection of March 4, 1905, that is, a collection made up of “spotted” (S) (Fig. 1) and “black” (B) (Fig. 2) but no “all black” (AB). Four generations were reared to maturity from this collection. The following table gives the data in regard to the occurrence of the sport AB with the character of the lineage in each genera- tion. (The term “sport” is here used in the sense of a singular and decided variance from the normal type.) Sport in Melasoma and its Behavior in Heredity 223 There occurred, therefore, in the breeding room, a total of 20 AB, or sport individuals in a total of 11,369 individuals reared during the breeding season, most of these coming from the imme- diate collection of March 4. Inspection of the table shows that in the first generation, 168 matings were made, one brood only being reared from each pair. Parentage was represented by both SxS and B xB matings. The sport AB was found in the progeny of each series, ten in the former, six in the latter. In other generations matings were made in the S line only. In the second and fourth generations, single broods only were reared from each pair, as in the first generation, TABLE I Seb eGr iP | No. | No. | | | as ss | Grand- as Grand- Parents. Matings ‘broods Total No. AB | *0ta | parents. Recpes | Barents. made. reared. tnd ote | ior Ist gen. Sxs 11g | 119 hes 2\ é | | I BXB 49 | 49 | 5034 | 3 5{ 2d gen. | | SxS Ses 45 | 45 1050 | I oO I 3d gen. SxS SXSi| Ss 32 | 180 4730S. | 3 4th gen. SxS SxS SxS SxS 19 19 | Fyn) |@) 8) ° RG Eal er cariate wycrayn et echo hed eeu eer one vas 264 | AT 2a nes GON || 20 while in the third generation, an average of five or six broods were reared from each pair (a minimum of two, a maximum of fifteen broods). With three exceptions, not more than one AB individual occurred in any one brood. In the first exception two individ- uals (a male and a female), occurred in a single brood, in the second exception five males occurred similarly, and in the third exception two males. We find, therefore, in a total of 264 matings, fourteen only producing sports. That the occurrence of the sport was normal, that is, was not due to laboratory conditions, is evidenced by the fact that but com- paratively few sports occurred (20 out of a total of 11,369 indi- viduals.) Since all broods in the breeding room are under prac- tically the same conditions, had an extrinsic influence been at work 224 Isabel McCracken tending to produce this variant, certainly more individuals would have shown the effect. The numerical results in second, third and fourth generations of the occurrence of the sport might have been different had con- tinued breeding of the progeny of the first B parents been carried on and if a larger number of broods had been reared from each pair. There was found to be no greater likelihood of the recurrence of a sport in a brood from parents ingwhose lineage sports had pre- viously occurred than in broods from parents in whose lineage no such sports were known to have occurred in so far as this point was tested. Numerous matings were made between individuals having AB sisters or brothers, but in no case was there a recur- rence of the AB character. That the sport is of occasional occurrence in the field has already been noted. The stable character of many sports or aberrant Variations in animals such as this appears to be is undoubted, since it has been shown several times, notably in the race of Ancon sheep from a single short-legged, long-backed ram, in 1791, and the production of the Mauchamp-merino breed of sheep in 1828 from a single ram with long, smooth, silky wool.’ Also more recently the production of a race of polled Hereford cattle in Kansas in 1889 from a single polled bull.’ The main purpose of the present investigation was to deter- mine, in case the new character should be found to be stable, its hereditary value in relation to the characters of the parent species. The hereditary value of each of the dichromatic extremes with relation to the alternative extreme had been previously deter- mined.! To test the hereditary value of the Melasoma sport, the first two sports occurring simultaneously in the first generation were used as parents for succeeding generations. Other sports were mated with the parent forms. 2Darwin, 1868, Animals and Plants, vol. 1, p. 126. 3Guthrie, 1906, Proc. Amer. Breeders Assoc., vol. 11, p. 93. 4McCracken, 1906, Inheritance of Dichromatism in Lina lapponica. The Journal of Experimental Zodlogy, vol. ii, pp. 320, 336. Sport in Melasoma and its Behavior in Heredity 225 In the first matings of AB x AB, the male parent was of S par- entage, the female of B parentage. Five broods were obtained from this mating, a total of 130 individuals. Of these 77 indi- viduals were similar to the immediate parents, that is, were AB in character, and 42 individuals were similar to the grandparents on the female side, that is, were B in character. In eleven indi- viduals the wing covers were black, while the character of the thorax was a mosaic of the thorax of AB and B (Plate, Fig. 4); that is, the emarginate area was in part black and in part red. The black blotch upon the red might be in any position, covering the anterior half, the posterior half or the median two-thirds, as indicated in the figure. We find here, therefore, a series of Variations arising in the progeny of a mating between two similar sports or aberrant variations, so that if the latter had not arisen, frst, it would have been considered but an “extreme variation”’ in a series. No such mosaic was observed in field collections. There was no recurrence of the “S” type. The sport or parent character predominates somewhat in the offspring over the B type (1.8 : 1). The mosaic type appears to be a heterozygous form as in mat- ings of this type (I) the offspring always revert to the AB or B type with only an occasional I individual. With the 77 AB, the 42 B, and the 11 [ individuals thus obtained, six categories of matings were made as shown in the following diagrams. ‘These diagrams show the pedigree for three genera- tions and the character of the brood produced in each. DIAGRAM I—Mating Category a First Matings i x | Bix First Generation S, B and AB sport B and AB sport Second Generation AB, B and I (7:3.3:1) Third Generation AB ABand B (12:1) AB and I (6.9:1) AB, Band I. 226 Isabel McCracken DIAGRAM II—Mating Category b First Matings S xX ‘ i xb | BxB First Generation S, B and AB (sport) B and AB (sport) | Second Generation AB, BI (mixed brood) B (all) Third Generation B (all) DIAGRAM IiI—Mating Category c First Matings Ss xs BxB SxS S x's First Generation S, Band AB (sport) Band AB (sport) S (all) S (all) Second Generation AB, B and I (mixed) S (all) Third Generation S (all) DIAGRAM IV—Mating Category d First Matings Soxcus BxB First Generation S, B and AB (sport) B and AB (sport) Second Generation AB, B and I (mixed broods) Third Generation B (all) B and AB (mixed broods 3.6:1) Sport in Melasoma and its Behavior in Heredity 227 DIAGRAM V—Mating Category e First Matings : x ' 1 x ‘| r i First Generation 55 A ae (sport) plas (sport) BX i Second Generation AB, ae I (mixed) | (all) Third Generation B (all) Band AB (3.6:1) B,AB and I (mixed) DIAGRAM VI—Mating Category f First Matings S xX | me S, B and AB (sport) B iP AB (sport) AB, B and B and AB (3:1) B and AB (with an occasional I) _ The following table gives the results in total as to the character of the individuals reared in the third generation in the different categories, from the first matings, the lineage for which is given in these diagrams. In the papers previously referred to on “heredity of dichromat- ism,” it was shown that the color type S is dominant over the color type B, not at first completely so 1n every Case, but showing from generation to generation an increasing prepotency in that direction. B was considered recessive in that it disappeared wholly or partially in the first generation of a cross between S and B, but bred true from the time of its reappearance in the sec- ond generation. Because B in its ontogeny passes through the S stage, it was suggested (’05) that S represents possibly the older, B the newer type. AB in its ontogeny passes through the S and B stages. Its normal infrequency of occurrence makes it appear 228 Isabel McCracken to be either the newest type or an atavistic form. Upon the assumption that it is a new type, we would expect its behavior to S and to Bto parallel that of BtoS. Our expectation isin the main fulfilled. Inspection of Table II shows that in the progeny of AB xB (category b) all the offspring are B. In the progeny of TABLE II (Third Generation) Ratio of Mating Tot.No. Char.of | Tot.No. Tot. Tot. Tot. Parents AB.B or Category Broods Broods Indy. AB B I z AB :I a ABXAB 61 SRorsoeeee ° 1B) cigcodoc ° AB (all)..24 586 AB& B.. 6 133 123 | 10 12: 1 AB&I .27 651 569 $2) | 5Os9:)an AB,B&I 4 105 56) 30 |. 189) 5i6:9encs b ABXB 5 Byiversisierie 5 | 135(allB)| © fe) c ABXS 17 Seite 17 | 448(allS) oo fo) ° d BXB 16 Seasons ° (both parents Be (Gall) peer 52 ON) 52 from AB) AB&B .14 498 108 390 © | 229.6 e BXB 39 Stace s: ° (one parent only B (all)...14 381 381 from AB AB&B .23 62 139 490 1-13.06 AB,B&I 1 17 I 15 I Bisdieace 30 29 I if IxI 4 Sine fe) (doth parents AB(all) ..0 from AB.) B(all) ... 0 AB&B . 3 80 20 ~©60 13 AB, B&I 1 22 Aas oy AB xS (category c) all the offspring are S. That is, both S and B dominate AB completely in crosses between S and AB or B and AB. In these two categories AB is completely recessive, in the sense that it does not appear in the soma of any of the offspring. In BxB matings, extracted B (category d, Table II), both parents having been born of AB parents, two kinds of broods are Sport in Melasoma and its Behavior in Heredity 229 produced; that is, broods in which all the individuals are B and mixed broods of B and AB, individuals of B character predomi- nating. In these mixed broods the proportion of B to AB is 3.6 to 1. This parallels the history of the behavior of S toward B in matings of dominant S hybrids as previously determined. In B xB matings (category e, Table II), in which one of the parents was born of AB parents, two kinds of broods also appear; that is, broods in which all the individuals are B, and mixed broods of B and AB, or B, AB and I or B and I, the character B predomi- nating. It is noticeable that in this case, in which one-half as many AB ancestors are represented, the proportion of wholly B broods exceed that of the former cross by about 3 to 1, while the proportion of B to AB in mixed broods remains the same, 3.6 to 1. In neither case were any broods produced that were wholly AB in character. In I xI matings (category f, Table Il) mixed broods result in which B is the predominating character, AB taking a second rank and | recurring but rarely, thus showing the | emer atns char- acter of I. The hereditary value of the character B in the first generation of a cross between B and AB (that 1s, B not previously contami- nated by AB) appears from these data to be, therefore, equiva- lent to the hereditary value of pure S to the character B. The hereditary value of pure S and pure B with reference to AB have the same equivalency. The stability of the sport character AB is absolute in neither first nor second generation matings, but comparison of data of first generation mating (Diagram I, category a) with Table II, a, shows an increased preponderance of the sport character in the latter case. AB in the second generation breeds true in more than two-thirds of all broods produced and in all mixed broods greatly predominates in the offspring. It is again noticeable that no S$ individuals appear in the offspring, pleenets reference to Diagram I shows that as many S as B individuals are represented in the ancestry and reference to Table I shows AB appearing in broods of SxS parentage as well as those of B x B parentage. ‘This sug- gests that the type AB is after all but a new phase of the type B 230 ~ Isabel McCracken and that its appearance in the progeny of S xS is coincident with the appearance of B in the progeny of SxS. _ Its failure to trans- mit § while transmitting B is in harmony with this interpretation. The character AB therefore answers to the requirements of Mendelian recessiveness with respect to its behavior toward the alternate characters, S and B; that is, it disappears in a first cross. Since the matings in category a represent merely second generation matings from the sport AB, the data here cannot be consistently compared with the data from matings of true recessives. Beha- vior of AB as a recessive was tested later in the season with extracted AB from hybrids 8. For fourth generation data, matings were made in the following categories: Category a. AB xAB, both parents of AB ancestry, for two gene- rations; that is,second generation from sport parents (Diagram I). Category b. BB, both parents being wholly of AB ancestry, for two generations (Diagram I). Category c. SxS, both parents being hybrids from S xAB (Diagram III). Category d. S xAB, the S parent being of pure S ancestry for two generations, the other of AB ancestry for two generations. The following table gives the results of these matings : TABLE III (Fourth Generation) | | Tot. 1 | Ratio | Ratio Mating Grand- | Tot.No.| Character of | __ Pots Lota | oe Cat ; | Parents Ressa noes | No. AB B S Of >|) 10k ren | | Br oods fey rag a | ish ie Indiv eet B: ABS :AB | a ABXAB |AB*¥XAB*/ 18 (AB(all)..... 16] 442 | 442 B& AB ... 2, 76 | 58 18 | BS) , | b ABXAB | B¥XB* if | B& AB >. .1n| 327 |" 57*| 279 eu | | c ABXS — S*xS* 7 | AB&S.... 3)\ : 25 | 47 | 208 IdI | AB,B&S.. 4 Me) |} 83 | 10:32 d | ABXAB) |, SXABS |) 013° /)SGI) 04.2099)" 96 SXSz) | | —___ | e * Ancestry of each parent is similar. Sport in Melasoma and its Behavior in Heredity 231 It is noticeable that the mosaic I noted in the previous genera- tion (Table IT) failed to recur. Table III, category a, shows AB again not breeding wholly true, but predominating more extensively than in the previous genera- tion; that is, there is relatively a larger number of wholly AB broods. Its preponderance in mixed broods appears to be not so great, though fewer mixed broods occurred. Here, again, asin the previous generation, we find an absence of the S character in the offspring. Extracted B, Table III, category b, that.is B born of AB par- ents, produces broods with the B type predominating (5 : 1), and in greater proportion than in the previous generation (Table II, category d). “The reverse might reasonably have been looked for since in the latter case there were but two generations of AB parents, whilst in the former there are three generations of AB parents. It is in line, however, with the normal prepotency of B over AB. Hybrid S (Table III, category c); that is, S from S XAB, a cross that had produced only S offspring when mated with a simi- larly produced individual, produced mixed broods of two kinds; that is, broods of S and B, the S character predominating, or broods of S, B and AB, the S character again predominating, AB taking second rank and B occurring as a minority. It 1s possible from the known behavior of S and B with relation to each other, that B from hybrid S is but an offshoot from S and therefore must in this connection be considered with relation to S only. Category d shows pure S (pure with reference to AB) again completely dominating in the progeny of S x AB. Individuals of this generation ceased feeding in the laboratory during the latter part of August, 1905, and were placed for hiber- nation in a dark, cool, well-ventilated shaft connected with the breeding room. ‘They were isolated by broods. During the first week of March, 1906, the insects began to move about in the cages. [hey were then brought into the breeding room and feeding began. A large number of S individuals sur- vived the winter, while but twenty-five AB survived. On March 5, 1906, matings were made in the following cate- 232 Isabel McCracken gories for a continuation of the study of the hereditary behavior of the AB character. Category a b C d h 1 AB (male or female) of AB parents xS of S parents. AB of AB parents xAB of AB parents. AB of AB parents xB (collected ’06). B of B x AB parents x B of AB XAB parents (hybrid B xextracted B). B of B xAB parents xB of B x AB parents (hybrid B xhybrid B). S of SXAB parents xS of S xAB parents (hybrid S xhybrid S). B of AB xABxB collected ’06 (extracted B x pure B (pure with ref. to AB). B of BxAB xB collected ’06 (hybrid B xpure B). AB (’06 sport) x AB (706 sport). The following pedigree-diagrams are given for categories d, e, f, g, and h, in order to show more graphically the character of the lineage. DIAGRAM VII—Category d First Matings SexXis Bx 1 First Generation AB (sport) AB (sport) Second Generation AB, B and I (mixed) Third Generation AB and B (12:1) AB (all) B (pure) Fourth Generation AB (all) AB and B (3:1) B (all) Fifth Generation Band AB B (all) (14:1) 9 broods 1 brood S port in Melasoma and its Behavior in Heredity 233 In succeeding diagrams the ancestry of at least one parent is similar to that in Diagram VII for the first two generations and hence is not repeated. DIAGRAM VIII—Category e Third Generation AUB (all) cee be OG Fourth Generation B (all) Fifth Generation B and AB (1:1) B, AB and I (9:7:1) 1 brood 9 broods DIAGRAM IX—Category f Third Generation AB (all) x S Fourth Generation S (all) Fifth Generation S, B and AB (2.4;1:1) 16 broods DIAGRAM X—Category g Third Generation AB (all) Fourth Generation AB and B (3:1) | °06 Fifth Generation B and AB (2:1) B, AB and I (7:2:1) 14 broods 6 broods DIAGRAM XJ—Category h Third Generation AB (all) B’05 Fourth Generation B (all) B ’06 Fifth Generation B (all) B and AB (5:1) 3B, AB and I (13:1:!) 10 broods 2 broods 3 broods 234 Isabel McCracken Behavior of the alternate color characters in this generation is entirely consistent with that in previous generations. In ‘Table IV, category a, S is dominant in every brood (com- pletely dominant in 14 broods, transmitting B in 22 broods). In no case does the allelomorph, AB, appear. In category b, AB is approaching a stage of purity (entirely pure in 27 out of 30 broods). In the three broods in which mosaics occur, there are but four in one brood and two in each of two other broods. Comparison of this category with category a of Tables IT and III shows here, in the 4th generation from the sport parent an increasing stability of the sport character. In category c, type B is almost completely dominant over the type AB (entirely so, if I is considered to be a modified B in char- acter, as its behavior appears to indicate). In category 7, where two ’06 sports were mated, the result par- allels that of similar matings in the previous year (Table II, cate- gory a),in the predominance of AB in mixed broods. No broods wholly AB were obtained. _ Submitting categories d, e, g and h to a closer analysis, we obtain the following data: Category d Extracted B (from AB AB) Xhybrid B (from ABB) =1B brood, 9 mixed broods (B: AB as 14:1). : : : AB: I —g brood e Hybrid B Xhybrid B=10 mixed broods ee AB tas a Se eee es BAB sb —6 brood g Extracted BX pure B* =20 mixed broods te AB cae va ae : ‘ JS aR —3 brood h Hybrid BXpure B*¥ =10B, 5 mixed broods | AB ae He 2 *Pure in the sense of having no previous contamination with AB. The following results are noticeable: 1 The total exclusion of S. 2 The unexceptional predominance of B over AB and I in mixed broods. 3 The large preponderance of exclusively B broods where pure B meets hybrid B. TABLE IV_ (Fifth Generation) | ; man | | ; Proportion of Mating | GGG-Gr. Goveat-cr. parents Great Gr. Grandparents Pasa Total No. | pool No. | Character of the Resid category, parents . parents | | Broods | Individuals Broods | | | ; tive Classes a f | a Ss | | | $ ae |sport ABXsp. AB | ABXAB <> CACM Pee ret | BXB | oN | | ABXAB AB* XS ’o6 36 | 1000 Oe boston 14 BXB) ss os | Siand'By 022) nena: 7 | | ABXAB Ss x Ss \ | “ “ | e BxB{ | | | b “ | « “ “ “ “ “ 30 | 721 AR eee: 27 «“ | “ “ “ “ | “ “« } AB*X AB AB and I 3) 751 c | ef Ms YW | (Gee | ee cs AB*X B ’06 22 | 696 SIs enero tn é 10 | | BandI 12] 112% | ne A Wee lee « NG | CABYAB. “)./ ae 10 230 Bees 1 a uss aS an || CO AB XB ’05 i | BandAB 9 14:1 = os - ——— e fe Weiss co eS PAB XR Ror 25 10 311 Reece ane ° sf ot < i PAB OCB cordial | BandAB 1| 1:1 | |B, ABandI 9 9:7:1 ee | | | if “ “ “ce “ “ | AB x S Pe | « é it a ABXS } SxS 16 637 | S,BandAB16 2.42131 Voss se tie oe eT | i g ot |e if | eAb A | BYEXB’06 | 20 369 BandAB 14 2:1 A oe eaeess Fee Se h | « s ut ChE ABXB’o5 B*¥* xX B ’06 15 407 [Bn Call) racers 10 | BandAB 2) 5:1 | |B, AB and 13 relay Bi i | | AB ’06X AB 06 dee | 2x6 BandAB 5] 123 | | B, ABandI 3) 1:2 *The ancestry of the AB parent only is given; in category a mated with pure S, in category b, mated with AB of similar ancestry, in category c, mated with pure B. ** The ancestry of this B parent only is given, the other being a pure B. Sport in Melasoma and its Behavior in Heredity 235 4. Extracted B (B from ABxAB) transmits the sport character. For sixth generation data the following categories of matings were made: Category a SxS, of ABxS parents. 6 AB XAB, of AB and AB parents. cule some distance between the mm. adductor longus and pectineus, where it ends. After giving off the cruralis and several branches to the abdom- inal walls, the eighth nerve passes abruptly in a lateral direction and intertwines in a complex.manner with the seventh. From this n.cut.fem.lat. rabdom.n.spinalis VII n.cruralis Fig. 8 Experiment I. Section through the thigh of the primary “‘aneurogenic” leg near the lateral surface, showing the entrance of the n. cruralis. 67. plexus, the details of which are difficult to make out and to repre- sent in the diagram, a number of nerves are given off. One runs to the secondary limb, becoming its n. ischiadicus (Fig. 10). Three other branches arise separately from the plexus, but before Experiments in Transplanting Limbs 259 finally dividing, run together only to divide again into a number of branches; several of these run down the thigh of the secondary limb and are distributed to the region over the m. gracilis minor, taking.a course which is intermediate between the normal position n.ischiadicus r.cut.fem. post. r.cut.fem.post. acces. n.ischiadicus ext.acc. Fig 9 Experiment I. Section through the axis of the thigh of the primary “aneurogenic” limb. r. cut. fem. post. acces., nerves supplying the region of the accessory limb usually innervated by, the posterior cutaneous nerve. 7. ischiadicus ext. acc., sciatic nerve of accessory limb. 67. of the r. cutaneus femoris medialis and r. cutaneus femoris pos- terior. One of the branches enters the primary limb as the r. -cutaneus femoris posterior. The largest trunk from the plexus 260 Ross Granville Harrison becomes the n. ischiadicus of the primary limb, which may readily be followed to the knee in a single section (Fig. 9). Within the primary limb the nerves may be clearly made out. The large sciatic nerve passes distally parallel to the m. ilio fib- Primary Limb Prone r.cut.crur.lat n.peroneus Pes rprof ntibialis ‘ah 3 r.cut.crur.med.sup. eA r.superfic.t.tibialis r cut.crur post. r-cut.fem.post, r-cut Fempost.” r prof post. n.ischiadicus ay Sea x) % at XK ae EP OSD 9 NL ONS Ne Fig. 10 Experiment I. Section through the thigh of the primary ‘“aneurogenic” limb near the medial surface, showing also the secondary limb cut through the hip joint. *a, r. cutaneus femoris posterior, derived from n. ischadicus; b, accessory nerves derived directly from plexus. 67. rilaris and gives off near its beginning the r. profundus femoris posterior. Before reaching the knee it divides into two subdivi- sions, the n. tibialis and the n. peroneus. ‘The latter gives off a Experiments in T rans planting Limbs 261 nerve which follows the course of the r. cutaneous cruris posterior, a branch which normally comes from the tibial nerve. A little further down the r. cutaneus cruris lateralis is given off to the skin"(Fig. 10). Further down in the shank the peroneal nerve divides into a medial and a lateral ramus, the lateral being much the thicker of the two, especially at the point of origin. Lower down the medial ramus becomes stouter and a well defined branch to the m. tibialis anticus brevis is shown. ‘The two peroneal nerves pass into the t lat. n.peronei r.med.nperonei rprof.n tibialis rsuperfic.ntibialis reut.crur lat Fig. 11 Experiment I. Section through the two “aneurogenic” limbs; the primary limb is cut through the tarsus and the accessory limb just below the knee. 67. tarsal region (Fig. 11). Below the middle of the tarsus they come together again to form the n. peroneus communis inferior. Before the dorsum pedis is reached a nerve 1s given off from this branch which may be traced out into the foot as the n. interstitiales dor- salis primus. In the foot itself the nerve breaks up into two other nn. interstitiales dorsales. One of these interstitial nerves, the third, could not be traced and all the nerves are very fine at this level. It will be seen from the above description that the n. peroneus has a normal distribution. The relations to the muscles and other 262 Ross Granville Harrison structures of the leg are also normal. ‘The only anomaly observed was in the origin of the r. cutaneus cruris posterior. The distribution of the n. tibialis is also the same as in normal limbs. First the nerve divides into a smaller r. superficialis and a larger r. profundus (Fig. 10). The latter then gives off a well developed trunk, the r. cutaneus cruris medialis superior which may be traced for some distance down the shank. ‘The two rami of the tibial may be followed into the tarsal region and show nor- mal relations to the muscles (Fig. 11). The ramus profundus passes into the planta pedis in the proximal part of which it breaks up into the four nn. interstitiales plantares. “The manner in which these nerves arise is slightly different from the normal as described by Gaupp (Fig. 12). The r. circumflexus could not be traced. A B _—__r-prof.n.tibialis—— Vv i MT WV TT il Fig. 12 Diagram of nerve supply to the planta pedis: 4, according to Gaupp; B, as found in the primary “‘aneurogenic” limb in Experiment I. 1, I, III, IV, nn. interstitiales plantares. The accessory limb receives from the plexus a much smaller n. ischiadicus than the primary limb (Figs. gand1o). Inaddition to this it receives several short twigs that run subcutaneously along the inner posterior surface of the thigh, supplying an area of skin normally innervated by the r. cutaneus femoris posterior and to some extent by the r. cutaneus femoris medialis. ‘The former nerve is small (Fig. 10), and arises in the normal way from the sciatic highupin the thigh. No traces of the n. cruralis have been found. ‘The sciatic gives off a distinct though small ramus pro- fundus posterior, though I have been unable to detect a ramus profundus anterior, which normally also arises from the sciatic. In the lower part of the thigh the sciatic nerve divides as usual Experiments in T rans planting Limbs 263 into the peroneal and tibial nerves. The former gives off at the knee (Fig. 11) a ramus cutaneus cruris lateralis and continues down the shank whence it may be followed into the tarsus. This nerve is extremely difficult to follow and it is uncertain whether it divides into its r. lateralis and r. medialis. Certain it is that at the tibio-tarsal joint and further on only the r. medialis is present. This is, however, quite well defined and may be traced through DOO paso te n.ISCHIAdICUS’ \n.ischiadicus rr.cut.fem. post’ . ‘ycut.aberrans Fig. 13. Section through the two normal transplanted limbs in the upper partof the thigh. * The Tamus cutaneus femoris lateralis is some distance from its normal position. 67. the tarsal region. I have not been able, however, to make out the nn. interstitiales dorsales. The tibial nerve, soon after its origin, divides into a super- ficial and deep ramus of which, contrary to the normal condition, the former is distinctly larger (Fig. 11). This may be followed in its normal position through the shank and tarsus into the foot. The deep branch may be followed through the shank. At the tibio-tarsal joint it approaches very near to the superficial branch, 264 Ross Granville Harrison then it dips behind the tendon of the m. tarsalis posticus and may be followed for some distance through the tarsus. I have been unable to trace it out into the nn. interstitiales plantares. The normal transplanted, or “euneurogenic”’ limbs (Fig. 13) receive nerves derived for the most part from the eighth spinal nerve. ‘The seventh gives off a twig to the skin which becomes the r. cutaneus femoris lateralis of the primary limb; this nerve is, however, somewhat out of place. The main nerve (from the eighth) runs in between the pelvic cartilages of the primary and accessory limbs in three branches, one of which is considerably larger than the others. ‘This gives rise to the sciatic nerves of the two limbs and to the crural of the secondary. ‘The other two run distally between the pelvic cartilages, and emerge in the skin of the inner side of the two limbs becoming in each the r. cutaneus femoris posterior. Into the primary limb the greater part of the fibers of the main trunk enter as the sciatic, eee almost immediately gives off a stout r. profundus posterior which takes its normal course (Fig. 13). The sciatic may be clearly followed to the knee, breaking up into two main divisions, the peroneal and tibial. Below the knee the details of its distribution have not been studied. The accessory limb receives nerves which are considerably smaller. ‘The crural arises from the main trunk intended for the two limbs before this divides into thetwo nn. ischiadici. - This nerve skirts along the surface of the m. iliacus internus and may be followed as far as the m. pectineus. As the n. ischiadicus, which is smaller than that of the primary limb, enters the accessory limb it gives off an aberrant branch which runs to the skin over- lying the m. gluteus magnus. ‘There is no nerve of this size in the normal limb at this place. Higher up, 7. e., before the sciatic actually enters the limbit gives rise to a smallr. profundus posterior which passes ventral to the m. piriformis and may be traced along the dorsal surface of the mm. gemellus and quadratus femoris. It is not nearly so large as the corresponding nerve of the primary limb but it is normally situated with respect to the other structures of the limb. The sciatic nerve may be followed through the thigh. It divides Experiments in Transplanting Limbs 265 as usual into the peroneal and tibial nerves. “The former runs to the skin as the r. cutaneus cruris lateralis and I have been unable to find a continuation of the main trunk down the shank. ‘The tibial nerve, as far as can be made out, runs entirely into the r. superficialis, the latter may be traced past the tibio-tarsal joint. In comparing the innervation of the four transplanted extrem- ities in the present case, it is seen that the primary “aneurogenic’”’ limb has the most complete system of nerves, no important nerve being absent. Next in order comes the primary normal trans- planted leg, in which, as far as studied, only the n. cruralis is lack- ing, though there is a nerve which probably represents the r. cuta- neus femoris lateralis. “The accessory aneurogenic leg is third; it lacks the crural nerve and some branches below the knee; the nerves could not be traced with certainty into the foot. ‘The least complete system of nerves is in the accessory normal transplanted leg, where the sciatic nerve and its branches are much smaller than in the other limbs and no muscular branches are found below the knee. Oddly enough, however, there is a well developed cru- ral nerve in this limb though it is entirely absent from the primary. Experiment II." For this experiment larve of Bufo lentigi- nosus were used. ‘The procedure differed from the previous case, in that no normal limb was transplanted to the left side. “wo nerveless hind limb buds were used; these were taken from a specimen from which the spinal cord had been removed one week before. The first limb transplanted was accidentally pushed through the wound into the body cavity. “The second remained attached in the body wall. ‘The larva grew rapidly and was pre- served in Tellyesniczky’s fluid thirty-five days after transplan- tation of the limb. Asis readily seen in Fig. 6, the second of the transplanted buds has developed into a pair of hind legs, which are connected with the body by a short narrow stalk. Dorsal to these two legs, which are normally formed, there is an irregular mass, which sections show to have developed out of the bud that was pushed into the peritoneal cavity. “Iwo hind legs are dis- tinguishable in this mass but they are very irregularly developed 77Record number, Tr. Ext. 15. 266 Ross Granville Harrison and the sections are almost impossible to interpret. ‘The descrip- tion will therefore be confined to the other pair. While these limbs are further advanced in development than those in the pre- vious experiment, it was apparent from the observation of the living specimen that they were not so well formed. They had a slightly atrophic appearance and were never seen to undergo even the slight twitching movements observed in the first experiment. Sections show that a small nerve trunk, which arises from the eighth spinal nerve of the host enters the stalk which connects the limbs with the abdominal wall, and running between the two pelvic cartilages, 1s continued into the accessory or super-regener- ated limb. ‘This is all the more remarkable because no nerves could be traced into the primary limb. ‘The nerve in the accessory leg follows the course normally taken by the n. ischiadicus. Above the knee joint it divides into two trunks, one running to the flexor and one to the extensor surface of the shank. Both of these are cutaneous nerves and correspond respectively in distribution to the r. cutaneus cruris posterior, which arises normally from the n. tibialis, and to the r. cutaneus cruris lateralis derived normally from the n. peroneus. No muscular nerves can be made out below the knee. The imperfect innervation of the limbs in this case as compared with the previous one is due no doubt to less firm implantation into the tissues of the host. | Experiment I1I.?* In this case a hind limb bud taken from a normal larva (Bufo lentiginosus) was implanted in the left side and one taken from a nerveless larva on the right. The spinal cord of the latter had been excised one week before. Both limbs developed well and produced accessory limbs. The normal budonthe left produced a typical limb scarcely distinguish- able from the primary. ‘The nerveless limb produced an imper- fect appendage, which in turn bore an accessory bud. ‘The speci- men was preserved forty days after the operation (Fig. 5). The limbs derived from the normal transplanted bud receive a large nerve from the seventh spinal nerve of the host. Two 28Record number, Tr. Ext. 12. Experiments in Transplanting Limbs 267 small branches are given off to supply the primary limb. One fol- lows the sciatic artery a short distance down the thigh, and is to be regarded as a rudimentary n. ischiadicus. The other runs to the skin of the thigh and corresponds in position to the r. cutaneus femoris lateralis. The secondary limb receives a much larger bundle of fibers. These run into a large n. ischiadicus, which however ends before it reaches the knee. It gives off in the upper part of the thigh a distinct r. profundus posterior. The limbs derived from the nerveless bud also contain nerves. A branch from the seventh spinal nerve supplies them. ‘This nerve runs first as a compact bundle. Just before passing the pelvic cartilage it becomes frayed out to some extent, but may nevertheless be followed nearly to the knee, giving off a r. pro- fundus posterior. The secondary limb is much less advanced in development than the primary but it receives a large branch of the above mentioned nerve, which runs into the thigh for a short distance in the position of the n. ischiadicus. In this case, as is readily seen from the figure, all four of the transplanted limbs are considerably less advanced in develop- ment than in the first case described. It is possible that had the larva been kept alive for a longer time, the nervous system of the limbs would have become more complete. It is worthy of note that in this case the primary normal leg has the least complete innervation of all the four transplanted appendages. Experiment IV.** ‘This experiment, made upon Rana larve, differed from the others in that the limbs were transplanted to the back immediately behind the anterior lymph hearts. As before, a normal left bud was placed on the left side and a nerveless right on the right. The larva from which the latter was taken had lived nine days after extirpation of its spinal cord. Sections of this larva show that there are no nerves posterior to the vagus, the funicular fibers not even having grown out from the brain. The yolk is entirely absorbed except for a few granules in the intestinal epithelium. 29Record number, Tr. Ext. 11. 268 Ross Granville Harrison Each bud developed into but a single appendage, and neither of these were so far advanced in development as the limbs in the other cases described. “Twenty-six days after the transplantation the specimen was preserved and afterward examined in sections. Both of the transplanted legs are innervated principally by the r. lateralis vagi. In the case of the normal transplanted limb on the left side a branch is given off from this nerve, which after skirting along a large vesicle, formed from the transplanted tissue, finally enters the thigh. Here it may be followed for some distance as the n. ischiadicus. In addition to this nerve a small branch from one of the spinal nerves extends out along the lateral surface of the thigh in the region normally supplied by the r. cutaneus femoris lateralis. In the nerveless transplanted limb, which is cut more favorably than the other, the n. ischiadicus formed by the lateralis vagi may be readily traced to the knee. In this region it divides, one branch running to the skin where it may be followed some distance fur- ther. The other ultimately becomes lost in the mesenchyme. No traces of lateral line sense organs in the leg could be found. There are irregularities in the development of the cartilages in this limb and the muscles are scarcely differentiated at all so that the topographical relations are to some extent uncertain. In comparing the above experiments it is seen that with a single exception all of the transplanted limbs contain nerves. There are great individual differences as regards completeness of inner- vation but in this respect the corresponding limbs in the different experiments do not occupy the same relative position. These features are expressed in the accompanying table. From this it is clear that the limbs which have been taken from nerveless individuals have fared rather better as regards innervation than the normal transplanted limbs have, and also that while the acces- sory limbs are less completely innervated in three cases, they are more completely innervated in two. The most constant nerve is the ischiadicus, and as might be expected the proximal regions of the transplanted limbs are the most completely supplied. In general the cutaneous nerves are Experiments in Transplanting Limbs 269 more fully represented than the muscular. With regard to the former it may be pointed out that some variations in position and origin have been observed. These have been noted in both the “‘aneurogenic” and normal limbs and in the primary as well as in the secondary. It follows that we cannot discriminate between the different types of limb as regards their ability to acquire a normal nervous system. Table showing the relative completeness cf innervation of the limbs in the individual cases. 2 EXPERIMENT I ExperIMENTIIT | Experiment III EXPERIMENT IV < < = hp eae se a ial an w Doe a= | mo Se = i=) nw oe co Oo n+“ ine) | fo) pay eS Wo} ao} om] | = 40 5 = oO Bia ope) Sica o Oo | Bow oy KS) lard fea 5 oy eo o a ao o oS. bo & eee AN a et omr = ZH ARS oe 8|no Ss Ellas oe SIiIACO | o S bia Ble det || a Se eis | ee ee Ne) See || o = 2 ~~ o Pa Bias | vo [az A Galli dal o Ma 2 e © | & has < a o2eloew Slo 8 elas So 2 Blase Slo 2 | a Sd z § Sa s/S Bk Sag s|s Ses ge sae eggs ses , £s [eb bly Sa ee Sly S alee biy & wise bla bE 2h $7 POk seal gOs|se4jgos seal gor a i) fa i) x — a | Ay | — —— ? ae | Primary norm x 2 x eee leone seen (ewe GE) 3 Access. norm, x 4 x I Pri. nerveless x I fe) 2 | 2 x 1(?)* Accessory | | | , | nerveless x 3 x I x 3 | *It is uncertain which of these should be regarded as the more completely innervated. T rans plantation of Normal Limbs to Nerveless Regions The object of the following experiments was to test the power of peripheral nerve fibers to develop when entirely cut off from nervous connection with the centers. It has already been shown that both Braus and Banchi have failed to establish their claim that the nerves have this power, for in their experiments the nerves studied were contained in appendages that were implanted into regions where numerous nerves were present. The only crucial test of this question by means of the method of transplan- tation is to graft tissues containing developing nerves to the body of an individual in which nerves are entirely lacking, or at least to an extensive nerveless region. While no difficulty has been encountered in obtaining tadpoles with extensive nerveless regions, it haS heretofore been found 270 Ross Granville Harrison impossible to keep such specimens alive after the yolk is gone. Even when the brain and cranial ganglia are left intact so as not to interfere with the normal movements of the mouth parts and gills, the larvae soon succumb, owing to their inability tomove about and obtain food. During the course of my experiments in the spring of 1906 a method was devised for providing the cordless, and therefore paralyzed, larve with nurses. ‘This is accomplished in the following manner: We start with embryos about 3 mm. in length, in which no nerves are as yet-differentiated. After removal of the entire medullary cord caudal to the vagus region, a small piece is cut off the side of the belly of the embryo and a similar piece is taken from the opposite side of a normal embryo. The wound surfaces of the two embryos are then brought together and the embryos are held in place for an hour by means of pieces of silver wire, as described by Born, after which time they are permanently united. ‘The further development of the pair takes place normally except as regards the direct effect of the wound healing, which brings about the formation of intestinal and vas- cular anastomoses between the two. In this way the normal component, which moves about and obtains food, is able to sus- tain the pair for some time. When the yolk is about absorbed the experiment is completed by transplanting to the nerveless component a limb bud taken from a normal larva of the same age or a little older. Such limbs contain, as already described,*° the terminal twigs of nerves. The limb is grafted to the pos- terior part of the trunk a little dorsal and cranial to the natural hind limb, and in all cases was put on the free side of the body, v. ¢., the side away from the nurse. The trunk region of the larva remains nerveless except that the r. lateralis vagi is present. There may be some extension of fibers from the nurse into the tissues of the nerveless component immediately adjoining the former, but only in one instance has a nerve been observed passing from the normal component to the opposite side of the other. This nerve passed under the notochord of the nerveless component after giv- ing off twigs to the axial musculature and finally ended in the skin wSeelp.253. Experiments in Transplanting Limbs 27 of the opposite side some little distance from the transplanted limb. While it is necessary, therefore, to guard against contam- ination even in these experiments, still in all the cases but this one there is a very extensive nerveless region and the limb which is transplanted to this is far removed from the source of extrinsic nerves. The outcome of these experiments was not altogether satis- factory because’the limbs transplanted to the nerveless individuals in no case developed rapidly and hence even after the expiration of three weeks or more they are nothing more than mere knobs (Fig. 14). The natural hind limbs of these specimens are like- Fig. 14 Four nerveless larve, attached to normal nurses. e, transplanted extremity. 2. wise poorly developed and it seems probable that the nutrition is insufficient to provide for a normal rate of growth. By modif- cation of the method, I hope during the present season to obtain more conclusive results. It will not be necessary to describe the individual experiments in great detail. Seven cases have been examined in all. The specimens were preserved two, three, five, seven, ten, twelve and fifteen days, respectively, after transplantation of the limb. In four of these no traces whatever of nerves could be found in the transplanted limbs or anywhere near them, although at the time 272 Ross Granville Harrison of transplantation the limbs did contain fine nerve twigs. The three specimens that do contain traces of nerves are those which were preserved two, seven and ten days, respectively, after the operation. In the first of these there is a short twig present which contains several Schwann cells and fibrillz in a state of disinte- gration. “The same is true of the ten-day specimen. In the seven- day specimen considerably more than the usual amount of tissue had been transplanted and this included part of several myo- tomes; hence this experiment is directly comparable to those of Banchi. In the transplanted tissue several structures. are to be found which are undoubtedly degenerating nerves. One of these is a chain of cells without any visible fibrilla; the others show fibrilla in addition to the cells, though the former are indis- tinct. The results of these experiments plainly indicate that there is no progressive development of the nerve after severance of its con- nection with the center. On the contrary there are rapid regres- sive changes, which in the majority of cases result in the entire disappearance of the nerves within a few days after they are cut off from their centers. “The one case, in which the larger trunks were transplanted and which showed definite traces of the trans- planted nerves after the expiration of ten days, may be taken as indicating that larger nerves will persist longer than the finer ter- minal twigs. Two other series of experiments have been made for the pur- pose of testing ina more simple manner, the direct effect of the removal of the ganglion cell upon the growing nerve fiber. It was found that excision of the spinal cord in embryos of Rana palustris 8.5 mm. long results in the complete disappearance of the motor rami after a few days, and likewise the removal of the vagus ganglion in embryos 6 mm. long brings about rapid disin- tegration and early disappearance of the r. lateralis.** In these experiments sources of contamination from anastomoses were under control. The results are in entire accord with the results of transplantation of limbs just described, and they show that a 31These experiments were made in collaboration with Mr. Laurence Selling, who will report upon them later in full. Experiments in Transplanting Limbs 273 young developing nerve which has been deprived of its ganglion is doomed to early disintegration. CONCLUSION The immediate result of the experiments in transplanting ex- tremity buds to normal individuals is to show that, as far as the acquisition of a normal peripheral nervous system is concerned, it is quite immaterial whether the bud, prior to its transplanta- tion, has developed in connection with the central nervous system or not. It is likewise shown to be of no consequence, as regards its nerve supply, whether a limb develops directly out of the trans- planted bud or whether it arises later as an accessory or super- regenerated appendage. ‘These facts are in direct contradiction to the premises upon which Braus bases his support of Hensen’s theory. The whole superstructure of his argument therefore falls to the ground, and it will be necessary to build entirely anew in inquiring into the bearing of the experiments upon the prob- lems of the development of nerves. It is clear that the experiments do not tear directly enough upon the point to decide satisfactorily questions of histogenesis, though they do throw important light upon the manner in which the peripheral distribution of the nerves is brought about. The original experiment of Braus, confirmed in the present study, shows that a normal limb bud, when transplanted to practically any region of the body of a normal tadpole, will acquire a system of peripheral nerves, which do not differ appreciably from the normal in their arrangement, and which are connected with the nerves of the region into which the limb is implanted, although in normal individuals the latter nerves may have no relation what- ever with the limbs. ‘This fact, though in other respects of cardi- nal importance, affords no solution of our problem because it may be interpreted in accordance with either the primary continuity or the outgrowth theories; either the nerves arise 77 situ out of structures present within the limb at the time of transplantation, or they grow in from the nerves of the host and are guided to their proper places of termination by the other structures within the limb. OTA: Ross Granville Harrison Experiments which have already been reported by Lewis and by myself® are sufficient to decide in favor of the latter alternative. Destroy the nerve centers of an embryo no nerves ever develop. Transplant the centers containing ganglion cells to otherwise sterile (nerveless) regions, nerves will develop radiating from them, often following paths entirely unknown as nerve paths in the nor- mal organism; in one case even the peritoneal cavity was bridged. Alter in the most profound manner the path normally taken by certain nerve fibers, at the same time leaving the nerve centers intact, fibers will, nevertheless, develop in connection with the center, following the normal direction of growth, though in strange surroundings. Again,” if the ganglion cells are removed after the nerves are partially developed, further development ceases and all traces of the nerve may entirely disappear; but on the other hand, as is well known, if the nerve is removed leaving the center intact, a new nerve is soon formed in its place. ‘These facts show that the first essential for the formation of the nerve fiber is the ganglion cell. The only other condition, which, as far as known at the present time, is necessary, is that there must be a surround- ing medium in the form of living tissue. ‘There is no evidence, however, that any specifically formed or localized structures, esser.- tial to the formation of nerve fibers, are present within this medium.*4 Hensen’s theory supposes that the protoplasmic bridges con- necting the various cells of the embryonic body, play this part. 32H arrison ’06; Lewis ’o6. 33See p. 272 34Recently Held has maintained, in accordance with Hensen’s theory, that the outgrowth of the nerve fiber from the ganglion cell is only apparent and in reality is but a differentiation of preéxisting protoplasmic filaments found between the various organsin the embryo. ‘‘Entsteht also die Nerven- leitung durch die Umwandlung von Plasmodesmen in Neurodesmen” (op. cit., p. 188). A full dis- cussion of this work will be deferred to a future communication dealing more especially with histogenesis. It may, however, be pointed out here that it is by no means certain that the plasmodesms are not arte- facts—products of coagulation (cf. Harrison, Arch. f. mikr. Anat., Bd. 57, p. 421, and v. Lenhossék, Neurologisches Centralblatt, Bd. 26, p. 127). Nor can it be certain in consideration of the extreme minuteness of the structures in question, whether the fine filaments of the nerve process actually run into the fine protoplasmic threads or along them. At least it is very siginficant to find that Ramon y Cajal (07), employing methods essentially similar to those of Held, nevertheless gives his full support to the outgrowth view. Ex periments in T rans planting Limbs 275 Coming from a physiologist, the especial virtue of the theory, as might be expected, is physiological; it places the genesis of the permanent nerve paths upon the basis of functional adaptation; of all the numerous undifferentiated protoplasmic connections existing in the embryonic body, it is only those which function in conducting impulses that persist as permanent nerves; the remain- der atrophy from disease. In refutation of this hypothesis it may be pointed out, however, that the functional activity of a nerve has no appreciable influence at least upon its early development. Amphibian embryos reared in a solution of acetone chloroform® acquire a perfectly formed nervous system, and one capable of normal functional activity, though during the whole period of their development up to the stage when the yolk is entirely absorbed, at which time the peripheral nerves are all well differ- entiated, no functional activity of the nervous system is mani- fested. Furthermore, in the transplantation experiments just referred to, a number of nerves were formed to which no conceiv- able function could be assigned, as, for instance, the funicular fibers which, after the removal of the medullary cord of the trunk, . extend out from the brain and lose themselves in the mesenchyme. It is precisely in this connection that the experiments in trans- planting nerveless limbs are of great significance. A nerveless limb is taken from an organism that has undergone the greater part of its development after having been deprived of its spinal cord. As a consequence, no nerves are developed in the trunk region, and there is no evidence of nervous activity there, although in a normal individual during the same period, all of the important nerves, including those running to the extremities are visibly dif- ferentiated and for the most part are functioning. It cannot be supposed that the pre-nervous protoplasmic bridges, postulated by Hensen’s theory, would be able to survive this long period of dis- use, for, as experiments show, even visibly differentiated nerve fibers degenerate very rapidly after removal of their centers, often disappearing without leaving a trace in a much shorter time than that during which the nerveless individuals in question have been 8>Harrison 04. 276 Ross Granville Harrison without their spinal cords. And yet when the limb of such an individual is transplanted to a normal larva it acquires a complete and normally arranged system of peripheral nerves just as a limb _taken from a normal larva does. In other words, lack of func- tional activity, consequent upon the absence of nerve centers dur- ing a protracted and important period of development, does not in the least interfere with the later normal development of the nerves as soon as new nerve centers are supplied. ‘These nerves, there- fore, must be regarded as the product of the nerve centers alone. This answer to the question being firmly established as correct, the cardinalimportance of Braus’ fundamental experiment becomes apparent, for, contrary to the conclusion drawn by its author, it shows that the structures contained within the limb must have a very important directive action upon the developing nerve fibers, in that they determine their mode of distribution. The manner of branching cannot possibly be predetermined in the ingrowing nerves themselves, because in the normal body these same nerves have an entirely different distribution. Let us picture to ourselves what probably takes place after a limb is transplanted. It is put into a region well supplied with nerves. “Lhe wound made for the reception of the bud involves without fail injuries to the nerves of the region. ‘This stimulates the fibers to grow and in so doing some of them will come into contact with the cells of the transplanted bud, which at that time consists of a blastema of mesenchyme cells covered by epidermis but not visibly differentiated. As the bud grows into a leg and the blastema differentiates, the nerve fibers become arranged and segregated according as they are at- tached to the organs within the limb. _ In the limbs 1n normal posi- tion the development of the nerves must go on in the same way. Here too the nerve fibers reach the bud when it is still without visible differentiation. Contact with the cells contained within it being made at that time, the peripheral branches of the nerves are determined as the constituent parts of the limbs are segre- gated. ‘The fact that any nerve in whose way a limb is planted 1s capable of giving rise to intrinsic nerves having perfectly normal arrangement, shows that the nerves themselves must be guided in the formation of their terminal ramifications. Ex periments in T rans planting “Limbs iG) This interpretation is in accordance with Nussbaum’s law that the course of the nerve within the muscle is an index of the direc- tion in which the muscle has grown.” In other words, the orig- inal point of contact between nerve and muscle persists as the entrance point of the nerve after growth is completed. It is only necessary to suppose that action takes place at very short distances in bringing about first contact between the devel- oping nerve fibers and the cells of the limb bud. Failure on the part of Hensen and the later advocates of his theory to realize this, has led to the great magnification of the difficulties which an out- growing fiber would supposedly encounter in reaching its proper end organ. It is not necessary to imagine, as a number of writers do, that the growing nerve would have to wend its way through a labyrinth of differentiated tissues, extending from the hip to the toes, in order to reach its end organ, but merely that the nerve must grow independently as far as the base of the undifferentiated limb bud, the rest being provided by the development of the limb itself. “he above interpretation calls nothing mysterious, noth- ing hypothetical into play. It is based solely upon known facts and does not postulate the existence of invisible and otherwise unknown structures. It is the only explanation that can be ac- cepted in the present state of our knowledge. Moreover the varia- tions in the distribution of nerves within normal limbs and espec- ially the slight aberrations from the normal which have been noted in the position of some of the nerves in the transplanted limbs meet a ready explanation on this basis. The foregoing suggests the consideration of certain meristic varia- tions in peripheral nerves.*7 It has long been known, having been pointed out especially by Furbringer, that the nerve plexus from which a limb is supplied might in two cases have a different metameric origin, and yet the nerves arising from the plexus might be distributed in the same manner in each. Gegenbaur, who, like Furbringer, held closely to the theory that muscle and nerve form an inseparable unit, admitted the difficulty of satis- 3$Nussbaum 794. 37This matter was brought up by Dr. McMurrich and Dr. Bardeen during the discussion of my paper at the Toronto Meeting, and has been fully discussed very recently by the latter (Bardeen ’o7). 278 Ross Granville Harrison factorily explaining such variations.** ‘The interpretation of the transplantation experiments just given avoids this difficulty, for it brings out the fact that there are two main determining factors in the development of the innervation of a limb. The first of these is the position and extent of the extremity at the time of origin; this determines the source of the nerve supply. The second factor is the mode of segregation and growth of the individual structures of the limb, which determines the intrinsic distribution of its nerves. The experiments show that these two factors may be varied inde- pendently of one another. Variations in the position and extent of the rudiment of a limb, which may be assumed to occur fre- quently in nature, will, therefore, result in the ingrowth of differ- ent metameric nerves, and still the intrinsic distribution of their branches may remain constant, owing to the circumstance that the factors determining the latter operate in the same way regard- less of the source of the nerves upon which they have to act. Both individual variations and specific differences in the metameric ori- gin of limb-plexuses are naturally explained in this way. SUMMARY 1 Limb buds of tadpoles, when transplanted ‘to various parts of the body of normal individuals, develop normally and acquire usually a complete or partially complete system of peripheral nerves, which have normal arrangement and are connected with the nerves of the host supplying the region in which the limb 1s implanted. 2 The whole trunk region of an embryo may be made nerveless”’ by cutting out the medullary cord posterior to the ear vesicle, just after closure of the medullary folds. Limb buds taken from such individuals and transplanted to normal larve behave exactly like the normal limb buds as regards the acquirement of nerves. 3 Accessory limbs, which frequently develop from transplanted buds by a process of super-regeneration, receive nerves either dir- ectly from the host or from nerve trunks running to the primary transplanted extremity. Sometimes the innervation of the acces- ce 38* Die metamere Umbildung, wie sie sich als Verschiebung zeigt, bleibt damit ein Problem, dessen Lésung man sich vorlaufig nur mittels der Hypothese nahern kann.’’ Gegenbaur, op. cit., p. 613 Experiments in Transplanting Limbs 279 sory limb is more complete than that of the primary, though more frequently the reverse is the case. 4. Itis possible to keep a nerveless larva alive for a period of a month by grafting it to a normal larva, which acts as a nurse. When a normal extremity bud is transplanted to such a nerveless larva, the nerve twigs contained within the former soon degenerate and no signs of progressive development of the nerves in such cases are to be observed. ‘There is no evidence that an embryonic nerve can continue its development after its connection with the center 1s severed and prevented from being re-established. Cases which have been reported to the contrary are to be attributed to the pres- ence of anastomoses. 5 The nerves are not formed im situ in the transplanted limbs but grow into them from the nerves of the host. Experiments which have previously been reported permit of no other conclu- sion and this is strongly reinforced by the experiments with nerve- less limbs. Hensen’s theory of primary continuity between nerve center and end organ is untenable, nor does functional activity play any part in the early development of the nerve paths. 6 Nerves reach the limbs, both natural and transplanted, when the limbs are in the earliest stages of their development and are composed of an undifferentiated blastema of mesenchyme cells. The intrinsic distribution of the nerves is determined by the struc- tures within the limb, most probably at the time when the cells of the blastema segregate into the various definitive structures. ‘This follows necessarily from the fact that any nerve which 1s led to enter the limb will assume the normal arrangement for that limb. 7 There are thus two important factors determining the inner- vation of a limb: First, its position and extent at the time of origin; upon this the source of nerve supply depends. Second, the struc- tures within the limb itself; these fix the mode of distribution of the nerves. 8 These two factors are entirely independent of one another. Meristic differences in nerve supply of limbs, either between individ- uals or between species, which may exist without affecting the intrinsic distribution of the nerves, are to be regarded simply as due to variations in the original position and extent of the limb rudiments. 280 Ross Granville Harrison REFERENCES Bancut, ARTuRO, ’04—Sviluppo degli arti addominali del Bufo vulgaris innestati in sede anomala. Monitofe Zoologico Italiano, Anno 15. °o5—Sviluppo degli arti pelvici del Bufo vulgaris innestati in sede anomala. Arch. di Anat. e di Embriol. Vol. 4. °o6—Sullo sviluppo dei nervi periferici in maniera independente dal sis- tema nervoso centrale. Anatom. Anzeiger. Bd. 28. BARDEEN, CuHar_Es R., ’07—Development and Variation of the Nerves and the Musculature of the Inferior Extremity and of the Neighboring Re- gions of the Trunk in Man. American Journ. Anat., Vol. vi. BarrurtH, D.,’94—Die experimentelle Regeneration tiberschiissiger Gliedmassen- teile ber den Amphibien (Polydaktylie). Archiv f. Entwickelungs- mechanik. Vol. I. Braus, HERMANN, '03—Versuch einer Experimentellen Morphologie. Naturhis« torisch-Medicinischer Verein Heidelberg. (Medizin. Sekt.) Sitzung vom 17 Nov. Miinchener medizinische Wochenschrift. °04—Einige Ergebnisse der Transplantation von Organanlagen bei Bom- binatorlarven. Verhandlungen der Anatomischen Gesellschaft. 18. Versammlung in Jena. °o5 —Experimentelle Beitrage zur Frage nach der Entwickelung peripherer Nerven. Anatom. Anzeiger, Bd. 26. Caja, S. Ramon y, ’06—Studien iiber die Hinrinde des Menschen. 5 Heft. Leip- zig. (Cited from Schiefferdecker.) °07—Die histogenetischen Beweise der Neuronentheorie von His und Forel. Anatom. Anzeiger, Bd. 30. Doctet, A., ’04—Ueber die Nervenendigungen in den Grandryschen und Herbst- schen Korperchen im Zusammenhange mit der Frage der Neuronen- theorie. Anatomischer Anzeiger. Bd. 25. ForssMan, J., ’98—Ueber die Ursachen welche die Wachstumsrichtung der periph- eren Nervenfasern bei der Regeneration bestimmen. Ziegler’s Bei- trage zur patholog. Anat. u. zur allgem. Pathologie, Bd. 24. °oo—Zur Kenntniss des Neurotropismus. Ibid. Bd. 27. Frirscu, Gustav, ’87—Die elektrischen Fische. I. Abtheilung. Malapterurus electricus. Leipzig. FUrBRINGER, Max, ’79—Zur Lehre von den Umbildungen der Nervenplexus. Morphol. Jahrb. Bd. 5 Gaupp, Ernst, ’96-’99—A. Ecker’s and R. Wiedersheim’s Anatomie des Frosches. Erste und zweite Abteilungen. Dritte Anflage. Braunschweig. Experiments in T rans planting Limbs - 281 GEGENBAUR, CarL, '98—Vergleichende Anatomie der Wirbelthiere. I. Band. Leipzig. GemeELLt, Fra Acostino, ’06—Ricerche sperimentali sullo sviluppo dei nervi degli arti pelvici di Bufo vulgaris innestati in sede anomala. Revista di Pathologia Nervosa e Mentale. Anno 11. Harrison, Ross GranviLLE, ’03—Experimentelle Untersuchungen tber die Ent- wicklung der Sinnesorgane der Seitenlinie bei den Amphibien. Arch. f. mikr. Anat. Bd. 63. ’o4—An Experimental Study of the Relation of the Nervous System to the Developing Musculature in the Embryo of the Frog. Amer- ican Journ. Anat. Vol. 3. ’o4a—Neue Versuche und Beobachtungen iiber die Entwicklung der per- ipheren Nerven der Wirbeltiere. Sitzungsber. der Niederrheinischen Gesellsch. f. Natur u. Heilkunde zu Bonn. Sitzung 11. Juli. ’o5—Further Experiments on the Development of Peripheral Nerves. American Journ. Anat. Vol. 5. Hep, Hans, ’06—Zur Histogenese der Nervenleitung. Verhandlungen der Ana- tomischen Gesellschaft. 20. Versammlung in Rostock. Hensen, V., ’64—Ueber die Entwickelung des Gewebes und der Nerven im Schwanze der Froschlarve. Virchow’s Archiv. Bd. 31. ’*68—Ueber die Nerven im Schwanz der Froschlarven. Archiv f. mikr. Anat. Bd. 4. ’0o3—Die Entwickelungsmechanik der Nervenbahnen im Embryo der Saugetiere. Kiel und Leipzig. Lewis, WarrEN Harmon, ’06—Experimental Evidence in Support of the Out- growth Theory of the Axis Cylinder. Proc. Assoc. Am. Anat. Am. Journ. Anat. Vol. V. Nusspaum, M., ’94—Nerv und Muskel: Abhangigkeit des Muskelwachstums vom Nervenverlauf. Verhandlungen der Anatomischen Gesellschaft. 8. Versammlung in Strassburg. ; Rerzius, Gustar, ’05—Punktsubstanz, “‘Nervoses Grau” und Neuronenlehre. Biol. Unterscuhungen, N. F., Bd. 12. SCHIEFFERDECKER, P., ’06—Ueber das Verhalten der Fibrillen des Achsenzylinders an den Ranvierschen Einsuchniirungen der markhaltigen Nerven- fasern. Archiv f. mikr. Anat. Bd. 67. °o6a—Neurone und Neuronenbahnen. Leipzig. Torniger, Gustay, ’05—An Knoblauchskroten experimentell entstandene uber- zahlige Hintergliedmassen. Archiv f. Entwickelungsmechanik. Bd. 20. THE ENERGY OF SEGMENTATION AN APPLICATION OF PHYSICAL LAWS TO ORGANIC EVENTS BY E. G. SPAULDING Princeton University CONTENTS Mie PRC CUCHIOM Etec ete cite testers aya ersva oree er onorscahasou dian sieves ageMareneieve civ clisbenetnateraehors ctkeaetanateraponete 284 THE Wee Roig PESO} slates pose cont CORO IOSEE Caos nb) CONS Hea Oma oath ad cams 3 cto oo ac 285 eel ewbanst lea MCOnseratlonivesayse 0; for (U,—U,) =E =W in accordatice with the First Law, and so might be set in equivalence with pv, and this substituted, with the result an identical equation as before. By the results of the computation in accordance with the results obtained from measurement, namely, that an original increase of pressure following fertilization and a decrease of this following each cleavage take place, while the temperature and volume remain constant, the quantitative value of the resultant energy- changes will, then, have been determined. But, in accordance The Energy of Segmentation 305 with this mode of determination, W and U,—U, will not have been kept apart; accordingly it is quite possible that during the event of cleavage heat is absorbed from the environment, that the internal energy will have increased by a certain amount. How- ever, to determine the exact value of such an increase would seem to be impossible. Yet experiment makes clear that there is a change in pressure. To what is this due? Equation [8] shows that in general such a change must be due to the fractional change in some other intensity; for, with U,—U,=0, and W=pv, by substitution in [8] we get That this change should be the result of a change in tempera- ture is impossible if the view is correct that this last intensity 1s constant; yet correlative change there must be, and the evidence is that this is in the chemical conditions. we RECORD OF EXPERIMENTS The following is an epitomized record of the experiments per- formed, showing what solution-concentrations were sufficient to inhibit the earlier cleavages; the results will be found to show a fairly close agreement. ‘The figures given here are simply those of the concentration, obtained in each case by starting with a cer- tain number of cc. of a 3 m. sugar solution in sea-water for each lot of eggs transferred, and then diluting this amount with a cer- tain measured amount of sea-water, plus the 14 cc. of sea-water necessary for transferring the fertilized eggs. ‘The pressures thus obtained are here not reduced to atmospheres; that is done only for the typical experiment in which the computation is carried out completely. In all of these the eggs in the control segmented very uniformly and well. Experiment I, July 15: First segmentation stopped in sol. 10 cc. sugar sol. plus 11 cc. sea-water plus 14 cc. s. w.,"" and in 10The experimentation was done in the Marine Biological Laboratory at Woods Hole, in the summer of 1905. ls, w., sea-water. 306 E. G. Spaulding all stronger; continued in all weaker solutions. Second cleavage stopped by 10 cc. sugar sol. plus 13.75 cc. s. w., plus 14 cc.; the third by toec. sugar, plus 15 cc. s. w. plus 14; the fourth, by 10 ce. sugar plus 16.5 plus 1.5 cc. s. w. Experiment II, July 21: Made for purposes of refinement; found to be confirmatory of preceding. Experiment III, July 22: First segmentation stopped by 15 cc. sugar plus between 16 and 18 cc. s. w. plus 1} cc. Experiment IV, July 22: Inhibition point found to be between 16 and 16.5 cc. s. w. plus 15 cc. sugar plus I} s. w. Experiment V, July 24: First cleavage stopped by 15 cc. sugar plus 4 (16 plus 16.5) cc.s.w. plus 14 cc. “This means that the (16), and all stronger, stopped, while the (16.5) and all “up” from this “allowed to proceed.”’ This method of taking the “intermediate point’ was subsequently adopted in each case. Second segmenta- tion stopped by 15 cc. sugar plus 4 (20.5 +21) s. w. plus 14 cc.; third, by, 15 plus 4 (21.5 +22.5) plus 13 cc. Experiment VI, July 27: Temperature 23.5 C. Fine lot of eggs. Fertilized at 11 a. m.; repeatedly and frequently observed up to 3 p. m.; observations all confirmatory. First segmentation stopped by solution 15 cc. 3 mol. sugar sol. plus 16.75 cc. s. w. plus 1.5 s. w.; second by 15 cc. sugar plus 20.75 plus 1.5 cc. s. w.; third by 15 cc. sugar plus 21.5 plus 1.5 cc. s. w. Control; over 95 per cent of the eggs segmented uniformly. For the purposes of the computation to be made, the surfaces and volumes of each stage must be found. ‘This was done as fol- lows. In the one-cell stage the typical or “modal” sea-urchin egg is approximately spherical; accordingly the diameters of as large a number as the period of 55 to 60 minutes elapsing before the first segmentation allowed were measured by means of an ocular micrometer; from these data, widely divergent values being excluded, the average diameter was found, and the surface and vol- ume computed from well-known formula. In the two-cell stage the typical form is that of two oblate ellipsoids; the long and the short axis of each ellipse was accordingly measured for a number of eggs, the average for each taken, and the surface and volume computed. Difficulties in doing this for the four and eight-cell The Energy of Segmentation 307 stages were foreseen, but the results obtained for the first two stages showed that the volume after the first segmentation was the same as before it; it could, therefore, be assumed to be con- stant during the second and third and even subsequent stages, especially since general observation makes no change manifest. On the other hand, the increase of surface which was demonstrated to have taken place was shown to be not of direct significance in the computation made of the energy of segmentation. These measurements of diameters were made in Experiments V and VI on eggs taken, of course, from the control.” But it is evident that the numerical results thus obtained are in any com- plete computation to be combined with those obtained from the use of the inhibiting solutions on eggs of the same lot. Accord- ingly, it is the result of the complete computation from all the necessary experimental data, as taken in Experiment VI, that is presented below; and in connection therewith it may be remarked that, as between method and numerical result, it is the former rather than the latter that I would have regarded as the more worthy of emphasis. If the question be raised as to the accuracy of the numerical result, this can be estimated by considering the sources and probable limits of error introduced both by the method of compensating and by the fact that one is observing a “group” of eggs and must adopt the expedient of taking averages, etc. OBSERVED AND COMPUTED RESULTS IN EXPERIMENT VI Data: Segmentations stopped, the first by 15 cc., 3 m. sugar sol. plus 16.75 plus 1.5 cc. s. w. at 23. Gr sere: Now, it is well known that osmotic salteons:| in general follow the law for gases; and it is held, too, that there is no dissociation in a sugar solution. Accordingly the pressure of a mol. sugar sol. at o° C. is 22.4 atmospheres.” Miss Evis Berry kindly assisted me in the experiment in this way. 3An atmosphere is that unit of pressure which is exerted by a column of mercury 76 cm. in height of a density 13.596; this equals in C. G. S. terms 1013300 dynes, i.e., the pressure of such a column of mercury per sq. cm. The egg has of course an area which is only a small fractional part of a square centimeter. 308 E. G. Spaulding Making the correction for a room temperature of 23.5 C. accordance with the formula, p, =p, (1 + .00367 ¢), the osmotic pressure of the above diluted solution, 7. ¢., the pressure sufficient to inhibit the first segmentation and eee equal to the result- ant internal pressure, is 7.32 atmospheres. For the second segmentation, the numerical value of this inhib- iting pressure, as computed in a similar manner from the recorded figures above, is 6.53 atmospheres, and for the third, 6.40 atmos- pheres. MEASUREMENTS OF DIAMETERS AND AXES Average diameter of 20 typical eggs = .072 mm.; radius = .036. Area (4 2 r*) = .0164 sq. mm. Volume (# z r°) =.000214 cu. mm. Two-cell stage: each cell an oblate ellipsoid. Average axes of 20 typical segmented eggs: Lone-axis (G08 ncare oh aa tte eee 068 mm. Shore axis/(@)ac). tee toke Nee eee 039 mm. Area (J d xd” x7 ‘’) = OL TO isd: wun: for each cell; for both .0236 sq. mm. Volume (32d d” x 2) = .000206 cu. mm. for both cells together. From these values it is evident that, whereas the area has in- creased by .0072 mm., the volume has remained the same. It would now seem as if the data were at hand whose numerical values could be substituted in the “‘work integral” De W = -{ ae Pi which becomes W = v (p,—p.) when the volume remains constant, as in this case. However, before doing this, the question must be answered, as to what may be the value of that pressure which is due to the tension of the surface film or membrane of the develop- ing egg. For it might seem that the resultant internal pressure before and after each segmentation was equal to, not alone the opposing osmotic pressure of the surrounding sugar sea-water The Energy of Segmentation 309 solution, but, rather, to this plus the pressure of the surface film or membrane. Accordingly, the numerical value of this must be found, that it may be known whether it is significant for our com- putation or not. The formula by which this pressure due to the tension of the surface," if this be only a film like the surface of a drop of water, may be computed, 1s OE Boe in which ¢ is the coefficient of surface tension and r the radius of a sphere. ‘This ¢ is determined from the capillary action of a fluid in accordance with the formula, t= 4 grh D (g = action of gravity,r = radiusof tube, # = height to which the fluid is drawn up, D = the density). Pfeffer gives this coefhcient as .oI g. cm. in relation to that of water as unity. Since other determinations are lacking, I made use of this, although, of course, it must be admitted that this coefficient might vary greatly with different kinds of proto- plasm. Substituting this value in the formula, P —— oe r P = -0055 atmos. pressure For the two-cell stage, with each cell an oblate ellipsoid, the form- ula is more complicated: here p= * area (a = long axis, c = short) “The best treatment of the general problem of surface tension, etc.,which I have found is M. Heiden- hain’s Die allgemeine Ableitung der Oberflichenkrifte, etc., in Anatomische Hefte, erste Abteilung, vol. xxvi. Wiesbaden. 1904. °6 Plasmahaut u. Vakuolen, Abhandl. d. Math.-phys. KI. d. Sachs. Ak. d. Wis., 16, 185 (1891); cited by Hober, Physikalische Chemie der Zelle u. Gewebe, s. 38; Leipzig, 1902. This is the only determination I have been able to find. *For this formula I am indebted to Dr. C. R. MacInnes, of Princeton University. 310 E. G. Spaulding Substituting, we get p = .0063 atmos. Although, therefore, it appears from this that there has been an increase in the pressure which would result from the curved sur- face of the egg were this a film, it is also evident that this is of insignificant value in comparison with the values, 7.32, 6.53, 6.40 atmos. found for the inhibiting solutions. It falls “outside the limits of error,’’ and is, therefore, to be neglected in the applica- tion of the ‘“‘work integral.” The fact, however, that this pressure has such a small compara- tive value, results, evidently, from the substitution of .o1 as the coefhicient of surface tension of protoplasm. ‘The acceptance of this value is, of course, purely gratuitous; but if it be approxi- mately correct for the protoplasm of the sea-urchin egg, then the resulting small value of the pressure of the surface on the basis of the assumption that this is a film proves this assumption to be incorrect, and indicates that there must be a membrane, differ- entiated from the cytoplasm, to oppose the relatively high internal pressure as indicated by the strength of the solutions requisite to inhibit segmentation. There has been demonstrated, then, experimentally, an increase of 7.32 atmos. in the “resultant” pressure, as brought about by fertilization and the process following it up to the time of segmene . tation. As a result of these, the egg normally cleaves; it changes form, and it is now shown experimentally that as it does this the internal pressure therewith decreases; without fertilization these events do not take place. For the early segmentations, then, there are numerical data at hand from which the resultant energy change can be computed in accordance with the “‘work integral” W--\"odp Pp which becomes, when the volume is constant, is (A- Po) The Energy of Segmentation Zrt Substituting in this the numerical values obtained for pressures and volume, we find that there has taken place, as a result of fer- tilization and processes subsequent to this, an increase in the energy of the egg of Wra2 1,083,200 XX 00000021 cu, cms—1-507 Eres; that, analogously, after the first segmentation, the energy is G53) 1,012,200 X .o0000021 Cu. Ems — 1700) eras: This means, that, as involved in or as identical with the first seg- mentation, there has been a resultant energy decrease, therefore, of .168 ergs; or that it has taken this amount of energy, about 4 of the total increase resulting from fertilization, etc., to bring about this cleavage. As bringing about the second segmentation, we find by sub- stitution: (6.53 —6.40) X I 013 300 X .000 000 21 =.028 ergs of energy to have been involved. CONCLUSION This completes the computation based on the measurements taken in Experiment VI. It could, of course, also have been made for some of the other experiments, and, had our purpose been to determine as accurately as possible the numerical value of the energy of segmentation, then a large number of both experiments and computations would have been necessary, in order from these to get a mean result. But it has been not the numerical result but rather the method, that is, the practicability, on an experi- mental basis, of applying the “work integral” and so the other equations of which it is a special case, that has seemed the more important and been deemed worthy of emphasis. “Thus would I forestall the point of the possible criticism that the numerical results themselves are meager, and that they have been found for only two segmentations on data obtained in one experiment. For, while these are the facts, nevertheless, on the other hand, Experi- ment VI and its results can be regarded as typical of further pos- sibilities, while on the other the limitation to two segmentations 352 E. G. Spaulding may be regarded as due simply to certain difficulties in experi- mental procedure which, of course, further refinements may over- come. Accordingly, I shall consider that that which was my immediate purpose, namely, the application to an organic event of the same general principles as are applied to inorganic events, has resulted successfully, and that thus a basis is furnished for answering the other questions which were propounded at the beginning of my paper. However, before that is done, an interpretation must be made as to just what the results obtained show as to the character of the energy-transfer which is involved in each cleavage. Here the principles stated in our introduction and developed in our formu- lation must guide us. In answering this question it must be said, in accordance, first, with what was shown as to the conditions under which our measure- ments must be taken, and, second, with the hypotheses formed as to the forces, etc., in the egg as a system, that the numerical values obtained for the energy-transfer in the two cleavages are the measure, first, of the difference between the energy decrease and its simultaneous increase, E = W+(U,—U,), during the event of cleavage; and, second, of the resultant, in energy-terms, of all those subsidiary processes and changes, morphological and other- wise, which contribute to the event; some of these must be identical with W, others with U,—U,,; if there be any processes which do not so contribute either directly or remotely, then, of course, they are not included in this resultant. It is evident, then, first, that the result obtained allows for the possible increase in the internal energy of the ovum by the absorp- tion of heat, or other energy, though probably only the former, as simultaneous with a decrease in accordance with which work is done; and, second, that our result gives the measure, not of the entire energy of the cell, but only of that which, as an excess of the energy “lost” over that gained, is identical with the energy of cleavage. What, now, is the character of the energy-form in which there is this resultant decrease? ‘To this the answer is indicated, first, The Energy of Segmentation 212 by the hypothesis formed that in cleavage we are dealing with “forces” which are effcacious only as resultant pressures, and, second, by the nature of the factors actually determined by meas- urement, again pressures, that it is the “volume energy” which 1s so concerned. ‘This “volume energy,” here the energy of the col- loidal solution, is a function, first, of the number of molecules or of particles, and of their velocity, and, therefore, second, of the chemical splittings and combinings, and of the temperature, respectively. With the temperature and volume constant, the decrease in vol- ume energy demands a correlative decrease in the number of mole- cules, or of colloidal particles, or of both, as accompanying cleav- age. This decrease would take place as a result in turn of a com- bining, to a definite degree of course, of molecules and of particles, which chemical change would be accompanied by the passing of energy from the system (ovum) to the environment in the form of heat. At least part of the “resultant” decrease in the volume energy of the system is to be accounted for in this way. Con- cerning the remainder of the decrease the evidence shows that its reappearance is in the form of the increase in the energy of the surface and in the mechanical energy or work done in the moving of the “mass”? surrounding the system as environment. Under normal conditions it is with the intensity of the “surface pressure’? equal to the opposed intensity from within that an equilibrium of form continues. What now, finally, is the meaning of the fact that it has been possible to determine the energy of segmentation according to the method presented? That meaning | propose to summarize, for I believe it stands firm, even on the basis alone of the limited numer- ical results obtained. . As a first step in the demonstration it was necessary to state briefly the principles which it was my purpose to apply,etc. ‘These were then formulated and shown to be epitomized in the funda- mental equation dW W+ (U,—U,) =I dl The “work integral” was then shown to be a special case of this 314 E. G. Spaulding formula, with the result that the successful application of the former to segmentation would mean also the validity of the latter and therefore of the Four Laws for this event. To it there would apply, then, the principles of Determinism, Potential Difference, Conservation, etc. But these laws are, seemingly, largely if not wholly quantitative, while on the other hand the organism, e. g., the ovum, is qualita- tive as well as quantitative. What, then, is the relation of these Laws to the qualities, and what are these? To answer the former question first, it may be said, that qual- ities in both the inorganic and the organic world are, at the same time that they are qualities, also quantities; and quantities are either extensive or intensive. Of qualities certain empirical laws are discoverable, while between these laws similarities are in turn found which lead to the Four Laws epitomized in equation [8]. Thus we get a “natural classification” of laws. From this it will be seen that the generic characteristics, so expressed, have the relation to that from which they are derived of being ultimately incorporate in the concrete qualities, and that they do not, although they are predominantly quantitative, simply exist alongside of these as a separate and distinct aspect. Rather, the Four Laws express the common quantitative aspects of these concrete qualitative- quantitative phenomena. This view is directly opposed to that which regards the Four Laws, because quantitative, as “not touching” the concrete qual- ities, and then finds that these last, because not so “touched,” furnish opportunity, especially in organisms, for Indeterminism, Regulation, Freedom, Entelechies, etc. But what are the qualities themselves? Are they not of things, events and relations? Our answer is: Let “thing” be equated with system; then system implies parts, and these may be either atoms, or coexisting energies, or both. In either case some of the qualities of the “thing” result from the cooperation of the parts or elements, whose qualities are different from those of the whole which they form, the test being, that if isolated their qualities are found to be unlike those resultant ones; this bringing about by the parts of qualities which they themselves have not, may be called The Energy of Segmentation 315 “creative synthesis.’ Other qualities of the system are the same as those which the parts retain when isolated. ‘The latter give an additive result in the complex, the former do not. It is now possible to make a statement as to what the cell is, and, if we may generalize, to answer our major question as to just how different organic phenomena are from inorganic. According to our hypothesis the cell is a system, a complex of energies or of colloidal particles, etc. Some of these components can be isolated, and, with this done, are found to follow the usual inorganic laws; these they are therefore assumed to follow when in the complex. ‘The same assumption is also made for those components which cannot be isolated; that is, the contrary posi- tion that such an “exclusion” demonstrates the presence of an irreducible, organic, vital remainder is held to be incorrect in view of the successful application of the energy-laws to theorganism as a whole. The qualities of the cell, are some of them, identical with the qualities of the parts and are the additive result of these, while others are the result of the “creative synthesis” of two or more constituent energies, etc. All these qualities are at the same time quantities, either extensive or intensive. Now, without it being necessary to treat either these energies or the qualities of the system analytically, it has been possible, since at least some of them act together to produce, or are identi- cal with, the event of cleavage, to measure this as a whole and bring it under the Four Laws. ‘Thus are the subordinate events which contribute to this resultant event also brought into the range of the validity of these Laws. The qualities of the organism—which are also quantities—are, accordingly, shown to be qualities which on this quantitative side have certain characteristics which are the same as those of the inorganic world—namely, those characteristics which the Four Laws formulate. Conversely, the Four Laws, as formulating these common characteristics, and as epitomized in equation [8], bring the concrete phenomena, both organic and inorganic, and the series of empirical laws into a “natural classification.” But this does not do away with the fact that here in the so-called 316 E. G. Spaulding organic realm, as in the inorganic, there are specific qualities which differentiate each class of complex from every other class, or, indeed, each individual from every other. The organism may have, therefore, qualities which, as such, are specifically different from any found in the inorganic realm; a “reduction” of these to inorganic being as impossible as is that of one inorganic quality to another. On the other hand, these very qualities, in that they are at the same time quantities, are like the inorganic in that they have in common with these the characteristics formulated in the Four Laws. In just this respect there 1s no difference between or- ganic and inorganic; they are in the same realm whatever that be called. The only difference between organic and inorganic which still remains is, then, just that difference which persists between specific and specific, a difference which holds as good within the inorganic realm as it does between it and the organic. The only ground remaining for holding a distinction between the two realms is, that, taking the same level of classification or compari- son, the differences between certain complexes, called inorganic, is less than the difference between these and certain others called organic. But even this does not do away with the necessity of bringing all into one realm in which the principles of Conservation, Potential, Determinism, etc., are valid. I conclude, then, that all events, both organic and inorganic, take place in full conform- ity with these principles, and that there is no ground for holding or interpreting organic events, etc., to furnish contradiction or evasion of them. : FACTORS IN THE REGENERATION OF A COMPOUND HYDROID, EUDENDRIUM RAMOSUM! BY A. J. GOLDFARB Witu Two Ficures Mpa ECL Sri eins aS CALE MIEN Ge te i256 lal oscrere' er alovade s Janay evave.nictare craters) Sie isisie- shale Mes Ae eaRve re elsaaye ehemiosevets 317 2 Effects produced by removing lateral branches or pedicels...............-....0-2-055 sees 319 Aeee tects dueltorepional: ditterences—— apes? siac\cye-c.o.sters eros ores afoles sues) alayeiaheletie is ds teeysiet tat ote etelere ts 320 40) Causes andiconditions underlying heteromorphosis| .2- 22+.) sie) eel eien = lela ee ereiele aie 322 Rue CcenOsatce, wes WOGVeMents ang imtermal circulation © 7 eye sien «ofa stein olstaele ales ne Fol tiene aes 330 GMEStOlonsiOrrmiallOM eye sais) «21a y= re = faye eynis chee sfeye soca a os aca haterny Ghat Pe lara Gee Sia ees res eee 334 Fapate OtareceneratOnlewt of. «ajc ecisise cis seis atts aeeies moe Gracies Saidielem, see Moet acir ania Ae 335 REE CES OLA PTA VAL fete icliche tai oiola.ererase si cities te: sfohs te evteys al sueteesouelage re sin aleceneh slap tel sla Rea exre Lars 337 Gum tectsmolecontdche pr an irs sciiss Noemie cetasta dette anyanten sete hints ldc laden terest eters 341 POMmMIECIS OflACGOMOXY PEM stelel 1s /axsmckalaio sale niayoh.satetovapcts nlc felohere lorerale rar. alain EPS Shy eet 343 Mia itects) Okidinect sunlight st prep ge crete sie-ayatsicver unl ase wie aeie areiwreve wicuelsvrtewehe atexe nie Mere aT eo Toe 345 Ramp ects Of tempera CULE tr sweye ts isya.a exer eratelere scree renatonsTeap soa cr st veka oi vemerececaretsoors eee fe or ave ietere ser are 346 13 Effects of repeated removal of polyps from the same lateral branches................0++0++- 346 TA btrectsOminy Unies toraiterent parts onitherstemi etre ciys seen ee ern eeiere neni cinie annie 347 ES Litectsiof.diluted aud’ concentrated sea-watenj40 1; os erin ioe eee Bonny ses vee orene 348 MOBENI ie steht tate ans A Perea vere ene aNS toh eh ae setts SLATE a orc ATTA Atos SOE 353 PRELIMINARY STATEMENT Loeb’s pioneer experiments on regeneration in hydroids, have stimulated a large number of investigators to study the effects of external and internal factors in these animals, especially upon unbranched or slightly branched forms like Tubularia. Most hydroids are affected by the same agencies, but not to the same degree; that while gravity is the detcomsmne condition in one hydroid, contact or regional differences or “ polarity,’’ determines ‘I am deeply indebted to Prof. Thomas H. Morgan, who suggested these studies and who ren- dered much valuable advice and assistance to me throughout the course of these investigations. My thanks are due Prof. Edmund B. Wilson for the privilege of occupying the Columbia University Table at the Marine Laboratory at Wood’s Hole, Mass., and to Prof. C. W. Hargitt for many valuable suggestions. Tue JourNnaAL oF ExPERIMENTAL ZOOLOGY, VOL. IV, No. 3- 318 DiS Goldfarb the kind of regeneration in other hydroids. In the following study of Eudendrium ramosum, I have attempted to examine nearly all the known factors, external and internal, that enter into the life of this hydroid, especially those that take part during growth and regeneration. Eudendrium ramosum consists of one or more main stems, bear- ing pinnately arranged lateral branches which, in turn, branch again and again, hnally Eades in pedicels each bearing a polyp. When kept in an aquarium the polyps disappear and regenerate period- ically. A few preliminary experiments made it clear that the method,” previously used, of adding the number of hydranths regenerated on a stem, each successive day after amputation of the polyps, did not give an accurate idea of the actual number of different hydranths regenerated in a given time. If, for example, one or more hydranths should regenerate on a branch at about the same time that an equal number of other hydranths degener- ated, the records would not show the formation of new polyps. In the following experiments the exact number of different hydranths produced each day was recorded by the aid of daily diagrams of each stem and branch showing the presence or absence of polyps, buds and stolons. Hydranths appear within two or three days after amputation. Later some or all of the regenerated hydranths may disappear to be replaced in part or in whole by new hydranths; or other cut ends, devoid of hydranths, may regenerate them now for the first time. In order to condense into the smallest space the data essential to an understanding of the phenomena, the number of hydranths that appear within three and six days respectively, after the removal of polyps, are quoted in the following tables, unless specifically mentioned to the contrary. When fractions are used the numer- ator represents the number of new hydranths formed in the time stated; the denominator indicates the number of lateral branches or pedicels removed. For convenience these fractions are usually reduced to per cent. It is nearly impossible to obtain stems absolutely alike in all *Light as a Factor in the Regeneration of Hydroids: Goldfarb, Journ. Exp. Zodl., 1906. Factors in Regeneration 319 respects. For practical purposes, stems that resemble each other in size, number and size of branches, that come from similar regions in the colony, and that are removed from their habitat at the same time, will be called “similar stems.” EFFECTS PRODUCED BY REMOVING LATERAL BRANCHES OR PEDICELS Experiment 1. ‘This experiment was undertaken to determine whether stems, bearing lateral branches but with the pedicels and _ their polyps removed, would regenerate a greater or less per cent of polyps than stems with all the lateral branches trimmed off close to the main stem. On one side of a large stem the branches were cut off close to the main stem, while on the other side only the pedicels were removed. From a second stem the lateral branches were amputated on both sides; and from a third the polyps only were removed. ‘The records for each stem were as follows: TABLE 1 Regenerated in 3 days 6 days No. of stems Pedicels only Branches Pedicelsonly Branches removed removed removed removed 2 2 11 3 RecieksLemwi ei = etn) Wiel aveiasoliey sania tno sou Stee Mlometaiwtnlialtebulieyahahel sist |e *ie si te) (oi lel atin fel a! of) \s\ s,s, /0/o\"ejrs) 6 -o/'e il Ti a 4 5 ; 14 DOM Aiieeaibtl ohis)m (a fo Palka, la tele’ ‘wie) allel is} /ex'eletco)! sat'e fel si of seta Ya'le: 78 17 18 7 24 1 34 21 5> (ie VE 6 ho Sut GM ORS DEG Ae eee = 4 Ea 5a 17 Titel Begey are geen II OS eee tees 24% 35% 42% 717% * The figures for each side of the stem given separately. The conclusion is obvious, viz: that colonies from which all the branches have been removed regenerate more hydranths than those from which nothing but the pedicels and their hydranths were ampu- tated. ‘Yhis conclusion was corroborated by later experiments. All the lateral branches were removed from stems used in the succeeding experiments. 320 A. fF. Goldfarb EFFECTS DUE TO REGIONAL DIFFERENCES? Experiment 2. Is the tendency to regenerate polyps more strongly developed in one region of the stem than in another, or is the same average number produced in all regions of the same size! Pieces from a series of large stems were compared and the number of polyps produced in each was separately estimated, viz: (1) The basal end of a stem, about one-tenth of the whole stem, (2) the basal half of a second stem, (3) an entire third stem, (4) and the two halves of this stem separately considered. TABLE 2 Number of hydranths regenerated on Distal half of Basal half of Entire stem entire stem entire stem Basal half Basal tenth Ee —6_ =o 5 f 20 10 10 9 7 ies 3 0 0. 3 day Sie ernie s i6 z 8 T0 13 12 ee _0 24 12 12 10 Average, 36% 70% 3% 17% 0% ard i4 =3r 6 [ 20 10 10 9 i03t 9 2 4 Gydays..--- - | 44 2 z as 22 17 3 prs Tz Tz 10 fc) Average, 83% 133% 33% 44% 0% No regeneration occurred on the small basal pieces until the seventh day after amputation. Even then very few polyps ap- peared. ‘The basal halves regenerated 17 per cent, the entire stems much more, namely, 36 per cent. More striking, however, is the difference in the regenerative power of the basal and distal halves of entire stems, for 70 per cent regenerate on distal halves, but 3 per cent on basal halves. The figures for six days reinforce these conclusions. Smaller stems, however, do not reveal this sharp contrast in the regeneration of the two halves of stems. Experiment 3. Similar stems were cut into three nearly equal parts. ‘The distal thirds regenerated two days after amputation; most of the middle pieces did not regenerate till the third day, and the basal pieces, not till the third or fourth day. The question of ’Some very interesting facts in this connection are given by Gast and Godlewski in Die Regula- tionserscheinungen bei Pennaria cavolinii, Archiv f. Ent., Bd. 16, 1903. Factors in Regeneration 321 rate of development will be discussed later. For the present it will suffice to state that because of this difference in the rate of devel- opment, the latent period’ was not computed and the number of complete polyps produced two days after their first appearance (which may be the fifth or sixth day after amputation), and the number produced within the next three days, were recorded. TABLE 3 Regenerated Average on 2 days 5 days 2days 5 days Distal thirds 29594 8 cr SFA/ SE SER VEEL 66% — 100% Middle thirds $ $339 $ HSEE| FEE RR EEE 61% — 101% Basalthirdls 2 §344yoerSS4/2 FF FM EEE 2% 17% We may conclude that the distal and middle pieces regenerate practically the same number of hydranths but far in excess of the basal pieces. Experiment 4. The last experiment was modified to the extent of using not the main stem but the /ateral branches from the distal, middle and basal regions of the stem. ‘The distal branches were small and delicate, quite different from the middle and basal branches which owing to the smaller stems used, resembled each other closely. TABLE 4 Regeneration on 2 days 5 days Branches, from/distal-recion\. «2.0.14 2 nea eae 37% 57% Branches fromiuaiddlexepionyjcp-pe sey sec e ticle 50% 79% Branches tromibasalsregions a.isye'-)-t eee eae ee 68% 104% Regeneration of lateral branches differs at different levels. It is greatest on branches taken from the basal regions (of small stems) least on branches taken from distal regions. An expla- nation of these phenomena will be attempted under the caption of “‘coenosarc.”’ From the evidence already cited we may summarize as follows: T he distal half of a stem regenerates a much larger number of polyps than the corresponding basal half. On the contrary, branches ‘See Rate of Regeneration. 322 A. “f. Goldfarb from the apical region give rise to fewer polyps than the branches from the middle and basal regions. While polyps at the apical end of stems are common, they are rare on the apical ends of branches.*® HETEROMORPHOSIS The phenomenon of heteromorphosis® in hydroids, has been investigated by Driesch, Loeb, Morgan, Stevens and others, par- ticularly upon unbranched or slightly branched colonies. It was hoped that experiments upon a much branched hydroid like Eudendrium ramosum, would afford further insight concerning: 1 The conditions underlying the formation of heteromorphic polyps. 2 The effects of such polyps upon the regeneration of other polyps on the stems. 3. The rate of development at different levels of the stem. 4. The coenosarc, its movements and its effects in the produc- tion of heteromorphic polyps. A great many observations on large stems of all kinds had shown that polyps are rarely produced at the basal region, partic- ularly at the basal cut ends, though common enough at middle or dstal regions. Smaller stems or pieces of large stems—bear- ing from 15 to 20 branches—regenerated more basal hydranths than the much larger stems, though still less than the number of apical polyps. Experiment 5. Warge stems with lateral branches removed were cut into three nearly equal pieces. Further details are given in Experiment 3. The number of hydranths regenerated at the oral and the basal cut ends of each piece, two and five days after their first appearance, are given in the following table: 5See Gast and Godlewski, loc. cit. ®The following papers on Heteromorphosis and Polarity give various hypotheses to account for these phenomena in hydroids: Bickford °94 J. Mcrph.; Driesch ’92 Biol. Cent., ’96 Vierteljahrs-schr. Nat. Ges. Zurich, ’97 Archiv f. Ent., ’99 Archiy f. Ent.; Loeb ’91, Ueber Heteromorphose, Wiirzburg; Mor- gan ’or Biol. Bull., ’or Archiv f. Ent., 04, ’05, 06 Journ. Exp. Zoél.; Morgan and Stevens ’o4 Journ. Exp. Zodl.; Stevens ’o2 Archiv f. Ent. Factors in Regeneration 323 TABLE 5 Polyps regenerated on 2 days 5 days y foral ES yO Re Se PBN Hien bs INRIA PRIMES a ARE eat Be 60% 60% ee oecian \ basal ETN Se ener Ca eR ore Anny art Sa Re fees CA RE Le Ne 30% 100% PONG loxe (le A EN a Gee Auer mre deer be Behe eee a 80% 100% Middle third of st J Seco oustem \ basal Cite Se oaceaan borer ahs sy Mocw're esshajehepevoueze create case 40% 70% E fonaliend sy ere gsy-sawins eis etastoos + stele eyes ese ernie ele Palo aS 70% 100% Basal f st ‘ geal third ox stem \ basal (ats Bean aot oreieic Soran Optra unions Din tonae 20% 30% With distal pieces excepted, the branchless parts of a stem regen- erate decidedly more hydranths at the apical end than at the opposite or basal end. Experiment 6. The above results contrast sharply with those in this experiment in which only /ateral branches from different regions of the stem were used. TABLE 6 Polyps regenerated on 2 days 5 days jOralienda mets yeee wae be eh see ec % % Branches from distal part of stem Gal eae ee Se \ basal ONAS ele tetera pee imen Aare uemenas 37% 50% : : foral rica eect Ee ree ARs Ce: MNO Lat 12% 12% a oval ep ot stem \ basal ENVAS Dey age pct PETER ce Ee ae 62% 112% Iherd serch Bags. eter terse she ete % 37% Branches from basal part of stem Jeralends Soe Se \basalends ...............0.000. sees 37% 87% The lateral branches from any level of a stem produce a far greater number of heteromorphic than apical polyps.7_ Under the caption “ccenosarc”’ this will be explained. Experiment 7. In Experiment 6 all secondary branches were removed. In the following experiment these were not removed nor were the polyps amputated. ‘The regeneration at the basal and apical ends only was recorded. While the polyps were dis- integrating, stolons were forming at all the basal cut ends, fasten- ing the pieces to the bottom or sides of the dish. Later stolons were present on several pieces at both oral and basal ends. “These stolons often grew to a remarkable length as long as or longer than the original specimen. Usually they gave rise to several branching stolons which regenerated one or more polyps, even as ™Driesch, Morgan and Stevens found that the apical ends of pieces taken from any region were first to regenerate and produced a greater number of polyps than the basal ends. 324 A. ¥. Goldjarb many as nine polyps were present at the basal end of one piece. These hydranths did not appear before the third or even the fourth day after removal of the branches. ‘The following table gives some of the details of the experiment. TABLE 7 Polyps regener- No. of branches Polyps regenerated ated at oralends that regenerated at basal end of 3 days 6 days Io days in 10 days basal polyps Dystalibranches vances ° 80% 160% 0% 90% Middle branches .......... ° 0% 100% 10% 70% iBasalibranchesmeseeeneeee ° 50% 120% 20% 50% The presence of polyps on the branch retards regeneration but does not prevent the formation of polyps at the basal cut ends. The figures, particularly in the last column, of the above table seem to indicate a maximum tendency toward the production of heteromorphic polyps on distal branches, less and less on middle and basal branches, respectively. SUMMARY The mid and basal thirds of a stem behave quite differently from the distal third and from the lateral branches, in so far as the rela- tive number of hydranths regenerated at the oral and basal ends is concerned. The distal region of stems and the lateral branches from any region produce a greater number of heteromorphic and fewer apical polyps, than do the median and basal thirds of the stem. Lateral branches from which polyps had not been removed likewise produce more polyps at the basal than apical ends. Experiment 8. In the preceding experiments the presence of lateral free cut ends may have introduced disturbing factors not yet fully considered. In order to avoid these influences, some or all of the lateral cut ends were ligated ona series of similar stems as follows: a the apical end only; : b the apical end and the two lateral ends on distal half of stem; c the apical end and all the lateral ends; d_ all the lateral ends; Factors in Regeneration 325 e all the lateral ends and the middle of the stem; 7 the middle of the stem only; g all the lateral ends, then stem was cut into two equal parts; h not ligated—control stems. The rate of development in the different series varied consider- ably. In the table the number regenerated for two and five days after the first appearance of polyps is given. TABLE 8 Polyps reg. on in Oraliends= cr 82% 66% 87% 88% Lateral ends...... 54% 25% 52% 44% aidaysi... «. Basalends) 5.7... - 61% 715% 75% 100% 88% 100% 100% Total Average.. 66% 70% 62% 94% 70% 72% 100% Oraltendsi 05 <1: /1-1¢ 123% 116 125 155 Lateralends ..... 94% 100 66 76 RIGAYSic..*\)) Basalends oi...12)-10 102% 123 100 155 122 116) 118 Averdgearrsr re: 106% 124 108 155 94 106 118 It will be observed, firstly, that hydranths are normally regen- erated more readily at the oral rather than the basal ends of stems. Secondly, with the same number of ligatures to the stem, the num- ber of polyps regenerated depends on the location of the ligated branch or branches; that is to say, tying a lateral end near the -basal region of the stem is less efficacious in stimulating basal regeneration than a ligature on a lateral branch nearer the distal end, which in turn is less efficient than a ligature at the apical end. ‘Thirdly, the total number of polyps produced or at least the number of basal polyps, is in a general way proportional to the number of ends ligated; the more ends ligated the greater the total and basal regeneration. It should be recalled, however, that ligating four lateral branches is less effective than one apical liga- ture. Fourth, a ligature around the middle of a stem increases the number produced at both oral and basal ends but not to the 326 Wile We Goldfarb same degree, causing a slight increase at the oral end, and a con- siderable increase at the basal end. Fifth, the influence that a ligature exerts, does not, as a rule, affect the next lateral free end to the same degree as the axial free end, particularly the basal free end, even Hogan such axial end is not nearest to the ligated branch. Sixth, while the figures in the table are not absolute they show in a general way which arrangement and number of ligatures are more effective in bringing about an increased basal regeneration. Experiment g. We may now turn our attention to heteromor- phosis shown by very small pieces of stems and branches. The apical ends of a series of stems, including the distal two pedicels and their polyps, were removed. ‘The apical polyp was then cut off. In one lot the lateral branch was ligated, in the second, they were not ligated; in both the axial ends were equidistant from the lateral pedicel. Polyps appeared at one or both ends either simultaneously, or more frequently the basal end appeared one day before the oral hydranth. Sometimes stolons were produced at one or bothends. The basal ends regenerated more rapidly and in greater numbers than the oral ends of these pieces. ‘Vhe actual figures are as follows: TABLE 9 Polyps regenerated on 2 days 3 days 4 days Oral ends 7.getast sy. < eae eee: ° 17% 25% Basal ends yor precise ctor te eet ra 7 25% 45% The same results obtain when similar pieces taken from dif- ferent regions of the stem are used. About three times as many polyps are produced at the basal as at the oral ends in six days. These facts indicate that very small pieces of stems, regardless of the region from which such pieces are taken, tend to regenerate a far greater number of heteromorphic polyps and to produce these faster than oral polpys. The following three experiments g give further evidence concern- ing heteromorphic regeneration on small pieces from similar or different levels of stems. Experiment ro. ‘The apical ends of a number of stems and branches were cut beyond the distal node. ‘These pieces about Factors in Regeneration : 327 14 mm. long and bearing a distal polyp, were kept carefully ori- ented. ake third day after amputation the apical hydranths had disintegrated. Onthe fourth day polyps were first regenerated and appeared invariably at the basal ends. One lateral polyp only was produced, two oral polyps appeared on the seventh day, there- fore not mentioned in the table; in no case were polyps regenerated at both ends. ‘The following table gives the actual figures: TABLE 10 Polyps Regenerated ; 3 days 6 days on : (Oma GCS. hme tera ceed eee c SR te cac ee Sea TEE be, PAE fo) ° JERS eis (a ePrice teens Eeene nce ork ene SEL ame EH ee fo) 3% Bra salnenshprcn errant creya ae ocie et ve Venice eerste ie ° 44% When pieces of the same kind as the preceding were ligatured at their distal ends, regeneration was not accelerated, nor was the number of basal polyps increased. The figures are remarkably like those in the preceding, viz: TABLE 11 Polyps regenerated es sae on : @raltendsy.. se neers saa eon tes er SR eae amen ° 7% * Brasellterrdsretsr cscs aictcie Per evekc teres chee ads eto barsusgsie ee a ats ° 43% * The coenosarc withdrew completely out of perisarc and then formed an oral hydranth. This matter will be taken up under “coenosarc.”” When pieces about 14 mm. long (smaller than those in Experi- ment 9) were used, the number of heteromorphic polyps was prac- tically the same as in the experiment just mentioned, though in pro- portion to the number of oral polyps there was a far greater basal regeneration in the smaller pieces. Furthermore, an apical ligature did not accelerate regeneration nor did tt increase the number of basal polyps. Experiment 11. ‘This tendency to produce more polyps at the basal ends is also clearly demonstrated with internodes of stems. ‘The internodes of several stems were cut off and oriented, and so arranged that it was possible at a glance to tell to which part of the stem each internode belonged. All doubtful pieces were rejected. The 55 internodes regenerated as follows: 328 A. Ff. Goldfarb TABLE 12 rps d Polyps regenerate days ise on Oraliend Sede sic ake hoe Cee aearere rete re 9% : 39% Basalends. cot sdteciae a alecolrer tater at Serene ofa rat eens ote 16% 52% In these pieces, each about 1 mm. long, more basal polyps are produced. A considerable number of internodes formed hy- dranths at each end, in some casessimultaneously. ‘This fact indi- cates that some of these very short pieces had the potency to pro- duce more than one hydranth. The majority regenerated but a single polyp. Regional differences were not apparent in this experiment. It cannot be said, for example, that regeneration was retarded or accelerated, increased or decreased in one region more than in another; nor was the formation of basal hydranths pecu- liar to any one region or level. From these experiments we conclude that in the small pieces mentioned, a greater number of heteromorphic than apical polyps are produced, that this increase 1s not associated with any particular level of the stem, that oral polyps are rarely formed on distal pieces, though many regenerate on pieces from middle and basal regions of the stem. Experiment 12. ‘Yo determine whether rapid growth first inone direction then in the opposite direction, could be effected, the new basal stems from a number of pieces were removed, viz: B pieces of the diagram. ‘There developed from the basal end b, of these basal stems many hydranths which grew with remark- able celerity in the original direction toward a; there also regen- erated actively polyps at the oral (c) and the lateral ends of the B pieces. [here was no reason to doubt that the removal of the C pieces would result in the formation of hydranths at the new basal end of C. If lateral branches instead of basal branches be repeatedly removed the results are essentially the same. The actual figures are given below: TABLE 13 Polyps regenerated on basal pieces of B 3 days 6 days 9 days Oralfendsis.. nite cian a ore oe etn ToT: o% 26% 53% Basaliends<;s% oer msie%, a e¥atnsee sls sssteeta snsenere 33% 40% 53% Polyps regenerated on basal ends of lateral branches Orallendstas venta cake eee eee ae 6% 36% 46% oF ¢ Basalend sive seen csctasestusiece om ie coe eiciw sestere aol 0% 239 Factors in Regeneration 329 The above tables give the number regenerated at the basal ends of pieces 3, 6 and g days after amputation. ‘The oral (and lateral polyps also) had not been cut off but disappeared within two days. Regeneration, therefore, at oral and lateral ends could not take place till one or two days after the basal polyps appeared. Regeneration in one direction, viz: from the basal end of B piece does not inhibit regeneration in the opposite direction at c, rapid growth and differentiation take place synchronously in op posite Poot ecis Corcl ena VA valle Peon ltasc pecans at P Staaealacaicena ae es pels © aval end pe oe Lo Fig. 1 directions. “Vhis would lend weight to the view that regeneration is not dependent on internal changes of the entire stem or even parts of the stem but is rather due to reactions at the cut end only. An examination of Tables 14 and 15 will make itclearat a glance that gravity is a considerable factor in determining the increase or decrease of the number of hydranths at the upper ends (toward the zenith), regardless whether such ends be oral or basal. Erect 8 Except for minute pieces, see also Experiment 25 on this point. 330 A. f. Goldfarb stems regenerate in 3 days, 50 per cent at the oral and 45 per cent at basal ends. Inverted stems—with basal ends pointing toward the zenith—in the same time, regenerates 35 per cent at the oral and 60 per cent at the basal ends. ‘The figures for 6 days empha- size the point, viz: Erect stems regenerate 50 per cent at the oral, 8o per cent at basal ends, inverted stems 35 per cent at oral and 115 per cent at basal ends. Injury to different parts of a stem does not affect the regenera- tion at the axial ends of the stem. Nor does increasingly diluted or concentrated sea-water influence the number of hydranths at basal or oral ends. Heteromorphosis occurs independently of these influences. COENOSARC® The coenosare may readily be observed in the more distal parts of a colony, for the perisarc there is thin, almost transparent and seldom covered with débris, polyzoans or hydroids, commonly found on the more basal regions. ‘The coenosarc is com- posed of a hollow cylinder surrounded by the perisarc. “lhe lumen, within the ccenosarc is cir- cular in cross section and differs from Tubularia”® in which the lumen 1s divided into two almost sepa- rate compartments by a central partition along the whole length of the stem. It has already been pointed out that the number of hydranths regenerated at the oral or at the basal ends, depends upon whether lateral branches or stems were used, and if the latter, upon the re- gion from which the piece was taken. Previous experiments and direct observations prove conclu- sively that the cenosarc withdraws from the distal *Gast and Godlewski’s account of the ccenosarc in Pennaria cavolinii is extremely interesting in this connection. Hargitt, G. T.,’o3 gives an account of the histologic structure of the coenosarc. Archiv fiir Ent.,. Bd. 17. 10Stevens, N. M.,’o2 records movements of the ccenosarc in Archiv f. Ent., Bd. 15. Stevens ’oI, 02, Archiv f. Ent., Bd. 13 and 15. Factors in Regeneration 331 ends of all branches, cut at any level, and from the distal ends of pieces from the apical region of the stem. The cenosarc also withdraws, as a rule, from the distal ends of small pteces from any part of the colony. In the diagram of a large stem a represents the oral, / the basal end. We can foretell with a fair degree of accuracy whether polyps will be formed at one or the other ends of pieces and the relative number regenerated at each end, provided we know the level at which the cuts were made. In pieces ) c, no hydranths will appear at 6, many at c; in pieces f h or g 7 more hydranths will appear at f and g, respectively, than at fh andz. In region af the nearer to a the oral end of a piece lies the fewer the oral polyps; in region j / the nearer to / the basal end of a piece lies the fewer the basal hydranths regenerated. The recession of the ccenosare from the distal end may extend only one internode or half a dozen or more, and the hollow peri- sarc thus produced often cracks and breaks off." In small pieces the coenosarc may move not only toward the basal end but through the basal end entirely free from the perisarc, leaving the empty perisarc behind. Placing inverted stems in sand accelerates the basal movement of the coenosarc, so that the parts embedded in sand become entirely empty; the ccenosarc is found only in the basal regions surrounded by water. In erect stems under the same conditions the coenosare withdraws somewhat from the dis- tal end while at the basal end it either (1) does not withdraw at all, in the majority of cases, and in spite of the adverse conditions, (2) or slightly withdraws, (3) or disintegrates, the result of the ravages of large numbers of ciliate protozoa. Two counteracting tendencies may be said to be present at every cut end of a stem or branch, the resultant of which deter- mines whether a hydranth will or will not be regenerated, and whether regeneration will or will not be retarded; first, the movements of the ccenosare from the distal cut ends already described, second—and | believe second in point of time—the regeneration of new tissue, which is negatively geotropic and, UGast and Godlewski, loc. cit. 332 A. “f. Goldfarb therefore, tends to grow upward. ‘The more ccenosarc present near the cut end, as determined by its width and density, the greater the regeneration. The less impeded, the greater the movement of the coenosarc. Whether enough ccenosarc is pres- ent or 1s regenerated, to make up for the recession of the coenosarc will determine whether or no hydranths will appear at the cut end. ‘The basal movement brings additional ccenosarc to the basal end, condenses it greatly and thereby increases the number of basal polyps, on small pieces, on lateral branches and on pieces from the distal region of stems. The cenosarc in the middle and basal parts of the stem does not withdraw from the oral cut end and, as a matter of fact, there are more hydranths at the oral than basalends. In this hydroid atleast, we donot need to call to our aid “formative stuffs,”!? and other hypothetical internal forces and materials to account for heteromorphosis, for the movements of the ccenosare account for the presence or absence of polpys at some levels and not at others. The coenosarc of stems kept long in the aquarium often becomes fragmented. Fragmentation is the result of a splitting of the coenosarc at one or another of the internodes; the ccenosarc thins rapidly at these points, and finally breaks into pieces entirely independent of each other. Each part may move into the nearest lateral branch and give rise to polyps, or it may remain in the stem. Not infrequently in the latter case it contracts at both ends ulti- mately forming an ellipsoidal dense mass of coenosarc near the basal end of the stem. It is believed that these hydroids winter over in this contracted condition. When stems are subjected for a long time to adverse conditions the ccenosarc forms this dense ellipsoidal mass. When such stems, which could no longer be made to regenerate, were cut into smaller pieces, polyps regen- erated provided the coenosarc was injured. With a low power of the microscope the lumen within the coeno- sarc 1s seen to be filled with a colorless fluid in which myriads of colorless granules float. These move slowly toward one end of “The hypothesis of specific stuffs, moving in definite directions, was developed by Bonnet, later by Sachs, and still further perfected by Loeb. For criticism, see Morgan ’o1 Archiv f. Ent., Bd. 11, *o2 ibid., °o4 Journ. Exp. Zodl. Factors in Regeneration 333 the stem; in pieces 1} inch long the trip from one to the opposite end takes from 14 to 3 minutes. As the granules accumulate at this end the stream moves more and more slowly until it finally ceases. There may be a respite of a few minutes and then the stream courses in the opposite direction at first slowly then faster and faster, and finally slowly again as the granules pile up at the other end. The performance is repeated over and over again; the time for each trip may vary considerably. Now and then groups of granules are violently whisked about or “tremble,” the result of ciliary movement of the endoderm. Granules were never seen to pass out of the prostomium of the living polyp. After disintegra- tion of the polyps dark red masses were frequently observed marking the spot where the polyps had been. The current in a lateral branch may be continuous with that in the main stem, moving at the same rate and in the same direc- tion, or it may be independent, and even contrary to the stream in the stem. The central stream may be continuous or divided into two or more independent streams. “Though the ccenosarc in branches and stem is continuous, the streams within the coenosarc of these parts may behave independently. ‘The contin- uity of the coenosarc may be permanently broken by pressing a stem firmly with the side of a needle. The ccenosarc is cut into two parts which separate more and more from each other. If gently pressed, a dent is temporarily produced in the coenosarc which slowly recovers its normal shape. The number of hydranths that arise from the basal cut end of pieces is closely associated with the amount of coenosare near such ends. Other conditions being equal the more ccenosarc at or near a cut end the more hydranths produced. Several condi- tions must, however, be taken into account. A large stem does not necessarily contain more coenosare per unit of length than another one-half as long. Much depends on the more or less con- tracted condition of the ccenosarc. If it be attenuated, as in rapidly growing branches, it will have per unit of length less regen- erative potency than the more concentrated coenosarc usually These dark red bodies are probably analogous with the red bodies resulting from the metabolic changes in Tubularia, studied by Bickford ’94; Stevens ’o1 and ’o2, and Loeb ’91, and Morgan loc. cit. 334 A. “Ff. Goldfarb found near the basal region of stems. Careful measurements of the diameter of the ccenosarc at different levels of large stems actually shows that normally the diameter of the coenosarc de- creases toward the distal end and conversely increases basally. The relation between size and the number of hydranths regen- erated, particularly at the basal end is shown by the following ob- servation. Pieces less than 1 mm. long never produced a complete polyp, though they often regenerated shoots at one or both ends. Pieces as small as two-fifths mm. long developed shoots at one end. Larger pieces 1 to 2 mm. long may regenerate one polyp at each end though usually at the basal end only. Still larger pieces regenerated two or three polyps from one basal end, whereas larger median or basal pieces paauced as many as nine polyps from a single basal end. STOLON FORMATION Little has been said concerning stolon formation, partly because of its comparative rarity, partly because stolons are often with difhculty distinguished from branches. A stolon,* root, or hydro- thiza is an outgrowth, positively geotropic and stereotropic in its reaction, which, when young, fastens itself by a sticky secretion to solid objects, and which does not directly give rise to polyps. In nature the stolon or stolon system anchors the hydroid. Less frequently, stolons may join two stems and sometimes the cceno- sarc of stolon and stem may fuse. In the laboratory, stolons may appear at any of the cut ends. Though most frequently observed at the basal, they may appear at any lateral or even oral end, singly or in groups of branching stolons. They may appear simultaneously at the oral and basal ends, or at one end only. The stolon may sometimes grow to a great length; in one instance a stem 27 mm. long regenerated a basal stolon 40 mm. long. Stolons may give rise to lateral branches usually pinnately arranged, which like pedicels end in hydranths. Some un- branched stolons after a time bend at their very ends and regen- The production of stolons has been shown in some species to be determined by gravity (Loeb ’gr), in others by regional peculiarities (Stevens ’o2), contact (Loeb), exhaustion (Driesch), by the kind of regeneration at the opposite end of piece (Morgan ’or). 15See Stevens ’o2. Factors in Regeneration 335 erate a polyp at the tip. In these, it is impossible to tell where the stolon ends and where branch begins. It is nearly always dificult to tell in advance whether a large growing shoot will ultimately become a branch or a stolon. Stolons in my experi- ments never regenerated stolons; when cut, only hydranths were produced at the cut ends, irrespective of the level at which the cuts were made. When stems are subjected to adverse conditions, the cut ends may regenerate new tissue, which becomes surrounded by a sticky perisarc, which does not differentiate into perfect hydranths. The new tissue is really a modified stem, which under these adverse conditions may increase in length and is then called a “stolon;” when it is amputated or when it grows into a favorable environ- ment the distal end frequently differentiates into a hydranth. When a large number of stems were placed in a shallow dish of water containing much débris, and the water was left undisturbed for many days, very few polyps appeared, while a remarkably large number of stolons were produced. When the water was frequently changed, however, the “stolons” invariably bore polyps. As the colony grows older, the perisarc of the stolons becomes thicker, less plastic and encrusted with débris. If the conditions remain constant, the “stolon”’ functions permanently as an anchor- ing organ and can no longer of itself produce polyps, though it has the ability to do so, if cut and removed to a favorable environ- ment. “Stolons” are not limited to any particular region nor is their formation influenced by gravity, size of the piece, kind of regeneration at opposite end, etc. ‘Their presence or absence, in this hydroid at least, is not an indication of the presence or absence of certain internal changes and is, therefore, useless as an index of the polarity of the stem. RATE OF REGENERATION" The term rate is here used to designate the interval between amputation and that point in the differentiation of the regenerating 1®Loeb (’91) determined rate of development and rate of growth for different hydroids. Morgan and Stevens ’o04, Journ. Exp. Zodl., made careful observations on the rate of development at basal and oral ends of pieces. 336 A. f. Goldfarb tissues at which the tentacles of the new hydranth are clearly discerned. Under normal conditions two days is required. Un- der adverse conditions regeneration may not take place for four, five, six or more days. Regeneration does not extend over this entire period, at least in so far as visible changes are con- cerned. ‘There 1s a latent period, during which no visible changes obtain and which normally covers but a few hours, but which under unwholesome conditions may extend over several days. Once ee begins dev elopment proceeds normally in about 12 hours. So that Saves conditions increase not the actual period of regeneration but this latent period. In large stems the younger (distal) portions always regenerate polyps at least one day before the older (basal) parts. When stems were cut into three equal parts, Experiment 3, new polyps appeared on the different pieces in the following order: TABLE 14 Polyps regenerated on 2d day 3d day 4th day Distalbhirdsty a2 r/c soe. seer te eee 8 polyps 1 additional polyp 1 additional polyp Middletthinds: ¥.2.- sew omnes somes ane fo) 9 polyps 1 additional polyp Basal thirds) vac eresjsicecece: am ae eer * fo) 6 polyps 4 additional polyps T he distal thirds regenerated polyps earlier than the middle pieces which in turn regenerated polyps before the basal thirds. ‘To determine whether the two cut ends at any level of a stem regenerate at the same time, the above pieces of stems were used. A, B,C represents the distal, middle and basal pieces, re- spectively. At the cut 5 c, bis the basal end of the distal b piece, and c the distal end of the basal piece. It was found that the b ends on 4 pieces regenerated on the average in practically the same time as the c ends on B pieces, and similarly for the b and c ends on the B and C pieces. It ¢ was found that the b ends regenerate first, just as often C gq 3s the c ends and as often as the b and c ends regenerated simultaneously. Although the distal region of a large stem regenerates polyps within two days or at least one day before the basal region, yet small pieces cut from the distal region (Experiment 10) never pro- duce polyps before the third day and often not till the fourth day. Factors in Regeneration 337 This retardation occurs irrespective of the region of the stem from which the pieces were taken, and is due solely to the small size of the pieces. Medium size pieces (with 10 to 15 lateral branches) regenerate polyps at all the cut ends, including oral and basal ends, in approx- imately the same time. Ligating the distal end or ends acceler- ates basal regeneration. Ligating the distal end of small pieces taken from the distal region of a stem, does not however accelerate basal regeneration, because the ccenosarc withdraws from the distal end and the ligature does not affect it. When the single lateral branch of small pieces (Experiment g) was ligated the rate of development was uninfluenced. Regeneration on small old (basal) pieces taken from large stems, is very slow; it may take six or seven days before polyps appear. Embedding inverted stems in sand and to a less degree suspend- ing them, stimulates the early formation of heteromorphic polyps, but the rate of development at other cut ends was not at all or but slightly affected. Lack of oxygen or low temperature or contact with a solid body or greatly diluted or slightly concentrated sea- water retards the development of the first formed polyps. In the meantime the stems become more or less acclimatized to the new conditions and polyps thereafter are regenerated at a normal rate. Injuries of various kinds, such as lacerating and slitting the stems in many places or disintegration of coenosarc in some of the lateral branches, does not retard regeneration at the other cut ends. All efforts to accelerate the normal period of regenera- tion in less than two days, failed. EFFECTS OF GRAVITY” Experiment 13. A series of stems were suspended vertically in a dish of water, some with the distal ends pointing upward (toward the zenith), the “erect’’ stems; others in the contrary direction, the “inverted” stems. The controls were placed hori- “Loeb (’91) believed that gravity and light were the external factors that determined the kind and direction of growth. Also see Driesch (’99). 338 A. Ff. Goldfarb zontally on the bottom of the dish. ‘The rate of development was practically the same in the three groups. TABLE 15 Polyps regenerated in 3 days . Regenerated in 6 days on Oral Lateral Aboral Total Oral Lateral Aboral Total LPAUE SUS. So Son gdocgnTaNoe 50% 54% 45% 53% 50% 86% 80% 83% Inverted Stems 5:5..5.02522: 35% 43% 60% 447% 35% 75%: 185%: wr5e0 Control Stems (eae eal 33% 37% 44% 37% 33% 48% 66% 48% (In Contact) INosof branches pisedt accesso eee eine er aiarand, siee ANNE relate snenenaretareerauarceaeters 574 An analysis of these figures gives some interesting details. In the first place, erect stems regenerate a greater total of polyps than the corresponding inverted stems. Secondly, erect stems produce a greater number of oral and a greater number of lateral polyps than do the inverted stems. The lateral branches of inverted stems that bear polyps bend upward toward the basal end of the stem, and these branches are invariably longer than on the erect stems. Thirdly, by far the greatest number of basal polyps are produced on inverted stems, from which they grow upward, and directly opposite to the rest of the stem. If the basal branch of erect stems are long they too bend upward. No emphasis is laid on the control stems in this experiment, as they were influenced by contact with the dish, a disturbing factor at the time not fully appreciated. | Experiment 14. Fine sand thoroughly cleaned was put in a dish of water. Erect and inverted stems were embedded in the sand to varying depths. The controls rested horizontally on the sand. ‘The diagrams in the table show the position of the stems; the horizontal lines represent the level of the sand, the parts below it represent the parts embedded in sand, the parts above the line, surrounded by water. The conclusions from the previous experiment were more than corroborated. Regeneration on inverted lateral branches is largely inhibited. ‘This is strikingly illustrated, on comparison of columns 2 and 4 and more particularly columns 3 and 5, in which the same number of lateral branches are free, but erect in the one series, inverted in the other. The erect stems of series 4 Factors in Regeneration 339 regenerates 33 per cent, the corresponding inverted stems of series 2 regenerates 8 per cent. Similarly, series 5 produces 36 per cent, the inverted ones 7 per cent. The branches in the two sets are too close together to warrant the belief that the difference in posi- tion on the stems can account for the large difference in regenera- tion. When inverted lateral branches do regenerate polyps, they invariable turn upward. Embedding stems in sand affects regeneration just as a ligature tied around the stem at the level of the sand would do, for in both TABLE 16 2 Jj y s 4 Polyps reg. on * : in Oraliendsiaeeneee % % % 0% 42% 8% Lateralends ...... 8 7 33 36 13 3 days... { Aboralends....... 33 58 66 25 otal eer cttrrr 33 24 18 28 37 14 Oraliend serreryteret= 50 58 33 Lateral ends ...... 38 21 41 55 28 Aboral ends ....... 74 140 158 58 6 days... — — — — — — Alia Gallleeretetereteieyersts 74 73 48 44 56 35 Wotalimumiber*of stems: aeressrrsicce hats sek ort osiessrarere Sil elo) hess reveal sae exe 72 Motalnumber oh branches:.raatstesia1-teiselcletlerel veiialeiateral retefolsiereveislalsie) 281 cases regeneration at the free axial end (oral or basal) of the stem is stimulated. The number regenerated at the upper ends of these stems depends on whether erect or inverted stems were used. Regeneration at the basal ends of inverted stems is very much greater than that at lateral or oral ends, forexample, 58 per cent are produced at the basalends,8 percentat the lateral ends, 0 per cent at oral ends, in one series. Again 66 per centregenerated at basal ends, 7 per cent at lateral and 42 per cent at oral ends of a second series. Furthermore, the total number of polyps produced ona stem 340 A. F. Goldfarb depends on how much of it is embedded in sand. ‘The deeper the inverted stems are embedded, the proportionally larger total num- ber is regenerated and vice versa. On the contrary, the deeper erect stems are embedded the proportionally fewer polyps produced. Experiment 15. These conclusions were corroborated and extended by the following experiment in which the sand was banked at an angle of 45° to the horizontal. TABLE 17 ne MT ae Ss Polyps reg. on in ( Oraliendsen 25% % 25% 50% % 66% 50% | Lateralends .... 54 80 75 75 fe) 66 56 3 days... { Aboralends..... 50 100 50 100 83 Vo otal i eis 50 83 TBs 7o 50 66 57 ( Oral’ ends: 22: iXe) 50 75 66 83 Lateralends .... 79 96 94 122 ° 83 92 6 days ... 4 Aboralends..... 50 150 100 100 116 Motalr erst 75 go 95 116 50 77 93 SUMMARY. r Hydranth bearing branches turn toward the zenith whatever the position of the stem, whether erect, inverted or inclined. Erect stems embedded in sand regenerate the largest total number of polyps, inclined stems less and inverted the least number. Regen- eration at the oral ends and lateral ends of similar inverted stems 1s largely inhibited, but at the basal ends 1s remarkably stimulated. It will be recalled that the coenosarec withdraws from the distal cut ends of erect stems; but a ligature or its equivalent embedding in sand, causes the coenosarc to move upward and regenerate an oral polyp. On inverted stems the ccoenosare under these influences, and particularly under the influence of gravity, with- draws not only from the distal end but from nearly all the lateral ends, toward the upturned basal end, resulting in an immense Factors in Regeneration 341 basal and almost 7z/ lateral regeneration. Furthermore, the total number produced varies with the amount of free ccenosare (not embedded in sand). EFFECTS OF CONTACT!’ Experiment 16. Stems were suspended on two horizontal threads, others were in contact with the bottom of the dish. Hy- dranths appeared on the two series at the same time. TABLE 18 Polyps regenerated on 3 days 6 days No of branches SUSpendedustemismicrae wcrc iietiysrerectes 46% 68% 248 Stemsionibottomlondishhs areata rata 41% 48% III This experiment agrees with No. 13, in that stems surrounded by water and otherwise under identical conditions produce a larger number of polyps than those in contact with a solid body. Experiment 17. Shallow V-shaped grooves were made in a cake of parafhn of such a width that when stems were placed therein the lateral branches were in contact with the sides or bottom of the grooves. A record of each lateral end was made, whether (a) it pointed upward and out of the groove and therefore not in con- tact with the parafhn, (b) it extended sideways and touched the sides of the groove, (c) it pointed downward, and therefore in con- tact with sides or bottom of the groove. In six days only 23 per cent of all the ends regenerated polyps. Further details will more clearly illustrate to what degree contact suppresses hydranth for- mation. During the first six days but 7 per cent of all downward pointing branches (in contact), 8 per cent of all sideways pointing branches (in contact), 43 per cent of all upward pointing branches (free) regenerated polyps. “Two hundred and thirty-six branches were used in this experiment. Experiment 18. Stems were put within glass tubes, 1 to 14 mm. inside diameter. One or both apices of some stems extended beyond the tube, in others several branches protruded beyond the tube, or short glass rings were so arranged that some of the lateral branches were free between the rings. In nearly every instance **Some hydroids react more quickly and more readily to contact than do other hydroids, Campan- ularia and Pennaria more than Eudendrium. See Loeb ’g1. 342 A. f. Goldfarb the lateral branches within the tube were in contact with the glass. The results were very decisive. TABLE 18 No. branches Polyps regenerated on 3 days 6 days removed @utiends: withinstheltubeaas-. ose aes o% fo) 88 @utiendssoutsidcofitubes ee saeenee sae een 56% 116% 34 In no case did regeneration occur on cut ends within the glass tubing. After six days these stems were removed from their tubes, and regeneration then proceeded normally at nearly all the cut ends. Experiment 19. Other experiments under somewhat different conditions illustrate further the inhibition due to contact. Each stem was rolled in a thick sheet of cotton, which was then immersed in water. No hydranths regenerated. Cotton abouthalf as thick was used and several hydranths appeared in three days, but by the sixth day they were gone and did not reappear thereafter. By using cotton one-half as thin again, a still larger number of polyps were produced, which disappeared less rapidly than in the pre- ceding case. ‘The results were similar when stems were placed between two flat layers of cotton, so arranged that a stream of sea-water continually flowed through the cotton. When removed from the cotton, the stems regenerated readily—provided they had not been kept too long in it. Experiment 20. On one side of a dish of sea-water stems were loosely placed, on the other side a larger number of stems were crowded together into a groove 4 mm. wide by 5 mm. deep and somewhat longer than the length of the stems. Table 20 gives the detailed results. TABLE 20 Indicates the total number of polyps observed on the following days 3d 4th sth 6th 7th 8th oth roth 11th No.of stems 1 Crowded stems........ 2 2 I 2 ° ° 2 I 2 10 2 Not crowded stems .... 10 23 10 2 3 6 6 9 16 6 qCrowded'stems!..5-.-).- 20 15 7 4 4 6 10 12 39\ . 12 4 Not crowded stems..... 40 35 20 27 36 22 20 20 19f 7 *Stems had been scattered. Thecrowded stems regenerated only at the free unentwined branches on the top of the pile, and regeneration was less than in Factors in Regeneration 343 the scattered stems. When the crowded stems were removed from the groove and separated from one another, there was a mark- edly increased regeneration. Experiment 21. Dr. Louis Murbach kindly permitted me to use an ingenious apparatus devised by him, by means of which a stream of air bubbles was introduced at the bottom of a vessel so that the contained water was in constant agitation. Stems placed in the vessels were whirled around making about 35 revo- lutions per minute in one vessel and about 28 in the other. Very little regeneration occurred (about 8 per cent), somewhat more in the slower stream, slightly less in the faster one. “Two days after the appearance of polyps they were gone, and no more regeneration occurred. If the current of bubbles was stopped for 12 to 24 hours, some regeneration would take place. After ten days the stream was stopped altogether, and there resulted a constantly increasing number of hydranths. Whirling stems through water at a com- paratively rapid rate affects regeneration in practically the same manner as contact, already discussed. Contact 1s unfavorable to and more or less suppresses the develo p- ment of polyps. It matters little whether branches touch each other, or collide with a solid object, or whether contact 1s due to growth within a confined space. In the latter case there is an increasing pressure proportional to the amount of growth. Contact may be reinforced by pressure resulting from the weight of a superim- posed layer of wet cotton; or contact may be the impact result- ing from the whirling of stems through water. Whatever the nature of the contact or pressure or weight or impact, regeneration of polyps 1s inhibited in proportion to the degree of pressure, weight, impact, etc. EFFECTS OF LACK OF OXYGEN’® Experiment 22. Sea-water was boiled to remove the oxygen more or less completely from it and the amount of water evapo- rated was replaced by an equal quantity of boiled tap water. An equal number of stems was placed in flasks filled to the brim with this deoxygenated sea-water. The smallest possible space Loeb ’g1 and 795, Pfliiger’s Archiy., vol. 62. 344 A. “f. Goldfarb was left between the water and the cork, then the flask was hermete ically sealed. ‘The following tables indicate in per cent the vol- ume of water after boiling. ‘These figures will serve in a general way to indicate the amount of oxygen removed. TABLE 20a No. of hydranths observed on the following days No. of stems 2d 3d 4th sth 6th 7th 8th ogth roth 100 1 Norm sea-water in Opentdishy ase: 40 18 8 7 8 II 13 8 6 2 Norm sea-water in Sedledsyessel peri iaers 14 13 12 3 5 6 5 3 to 10; from 97% to77% o ° ° ° ° ° ° ° ° T here was absolutely no sign of regeneration in the eight sealed flasks, Nos. 3 to 10 inclusive, from which the oxygen had been removed from the sea-water and its reabsorption prevented. Flask No. 2 which was also hermetically sealed, contained normal sea-water, and regenerated fewer polyps than the open dish No. 1. Regeneration in the former ceased altogether after eight days, while in the latter polyps continued to be produced. Regen- eration in the hermetically sealed flask No. 2 was inhibited, either because the supply of oxygen in the water was entirely appropriated by the previously developed polyps or because of the carbon dioxide produced by these polyps, or by both of these causes acting together. Additional evidence of a very interesting nature was obtained by filling wide mouthed jars about two-thirds full of oxygen—free sea-water as above. ‘There was, however, a layer of air over one inch deep between the water and the cover, which was sealed air- tight. TABLE 2o0b No. No.of min- Evap- No. of hydranths observed on the following days utes boiled oratedto 3d 4th 5th 6th 7th gth oth sith 12th 13th 14th I control control 8 2 3 7 18 7 7 2 3 3 4 2 4 98% 2 fo) 9 II Fl 2 3 2 2 I 2 3 8 82h 4 9 7 4 2 4 3 3 5 6 4 10 81 I fe) fo) 3 8 2) 3 ° ° ° ° 5 123 75 ° fe) 2 ° ° I I ° ° ° ° 6 15 73% I fe) I 6 5 7 8 2 2 2 ° 7) 174 70° fo} fo) 2 2 2 3 2 2 I ° ° * Probably an error. Factors in Regeneration 345 Polyps were produced in all the jars. ‘The greatest number appeared in the control dish No. 1. The previous experiment showed that no polyps are produced in water from which the oxy- gen had been removed. But in this experiment the water reab- sorbed enough oxygen from the overlying layer of air to supply the needs of the developing polyps. Where large quantities of oxygen had been removed from sea-water as in Nos: 4,°5, 6 and 7 a much longer time was required to absorb enough oxygen to permit regeneration to begin; and the available supply of oxygen was more quickly exhausted in these than in Nos. 2 and 3, therefore, regeneration ceased earlier in the former series than in the latter. It follows that where much oxygen has been removed from sea-water, polyps appear later and disappear earlier than on stems kept in water containing more oxygen. EFFECTS OF DIRECT SUNLIGHT”® Experiment 23. This experiment unfortunately was not carried to completion. “The apparatus was so placed that the sun shone directly upon the stems for about eight hours daily. The heat of the sun was guarded against by reducing radiation from the table and fixtures to a minimum and by surrounding the dishes by large volumes of water to which ice was sometimes added. Temperature records of each of the dishes were made at least three times daily. The stems were grouped as follows: 1 Dish was not guarded against the heat of the sun. 2. Dish was surrounded by 3000 cc. of water. 3. Dish was surrounded by gooo cc. of water. I, 2 and 3 were exposed to the direct rays of the sun. 4 Not exposed to the light, but kept in the shade of the room. 5 Not exposed to the light, but kept in a dark chamber. The temperature in 1 was, of course, several degrees higher than in any of the other dishes, especially about midday, when the temperature was often as high as 28° C. In 2, the tempera- ture was lower, while in 3, 4 and 5, which were practically the same, the temperature was lowest. *"Loeb ’92, ’96; Driesch Zool. Jahrb. ’g0; Goldfarb ’o6. 346 A. “f. Goldfarb The greatest regeneration occurred in 1, then came 2 and 3, which regenerated about the same number and less than 1; there was a decided drop in 4 and 5 which regenerated least. “There was a notable exception in one of the dishes of series 5 in which a surprisingly large number of polyps were produced. Durect sunlight stimulates stems to increased regeneration of polyps, even though the temperature of the water in which they are contained rises as high as 28°C. ‘There is, furthermore, an undoubted posi- tively phototropic bending of some of the new polyp-bearing branches. EFFECTS OF TEMPERATURE”! Experiment 24. In the previous experiments bacterial increase was guarded against, particularly in the higher temperatures, and hydranths prospered in a temperature as high as 28° C. Without this precaution such high temperatures would be fatal to the stems. At what higher temperature polyps would be regen- erated was not determined. Up to a certain point the greater the warmth the greater the regeneration. When stems were placed in a refrigerator in which the tempera- ture varied from 10° to 16° C. there was no mistaking the inhibi- tory effects of the cold. A large number of stems never regener- ated at all, and the total number produced was exceedingly small. EFFECTS OF REPEATED REMOVAL OF POLYPS FROM THE SAME LATERAL BRANCHES” Experiment 25. As soon as polyps were distinctly differentiated at the cut ends, they were again amputated. ‘This daily removal of poylps was carried on for a period of 31 days, and daily records made of the number and position of the polyps removed. During this time there were regenerated: 1 At 15 cut ends, including oral and basal, 59 different polyps; 2 At 11 cut ends, including oral and basal, 28 different polyps; 3 At 15 cut ends, including oral and basal, 64 different polyps, making a total of 41 cut ends regenerating 151 new polyps or 368 per cent in 31 days. *1Peebles ’98, on The Effects of Temperature on Reg. of Hydra, Zool. Bull. *»Hargitt, G. T., ’03, Reg. in Hydromeduse, Archiv f. Ent. ’03. Factors in Regeneration 347 Many cut ends regenerated but once during this entire period, others as many as ten times. The greater or less regeneration was not confined to any definite region of the stems. ‘The oral ends rarely produced any polyps for reasons already given. The stems continued to regenerate polyps normal in every regard, until the last days of the experiment whenthey decreased appre- ciably in size. Though observations ceased at the end of 31 days, regeneration of polyps would most likely have continued further. EFFECTS OF OTHER INJURIES TO STEMS If a stem or branch is cut at any level and the cut end is exposed to sea-water, a hydranth is normally produced. The question arose whether any severe injury except cutting a stem completely across would result in the formation of a polyp at the point of injury, and whether the regeneration of such adventitious polyps would effect regeneration at the neighboring cut ends. Experiment 26. All theinternodes of several stems were either. bored through with a needle, or severely lacerated or slit with a fine scissors. No regeneration at the injured internodes occurred except in two instances. The wound appeared to close up imme- diately, only a crack in the perisarc marked the place of injury. In the two instances, above noted, the hydranths grew out at right angles to the stem, in marked contrast to the hydranths on the rest of the stem, all of which pointed orally. The large num- ber of injuries on each stem did not reduce the number of lateral, oral or basal hydranths regenerated. Nor did the bending of large stems permanently into an acute angle, effect regeneration at any of the cut ends. When, however, one or two internodes on each stem were slit and the stems then bent so that the wounds were kept exposed to the sea-water, a large number of hydranths appeared from the bent ends. It made no difference whether the bend formed an acute or right angle. Nor was regeneration at any of the other cut ends, including oral and basal, effected by the formation of these adventitious hydranths. The ccenosarc from each of the two injured ends of a slit would grow out directly in line with the axis of the stem, and then fuse into a single branch which would regenerate a hydranth at the 348 A. F. Goldfarb distal end. Sometimes the ccenosare from the two ends would fuse close to the wound, or each wounded end may independently regenerate a polyp, or less frequently the one or the other end only, develops a polyp. Whether the injury was at the distal, middle or basal part of the stem did not influence the regeneration. TABEE 21 No. of stems No. polyps reg. at experimented upon Nature of the Operation injured ends in 6 days 6 Stems punctured or slit at every internode I 8 Stems lacerated at every internode 2 6 Stems bent permanently fo) 4 Stems bent and punctured at the bend ° 12 Stems slit and bent at point of injury II If a sufficiently large area of cenosarc is cut, irrespective of the level of the stem, and the wounds are prevented from closing imme- diately, a relatively large number of hydranths are regenerated. These adventitious polyps appear at the same time as the polyps on the lateral branches, and seemed in no way to effect the regenera- tion of the latter. Experiment 27. After stems had been experimented upon for a long time and could no longer be made to regenerate, they were cut into small pieces. Sometimes the lateral branches were also cut close to the stem. The pieces cut from the distal regions never regenerated, for the very obvious reason that coenosare is never present in the distal parts of stems. But the middle and basal pieces regenerated an incredibly large number of polyps at their cut oral and basal ends. Some of the lateral branches which had been cut close to the main stem also regenerated polyps. Here again injury to the coenosarc accompanied by exposure to sea- water rejuvenated the pieces in so far as rapid and extensive regen- eration of polyps is concerned. EFFECTS OF DILUTED AND CONCENTRATED SEA-WATER”™ Experiment 28. Sea-water was diluted by the addition of tap- water so as to make a graded series, with differences of 5 and some- times IO per cent, from normal sea-water to 50 per cent dilution. *3'These experiments are based on Loeb’s 91. See also Snyder ’05, Archiv.f. Ent. Large numbers of stems and branches were used. Factors in Regeneration 349 In the experiments of 1905 the total number of polyps was daily recorded. ‘These records agree with those of Loeb, that the num- ber regenerated increases with the increased dilution of sea-water. The maximum regeneration is reached in sea-water diluted 15 to 20 per cent; beyond this point, that is, in solutions more diluted, regeneration rapidly decreases; in 40 per cent, few polyps are produced, in 50 per cent, none. In 1906 the experiment was repeated, but the records indicate the number of different hydranths daily regenerated. The results are practically in accord with the data obtained by the other method in 1905. ‘The rate of development in both series was the same in normal sea-water and in solutions diluted as much as 15 per cent but beyond this point the greater the dilution the greater the retardation. TABLE 22 No. of different polyps regenerated in No. of branches Solution 3 days 6 days used Normal 46% 82% 234 5% 56% 79% 15% 62% 84% 25% 28% 73% 35% o% 34% Stems of Pennaria tiarella behave in quite the same manner as Eudendrium in dilute solutions of sea-water. The results in both hydroids are practically the same. Experiment 29. An effort was made to acclimatize the stems of Eudendrium and Pennaria to greatly diluted sea-water, and thereby to have them regenerate in solutions diluted 50 per cent or more. Sea-water was daily diluted 24 per cent more than the preceding day. Polyps appeared in all dilutions until 45 per cent was reached, beyond which regeneration ceased on Euden- drium stems, while Pennaria ceased at 50 per cent. In the very diluted solutions polyps were distinctly smaller than the normal polyps. ‘The experiment seemed to show that hydranths of Euden- drium and Pennaria could not be made to regenerate in solutions diluted more than 45 and 50 per cent, respectively, during the 25 days of the experiment. 350 A. Ff. Goldfarb Experiment 30. ‘The stems from the preceding experiments were removed from their solutions to normal sea-water. Four to six days after the transfer, hydranths appeared, first, on stems from the least diluted sea-water, later on stems taken from 20 to 30 percent dilutions while no regeneration occurred on stems kept in water diluted 40 per cent or more. ‘The number regen- erated increased daily but not to the same degree, so that by the eleventh day after the transfer the stems taken from the greatly diluted water regenerated as much as those taken from the normal or slightly diluted sea-water. TABLE 23 Stems in dilute solutions for 11 days were transferred to normal sea-water Solution prior to Per cent regenerated in transfer 3 days 6 days g days 11 days Norm g* 20 32 40 5% ° 9 21 28 10% ° 22 36 47 15% 16* 22 32 38 20% fo) 3 15 3° 25% fo) Io 14 27 30% ° 5 22 33 35% ° 6 26 40 40% ° ° ° ° * These polyps were present at time of transf r. The effects of concentrated sea-water will be taken up more fully in the following experiments. Experiment 31. A graded series was made by boiling and there- by concentrating sea-water. For example, 200 cc. of sea-water at 18° C. when boiled to a volume which at the original tempera- ture was 180 cc. constituted a go per cent concentrated solution. The water was filtered and aérated by thorough shaking. Every few days the water was replaced by water freshly concentrated to about the same per cent. TABLE 24a ; Polyps regenerated on the following days: No. of branches Concentration 2d 3d 4th 5th 6th 7th 650 Norm 11% 50% 55% 52% 32% 5% 92% to 95% ° 5 15 15 15 4 85% to 89% ° 2 6 12 12 10 777% fe) ° fe) I ro) ° 65% ° ° ° ° ° 62% ° ° ° ° ° Factors in Regeneration 351 TABLE 24b Polyps reg. in Concentration 3 days 6 days No. of branches Norm 51% 91% 719 90% 14 51 87% 7 32 85% 5 37 80% 76% fo) fo) 70% etc The more concentrated the sea-water the fewer the polyps pro- duced. Slight increase of concentration results in large decrease in the number regenerated. “The maximum concentration beyond which no regeneration occurs is about 77 per cent. Retardation observed in diluted sea-water likewise takes place in concen- trated solutions. It is almost needless to add that the greater the concentration the slower as well as the fewer the polyps produced. Experiment 32. “Vhe maximum concentration at which regen- eration of Eudendrium stems will take place was more definitely ascertained by transferring them daily to solutions more concen- trated by about 2$ per cent. Regeneration occurred in all con- centrations from norm to 573 per cent. Even in 50 per cent solutions, shoots were observed. ‘Thus by slowly increasing the concentration of sea-water, stems were made to regenerate polyps in a considerably greater concentration than had otherwise taken place, when stems were placed at once into the 574 per cent concentrated solution. . Experiment 33. Stemsthat had been kept in concentrated solu- tions were after 18 days transferred to normal sea-water. TABLE 25 Stems kept in concentrated solution 18 days were transferred to normal sea-water. On the fol- lowing days there regenerated: Previous Concentration 2d 3d 4th sth 6th 7th 8th gth roth Norm o% o% 2% 2% 47% 4% 6% 6% 10% 95% 2 2 2 2 4 2 4 II 28 89% ° $ 3 3 7 12 16 26 35 79% ° ° ° ° ° ° ° ° 73% ° Co) fr) ° fo) ° ° 352 A. F. Goldfarb The greater the difference in salinity between the solutions in which stems had been kept and normal sea-water, the greater the regeneration after the transfer. But this is exactly what might have been expected from Experiment 28. Seventy-nine per cent concentration or thereabouts, marks the toxic point beyond which stems do not regenerate when transferred to normal sea-water. The percentage of salts present in sea-water determines whether regeneration shall or shall not take place. An excess, or on the contrary, too little salts present in the solution prevents regen- eration. Whether the effects produced are the result of differences — in osmotic pressure or of the specific action of the salts or of both of these factors was not determined. Regeneration took place on the one hand tn solutions diluted to 45 per cent and on the other concentrated to 58 per cent. From these two extremes the number regenerated increases to a maximum not in normal sea-water but in 15 to 20 per cent diluted sea-water. CONCLUSIONS AND SUMMARY It is extremely difficult, even approximately, to distinguish the external from the internal factors in regeneration. Both. kinds of factors play important roles in the life history of Kudendrium and other hydroids. For convenience and for purposes of study each of the factors in these two series were separately considered, though it should be remembered at all times that this is an arti- ficial though convenient arrangement, and that these factors never act singly and independently of the rest. ‘These influences bring about various reactions, only some of which may be said to be adaptive. The following five factors may be said to result in adaptive changes in Eudendrium.” 1 Gravity determines the position, at which regeneration shall more frequently take place, and the direction of growth. It does not determine the kind of regeneration, for with rare exceptions only polyps are produced when regeneration occurs at all. On erect stems oral and lateral cut ends regenerate pro- fusely. On inverted stems,-regeneration is greatly stimulated, *4Loeb laid emphasis on but two factors, namely, light and gravity. Factors in Regeneration 353 at the basal end only, while polyp formation at the lateral and oral ends is largely inhibited. New stems and branches show a strongly negative geotropism, and grow upward irrespective of the position of the piece. 2 Sunlight. Stems or branches exposed to the direct rays of the sun regenerated a greater number of polyps than those kept in the shade of the room, the temperature in both cases being approximately the same. How much the increase was directly due to the effect of the actinic rays per se, or indirectly to the destruction of bacteria, or to the slightly increased temperature, or to all of these factors was notascertained. Manyof the stems and branches bend toward the sun, 7. ¢., they are positively heliotropic. 3 Temperature. Other conditions being favorable and equal, regeneration increases with increased temperature to the optimum, and decreases with the lowering of the temperature. At 10° C. regeneration is largely inhibited, while regeneration increases up to and including 28° C. temperature. One of the more important conditions, just mentioned, is exposure to sunlight for stems placed in water at a moderately high temperature and not exposed to direct sunlight produces far fewer polyps. 4. Any severe injury at any level of the colony, may cause polyps to regenerate, 1f the wound be exposed to sea-water. ‘Uhe direction of growth of the pedicels and the rate of development of the polyps are subject to the external conditions mentioned and to the internal conditions to follow. 5 Contact, pressure, impact, etc., are inhibiting influences which tend to prevent complete development at those ends that come in contact or are pressed upon by a solid body. Shoots are often produced, but further differentiation is stopped. Con- tact determines, in some degree, particularly on very young branches, the direction of growth, which is away from any solid body, therefore, is negatively stereotropic. “The amount of inhib- ition is proportionate to the degree of contact, pressure, impact, etc. 6 Large variations in the concentration of sea-water probably never occurs in nature and the reactions of Eudendrium to differ- ently concentrated solutions can hardly be called adaptive. The maximum number of polyps regenerated does not occur in normal 354 A. “f. Goldfarb sea-water but in solutions diluted with about 20 per cent of tap- water. [he amount of salts present in this solution is most favor- able to regeneration. As the quantity of salts is increased by concentration or decreased by further dilution, the number regen- erated decreases until the mimimum is reached on the one hand at 45 per cent dilution and on the other at 58 per cent concentrated sea-water. Stems transferred from concentrated to norm sea- water which is equivalent to placing them in dilute solution, regenerate according to the principle laid down, viz: the more dilute the solution, to a certain point, the more hydranths pro- duced. On the contrary, stems transferred from dilute to norm sea-water, which is practically placing them into more concentrated sea-water, do not regenerate less than stems continuously kept in normal sea-water. Regeneration is not inhibited until the solu- tion contains more salts than that normally present in sea-water, while the stimulating effects of diluted sea-water occurs, when either concentrated, normal or dilute solutions are diluted to what is equivalent to a 20 per cent dilute sea-water. Before summarizing the internal factors, the behavior of the ceenosare under different conditions might perhaps more profit- ably be taken up. The ccenosarc is circular in cross section, with no partition as in the case of Tubularia. Granules within the coenosarc stream alternately toward the apical and basal ends, either in the same direction throughout the colony or independently, in each branch, or even in different directions in the main stem. “The coenosarc itself can move en masse within the perisarc. With the basal two- thirds of stems excepted, the cenosarc invariably moves toward the basal end of the piece, 1. e., in all branches, in the apical pieces of stems, and in small pieces from any region of the colony. It may even move entirely out of the piece, through the basal end. Now, as regeneration occurs only where the cenosarc is present, it follows that whether regeneration shall or shall not take place at a cut end is determined by the migration of the cenosarc. This movement basally crowds or concentrates the cenosarc at the basal end of the piece and 1f conditions are favorable at the end polyps readily appear there. ‘The migration of the ccoenosare may be furthered in various ways, namely: Factors in Regeneration 355 1 Inverting pieces is almost certain to stimulate regeneration at basal ends; or, better still, embed inverted pieces in sand and a remarkable number of basal polyps appear on the free parts. 2 ‘Tying a ligature at the distal end of a stem or, still better, ligature the lateral branches, then ligate the middle of the stem, and a greatly increased number of heteromorphic polyps result. 3 By cutting small pieces from any part of the colony a far greater number of polyps is produced at the basal than at the apical ends. Thus though the cwnosarc 1s influenced by such external factors as gravity, ligatures, lack of oxygen, cold, and by internal factors, such as age, size of the piece, etc., the cwnosarc normally behaves in certain definite ways which, without the aid of hypothetical “specific stuffs” not only accounts for the absence of polyps at cer- tain cut ends but accounts for their regeneration atotherends. Under a given set of conditions we can foretell with a fair degree of accu- racy the number and region at which regeneration will take place. Furthermore, it is not necessary to have recourse to the stimu- lating effects of necrotic tissues thrown into the circulation to account for the regeneration of polyps. In Tubularia it has been maintained that the breaking down of the partition near the cut end throws into the circulation material which stimulates regen- eration at thatend. Inahydroid resembling Eudendrium, namely, Pennaria, Gast and Godlewski believed that the disintegration of polyps supplied the circulation with material which stimulates regeneration. In Eudendrium there is neither a partition as in Tubularia, nor were the polyps permitted to disintegrate, for they were cut off at the beginning of the experiment. Yet hydranths were formed within 48 hours often at every cut end. Even when hydranths were daily removed as soon as formed, other polyps were regenerated. . The following internal factors affect regeneration: 1 Age determines not the kind but the rate and number regen- erated. ‘The younger the region the more numerous and the quicker do polyps appear. alii statement is subject to special conditions already enumerated, such as ligatures, inversion of stems, migration of the coenosarec, etc. 356 : A. “f. Goldfarb 2 The influence of the presence or absence of lateral branches and their pedicels. Pieces with the lateral branches cut off close to the main stem regenerate many more polyps than similar stems from which only the polyps have been removed, for the reason that the coenosare tends to withdraw from the pedicels whereas it does not do so from the lateral ends cut close to the stem. 3 This suggests another closely related factor, viz: the influence of size (t.e., the amount of cenosarc) on the number and position of the regenerated polyps. Pieces less than 1 mm. long may regen- erate stems but never complete polyps. Pieces 1 to 1 mm. long regenerate but one polyp; more frequently however at the basal end. Sometimes one is produced at each end. On larger pieces two or three polyps may appear on an outgrowth at the basalend. Still larger pieces may bear as many as nine basal polyps at one time, while itis rare for more than one apical poylp to be produced. 4. The influence of the old tissue on the kind of regeneration. Polyps are replaced only by polyps; stems if injured give rise to polyps. Under certain unfavorable conditions proliferation of cells may take place but no differentiation into polyps occurs, and “stolons,’’ or modified stems, result. If these are cut and removed or grow into a favorable environment polyps, not “stolons,”’ are regenerated. The rate of regeneration varies with the size and age of the piece. Large stems produce polyps quickest at the distal, slowest at the basal region. Medium size pieces regenerate at all the cut ends at the same time. Polyp formation is greatly retarded on small pieces even if the pieces are taken from the distal region of large stems. ‘The two cut ends at any level.of a stem regenerate at the same time. ‘The presence of polyps does not prevent, but may retard, regeneration at the basal end. Unfavorable con- ditions, such as lack of oxygen, low temperature, greatly diluted and even slightly concentrated sea-water, gravity (on lateral ends of reversed stems) all retard development. Nothing availed to affect regeneration in less than two days. Zodlogical Laboratory Columbia University, New York New York City, March 1, 1907 STUDIES ON REGULATION XI FUNCTIONAL REGULATION IN THE INTESTINE OF CESTOPLANA BY (GG Wile, (leQUED With Twenty Text Ficures This Neapolitan form which has served for other experiments (Child ’o5a, ’o5b, ’o5c) is very favorable for the study of the intes- tinal changes which occur during form-regulation. ‘The intestine in normal animals is almost black in color and since other portions of the body are unpigmented is very distinctly visible in the living animal. Moreover, the regulatory changes are extreme and in some cases relatively rapid; and finally animals and pieces live for months in clear water without food so that it is possible to follow the intestinal changes during a long period. I THE TURBELLARIAN INTESTINE, ITS FUNCTIONS, AND FUNC- TIONAL FACTORS INVOLVED IN ITS DEVELOPMENT AND) REG- ULATION This part of the paper aims to establish a general basis for interpretation of the experiments and observations to be described later. It precedes rather than follows the descriptive part because it Is important, as well as economical of time and space, to be able to point out the bearing of the various experimental data, under each head instead of postponing interpretation to a general sec- tion where the chief points of the description must be reviewed. The basis of interpretation suggested here is, however, in part the result of these and other similar experiments, not a ‘precon- ceived hypothesis with which the facts are to be brought into accord. As will appear also, it is in line with previous suggestions which I have made concerning the dynamic or functional char- acter of form-regulation (Child ’o5a, ’o6a, ’o6b). Tue JourNAL oF EXPERIMENTAL ZoOGLOGY, VOL. IV, No. a 358 C. M. Child rt The Turbellarian Intestine and its Functions The names by which the various organs of the lower inverte- brates are designated do not necessarily serve to indicate with any degree of exactness their functions. We commonly speak of the alimentary apparatus of such forms as the turbellaria as an intestine, a digestive system, etc., but strictly speaking the func- tions of this apparatus are not identical in all respects with those of the intestine of the vertebrates for example. It is of course a digestive system, but it is more than that. In the first place, the turbellarian intestine undoubtedly serves as a place of storage for undigested nutritive material. Any one who has observed turbellaria feeding can scarcely fail to recognize that this is an important function of the intestine, at least in cer- tain species. Food is often taken until not only the intestine but the whole body is greatly distended. In fact I have often observed the bursting of various species in consequence of rapid intake of food. The opening in such cases 1s usually small and after out- flow of the excess of material soon closes. Under such conditions the intestinal walls must of course undergo great mechanical extension. Secondly, digestion is, at least in part, intracellular and the intestinal cells undoubtedly accumulate reserve material when food is abundant; in other words, when digestion proceeds more rapidly than material is removed. But besides the functions of digestion and accumulation of reserves the intestine in these forms is the chief means of distri- bution of the nutritive material to various parts of the body, ». e., it is in greater or less degree a circulatory system, a fact which has been recognized by those authors who have termed it the gastro- vascular system. As a gastro-vascular system it contains fluid laden with nutritive substances. ‘This fluid moves to and fro, enters and leaves the various branches and regions according to the muscular contractions of the body-wall. ‘Thus the intestinal wall is subjected to the varying fluid pressures which, however, are more or less typical for each particular region since the muscular contractions are in general typical. A wide range of conditions Studies on Regulation 359 exists, of course, for each region, but the conditions in the terminal regions, for example, must be in general typically different from those in the middle region. To sum up: the turbellarian intestine as an organ of digestion and a store-house of reserve material is undoubtedly the seat of typical chemical reaction-complexes. As a reservoir for the tem- porary storage of undigested food and as a vascular system con- taining moving fluid it is undoubtedly subjected to a typical com- plex of mechanical conditions. ‘These two groups comprise, I believe, the most important functional conditions for the turbel- larian intestine. The intestine of higher forms, or at least some part of it, serves as a place of temporary storage for undigested food and often, as in certain birds and mammals, undergoes a high degree of special- ization in connection with this function. But in higher forms where a specialized circulatory system is present, the intestine does not function to any great extent as a system for the distri- bution of nutritive material and is not subjected to the mechanical conditions which must exist in such a system, although of course mechanical functional conditions are more or less important fac- tors in the functional complex in all cases. There can be no doubt, however, that mechanical conditions constitute a much larger element in the functional complex characteristic of the turbellarian intestine than they do in higher forms. If functional factors play any part in development and regulation, we may expect to find the determining factors in the two cases different to a greater or less extent. 2 Functional Factors in Intestinal Development and Regulation It is a well-established fact that the mechanical conditions con- nected with the movements and pressure of fluid within the vessels are factors of great importance in determining diameter, distri- bution, angle of branching and character of the wall of the blood- vessels. Since this is the case it is natural to expect that similar conditions will play a role of greater or less importance in develop- ment and regulation of the turbellarian intestine. 360 C. M. Child Judging from the form of the turbellarian intestine in relation to the form and structure of other parts it is difficult not to believe that functional and particularly mechanical conditions are impor- tant factors in its development. In the rhabdoccels where no strands of parenchyma or dorso-ventral muscles oppose it, it forms simply a sac, filling the pseudoceel in part or wholly accord- ing to conditions. In the polyclads, on the other hand, the intes- tine might be compared roughly to an elastic sac placed in a space in the axis of the body and then gradually distended so that parts of it are forced into the parenchymal spaces toward the periphery of the body. If the fluid contents of the intestine move and exert pressure in typical directions it seems to me that the effect of these movements must necessarily appear in the direction and size of the intestinal branches. All the facts seem to indicate that the general direction and arrangement of the intestinal branches in the various parts of the body 1s determined, at least in large part, by the mechanical conditions resulting from movements and pres- sures of the fluid contents. By altering these conditions the arrange- ment of the intestinal branches can be altered, as I showed for Leptoplana (Child ’o4a). In the triclads conditions are similar but the intestine develops in different form because of the position and form of the pharynx. ‘The almost infinite variations in type of the “normal” turbellarian intestine in a given species simply show, in my opinion, how largely its form as regards details is a matter of chance, determined often by the presence, absence, or position of spaces, or dorso-ventral muscular fibers in the paren- chyma, by slight individual differences in movement or consti- tution of other parts, etc. If the functional conditions connected with the movements and pressures of fluid contents are essential factors in determining the form of the intestine, we may expect to find changes of form occur- ring when these factors change and the facts justify our expecta- tions. Starvation of a planarian results in degeneration and total disappearance of the most distal portions of the intestine in succes- sion: feeding results in the redevelopment of branches, but not necessarily in the same pattern, and increased distension of a normal animal results in the formation of new intestinal branches. 361 x Studies On Regulation But it is in connection with experiments on form-regulation that the extreme plasticity of the turbellarian intestine becomes evident. The changes in form and arrangement of the intestinal branches in the experiments of Lillie (o1) and Bardeen (’or, ’02, ’03) are sufficient to illustrate this point, although they do not demonstrate its correlation with the functional conditions result- ing from the movements and pressures of fluid contents. In most triclads and polyclads intestinal regeneration 1s usually much less complete than the regeneration of other parts when the animals are not fed. Moreover, and this seems to me to be a crucial point, it is much less complete in pieces without the ceph- alic ganglia than in pieces containing the ganglia (Child ’o4a). It can scarcely be supposed that there is any essential difference in nutritive conditions between pieces with and those without the ganglia. If anything, more nutritive material should be avail- able for growth in the piece without ganglia since it is much less active than the other. I do not believe, however, that such differ- ences in intestinal regulation can be due primarily to the differ- ences in nutritive conditions. ‘The only reasonable basis for interpretation seems to me to lie in the differences in activity. In the piece without ganglia the movements of the intestinal con- tents are less frequent and less energetic, and consequently the stimulus to intestinal growth in the new tissue is less than in the piece containing the ganglia. Observation of two such pieces and of the movements of intestinal contents in their bodies shows very clearly that the intestinal pressures and tensions are much greater in the piece containing the ganglia than in that without them. All the data thus far available seem to me to indicate that the form and arrangement of parts of the turbellarian intestine is determined very largely by mechanical factors due to the presence and movements within it of fluid contents. This statement is not to be interpreted, however, as signifying that nutritive factors play no part in determining intestinal form. The form must be altered to a certain extent by the presence or absence of reserve material in the cells, by the general metabolic conditions, the relation between intake and output, ete. But I find it difficult to under- stand how such factors as these can possibly determine the general 362 C. M. Child outline of the intestine, and the direction and arrangement of its branches. Lack of nutrition may of course determine the degen- eration of a branch or of branches, but how can the presence of nutrition determine the position and direction of new branches? On the other hand, the conditions above mentioned do account readily for position, outline, and arrangement of parts and experi- mental data indicate that they are the factors chiefly involved. The development of intestinal branches is simply another illus- tration of the fact which I have mentioned elsewhere at various times, viz: that the stimulus to growth is not identical with the presence of nutritive material, but that, on the other hand, nutri- tive material goes where the demand is greatest even at the expense of reduction and disappearance of other parts where the demand is less. This relation between growth and nutrition seems also to show why such extensive intestinal reduction occurs in many turbellaria during starvation: the demand for nutritive material is greater 1n other parts than in the intestine, consequently material passes from it to them. In short, I believe the whole problem of the “self-regulation of metabolism” during starvation and indeed at other times is essentially a problem of relative func- tional activity in the broadest sense. In Cestoplana the axial intestine extends directly through the median region of the body from end to end. The lateral branches are at right angles to the axial intestine in the pharyngeal region, but toward the anterior end gradually change their direction, and are directed more and more anteriorly: posterior to the pharynx exactly the reverse is the case (Fig. 1). The movements of intestinal contents in this species are briefly as follows: general contraction of the body forces the intestinal contents from both ends toward the pharyngeal region, and the axial intestine and the lateral branches of the middle region of the body become distended. General extension of the body forces the intestinal contents out of the middle region to a large extent and distributes them along the lateral branches even to the extreme terminal regions, if the contraction is strong. Under these conditions the intestinal contents move anteriorly in the prepharyngeal and posteriorly in the postpharyngeal region. Studies on Regulation 363 Local contractions and extensions of course cause local changes in the distribution of intestinal contents, but these follow the same rules as the more general movements. Evidently then, the intestinal branches in the middle region are filled and distended by the intestinal contents which accumulate in the middle region during contraction and the branches in the terminal regions by the contents during their flow away from the middle region. In the regions between the middle and end all intermediate conditions exist. “Those branches which are filled and distended chiefly by the fluid accumulating in the pharyngeal region arise at right angles to the axial intestine since their for- mation is correlated essentially to lateral pressure of the intes- tinal contents, escape in other directions being impossible. But toward the ends of the body the intestinal branches are filled and distended by fluid which is moving anteriorly or posteriorly. If mechanical conditions are factors in determining the form of the intestine, the intestinal branches in these regions may be expected in accordance with the laws of hydrodynamics to be directed more or less obliquely in the direction in which the fluid is moving. The intestine in Cestoplana seems to me to possess exactly the form which might be expected if movements and pressures of fluid contents are the chief factors in producing it. The fact that a gradual change in direction of the branches between the middle and the ends of the body exists is due simply to the gradual change in conditions. In the regions between the middle and terminal regions the branches are filled and distended in part by the fluid moving away from the pharynx, and in part by standing contents escaping laterally from pressure in other directions. The nearer the pharyngeal region, the more exclusive the latter condition of hlling and distension, the farther away the more exclusive the former. Hence we may expect to find with increasing distance from the pharynx a gradual change in the direction of the branches from a position at right angles to the axis to one oblique toward the direction of movement of the contents. Similar conditions in general, with of course various specific differences, exist in other polyclads and triclads, and the form of the intestine as a whole and of each of the long branches in many 304 GC. M. Ghild of the broader polyclads corresponds very closely to what may be expected if hydrodynamic factors play an important part in their formation. On the following pages the various regulatory changes in the intestine of Cestoplana under various conditions are described and their bearing on the above dynamic hypothesis of intestinal development is discussed. Perhaps it should be added in order to forestall objections that hydrodynamic factors are not considered as the only factors involved in determining intestinal outline and arrangement of parts in the turbellaria. It seems very probable that other factors must also play some part, though the facts seem to me to indicate that hydrodynamic factors are certainly of great importance. II THE NORMAL INTESTINE AND THE TYPICAL COURSE OF INTES- TINAL DEGENERATION IN THE ABSENCE OF FOOD t Descriptive The appearance of the intestine in newly captured animals differs to some extent, apparently according to the previously existing conditions. In Fig. 1 the terminal and middle regions of the intestine in a normal newly captured specimen are shown, somewhat diagrammatically. The intestine in this case is only moderately distended by its contents: in many cases it is so dis- tended that no spaces between the branches are visible and it appears as in Fig. 2. In uninjured animals kept without food a gradual reduction or degeneration of the intestinal branches occurs, though much more slowly than under certain experimental conditions. Intestinal reduction proceeds from the peripheral or terminal region of the intestine toward the middle. The first parts to disappear are the tips of the branches at the anterior and posterior end and as reduction of these branches continues branches nearer the middle region are affected until a condition resembling that shown in Fig. 6 is attained. In this case which represents a normal animal after about four and a half months without food, only short stumps of the lateral branches Studies on Regulation remain in the terminal regions of the body. With approach toward the middle the length of the branches increases until in the pharyngeal region they still retain their full length, though they are less distended than originally. At the beginning of the experiment the intestine of the specimen figured presented the condition indi- cated in Fig. 2. Undoubtedly intestinal reduction could be car- ried further in normal animals, but departure from Naples made it impossible to keep the speci- mens under observation longer. Intestinal reduction in this species consists in an atrophy and disintegration of the more distal portions of the intestine, not merely in a reduc- tion in size or contraction. Various stages can be more or less clearly distinguished, in most cases, though of course each gradually passes into the following. Starting with the normal well-filled intestine as in Fig. 2, or Fig. 1, the first changes consist in de- creasing distension, so that the individual branches become more clearly distinguishable. » 'Somewhat later the distal portions of these branches disinte- grate and form a longitudinal band of dark granu- lar substance, which appears somewhat like a longi- tudinal canal on each side connecting with the lateral intestinal branches (Fig. 3, also the pharyn- geal region in Fig. 6). Under high magnification, however, these longitudinal bands are clearly seen to be the débris of the disintegrated terminal regions of the branches. ‘The lateral bands make their appearance first in the more terminal regions of the body and progress toward the middle regions. as the ends of the branches undergo degeneration. But a part of the products of degeneration 366 C. M. Child appears within the intestine. As degeneration of the branches proceeds a fluid crowded with dark granular masses appears in the intestine and may accumulate and distend the remaining parts of the intestine in pieces of certain sorts to be described in another section. In normal animals, however, this substance never accumulates to any great extent but undergoes resorption almost as rapidly as it is formed and undoubtedly serves as nutri- tive material for other organs which are still functional. As reduction continues the lateral bands become less conspicu- ous, the dark color gradually fading out as they undergo resorp- tion, and the lateral branches undergo further reduction until their tips no longer extend to the region occupied by the lateral bands. At this stage the intestine appears as in Fig. 4 or as in the regions a short distance anterior and posterior to the pharynx in Fig. 6. Somewhat later still the lateral bands disappear entirely 4 i : : = 5. 2 a i 3 t + 1 Fics. 2, 3, 4 AND § or break up into parts which sooner or later disappear. Various stages in the disappearance of the lateral bands are shown in Fig. 6. And finally, intestinal reduction may proceed so far that only the axial intestine remains (Fig. 5). In some cases, as in this figure, the intestine still shows slight indications of the positions of the former branches, but often even these disappear and abso- lutely no trace or indication of branches can be discovered (Fig. 19). Often, as in Fig. 5, the lateral branches disappear before the last traces of the lateral bands, which may persist for a time as isolated groups of granules, presumably occupying the paren- chymal spaces originally filled by the ends of the lateral intestinal branches. The next stage is of course complete disappearance of the intes- tine from the regions concerned. ‘This stage is attained only in the terminal regions of the normal body and of pieces under cer- tain conditions. In Fig. 6 the intestine has disappeared almost Studies on Regulation 307 entirely from the preganglionic region, in which it is present in normal well-fed animals (Fig. 1). In all observed cases of intestinal degeneration, except under certain conditions connected with form-regulation, the course of the process of degeneration is essentially the same and _ passes through the stages described above. 2 Dtscussion According to the above account the intestinal degeneration begins at the extreme peripheral regions of the intestine and pro- ceeds “‘centripetally.”” The ends of the branches in the terminal regions are the first parts to disappear, and the last branches to undergo reduction are those immediately about the pharyngeal region. It can scarcely be supposed that the more peripheral branches or the more peripheral regions of each branch are less needed than the more central parts and so disappear first. The peripheral portions of the intestine would seem to be just as essential as other parts for proper nutrition. ‘The head-region is the most active region of the body and yet the anterior end of the intestine dis- appears earlier than any other part of the prepharyngeal intestine. But when the course of reduction 1s considered from a functional standpoint interpretation becomes easy. In the first place the quantity of intestinal contents undergoes gradual decrease from the beginning to the end of the experiment. In well-fed animals the intestine is greatly distended (cf. Fig. 2) with food at first. This nutritive material is gradually used up, but as degeneration of the intestinal branches occurs a part of the products of degener- ation appearsin the intestine as a fluid crowded with dark granular masses. In normal animals this too undergoes resorption almost or quite as rapidly as it is formed, and gradually decreases in amount as time goes on. ‘Thus even long after the food taken from without has disappeared the intestine is not empty, but the amount of intestinal contents is always decreasing. The move- ments of this dark substance in the intestine can be readily ob- served and the following statements regarding their relation to the general muscular contractions are the result of direct observation. 368 C. M. Child Under extreme conditions of intestinal distension with material all parts may be subjected to equal or nearly equal internal pres- sure but when decrease in the amount of intestinal contents occurs, ide t eles peg tatemreete aah: abe ariee ave re 5 ? : ri ' - + as a b 3 £ Fic. 6 as is the case when the animals are kept without food, the energy of the mechanical conditions connected with the contents must decrease more rapidly in the periph- eral than in the middle regions. ‘Thus, for example, if the intestine 1s only partly filled, the internal pressure on the walls in the extreme anterior and posterior regions is in general much less than in regions nearer the middle. In the first place, the fluid contents are forced into this region only during extreme extension and then appar- ently with much less energy than into regions nearer the middle. The consequence is that those regions in which the functional stimulus falls below a certain minimum gradually undergo atrophy and degeneration, and as the intestinal contents continue to decrease -in amount this atrophy and degeneration gradually extend toward the middle region, which is the last to be affected. Size of the lumen of the various parts and friction between the contents and the walls must also play a part in determining movements and internal pressures of the intestinal contents and both of these factors tend to reduce the energy of the functional conditions more rapidly in the peripheral than in the middle regions. The lateral intestinal branches in and about the pharyngeal region persist longer than any others, simply because the functional conditions are less altered there than elsewhere. In the first place, contraction which drives the intestinal contents toward the middle is usu- ally sudden and violent, in consequence of sudden ex- ternal stimuli, while extension is usually much slower and less extreme. Consequently the intestinal contents are driven into the lateral branches of the middle regions with great force long after they have ceased to reach the extreme peripheral regions at all. “The normal movements of the animal, especially Studies on Regulation 369 the very frequent slight contractions of the anterior and posterior end all tend to keep the middle regions of the intestine more dis- tended than the peripheral regions. Very probably the other functions, 7. ¢., the digestive and storage functions, also play a part in determining atrophy. Of course absence of intestinal contents from any part of the intestine means absence of food to be digested and stored up. Hence the cells of this region may atrophy or change their character because of the partial or total absence of the stimulus to the digestive function or because of malnutrition: or again degeneration may occur because the demands upon these cells for nutritive material are so great in relation to the supply, that they are exhausted or forced so far from equilibrium that continued existence is impossible: degeneration from either of these causes would affect the peripheral regions first and proceed toward the middle. But as will appear below, in certain experimental cases it 1s impossible to account for the regulatory intestinal changes on any other basis than that of mechanical stimuli from the contents. III INTESTINAL REGULATION IN CORRELATION WITH FORM- REGULATION OF PIECES The character of form-regulation in general in this species was described in an earlier paper (Child ’o5a). It will be recalled that regulation after removal of posterior pieces consists almost entirely in redifferentiation of the parts remaining, only a very small amount of new tissue being formed on the cut surface. Posterior regulation is qualitatively, 7.e., functionally, complete at all levels except anterior to, in, and immediately posterior to the cephalic ganglia. Regulation in the anterior direction, on the other hand, consists almost wholly of regeneration, except as regards certain cases of pharynx-formation, and 1s complete only at levels anterior to, in, and immediately posterior to the ganglia, being slight in amount elsewhere. As might be expected from these differences, intestinal regula- tion is much more extensive in correlation with posterior than with anterior regulation. But the most remarkable cases of intestinal 370 | COM Gen regulation occur in cases where return to the typical form of the species does not occur. In the earlier papers (Child ’o5a, ’o5b, ’o5c) dealing with this species the processes of form-regulation were interpreted as essen- tially cases of functional regulation, 7.e., “functional adaptation.” For example the redifferentiation into a posterior end of the pos- terior part of a prepharyngeal piece and the formation of the pharynx at a certain level of the old tissue was regarded as the result of a functional regulation in response to altered conditions, in consequence of which a portion of the body which had been functionally, as well as morphologically, prepharyngeal now became functionally posterior, 7.e., postpharyngeal, and in conse- quence underwent regulation, 7. ¢., functional adaptation of its structures to the new conditions. As [ have pointed out repeatedly in different papers (Child ’o5a, ’o6a, ’o6b), redifferentiation of old parts into parts similar to those removed can occur only when these old parts are capable in some degree of becoming the functional representatives or substi- tutes of the parts removed. If functional substitution for the part removed does not occur at all, form-regulation does not occur: if the substitution is confined to regions adjoining the cut surface the part is replaced more or less completely by regeneration, the completeness of replacement depending on the degree of functional substitution. As was shown in the earlier papers on Cestoplana (Child ’o5a, ’o5b, o5c), the phenomena of form-regulation in general can be readily and consistently interpreted on this basis and the differ- ences between anterior and posterior, preganglionic and post- ganglionic regulation, and regulation in the presence and in the absence of the ganglia, differences which on any other basis appear merely as isolated facts without special significance and without relation to each other, are clearly correlated and explicable. For the more complete discussion and interpretation of the experimental data in the light of this hypothesis the reader 1s referred to the earlier papers (Child ’o5a, ’o5b, ’05c, 06a, ’o6b). Since the phenomena of intestinal regulation are so striking in this species they were omitted from the preceding papers as desery- Studies on Regulation a7 ing special consideration. As will appear, however, they afford strong support to the hypothesis which has served for interpreta- tion Be the other phenomena: indeed a consistent interpretation seems scarcely possible on any other basis than that of functional regulation. I Intestinal Regulation in Correlation with Posterior Form- Regulation a Inthe Prepharyngeal Region The process of form-regulation in prepharyngeal pieces contain- ing the cephalic ganglia consists essentially (Child ’o5a) in the redifferentiation of the posterior part into a new postpharyngeal region and the formation of a new pharynx between this and the new prepharyngeal region. Regeneration is limited to the extreme posterior end of the piece and amounts to little more than the closure of the wound. The length of the postpharyngeal region thus formed, and consequently the position of the new pharynx, depends on the level of the posterior end of the piece. If the piece includes only the most anterior part of the prepharyngeal region (Figs. 11 and 12), the new postpharyngeal region is short and the pharynx appears near the posterior end. With approach of the level of section to the original pharyngeal region the length of the new postpharyngeal region increases and the pharynx is formed farther from the posterior end (Figs. 7 to 9). In all cases of regulation of prepharyngeal pieces containing the cephalic ganglia the lateral intestinal branches posterior to the new pharynx undergo complete disintegration within a short time after section, leaving only the axial intestine. ‘This is shown in Figs. 7 and 8 for a long piece, and in Figs. 11 and 12 for a short piece. In the first case the piece originally included that part of the body anterior to the line fin Fig. 1 and the new pharynx appeared at a considerable distance from the posterior end of the piece. During the first few days following section the dark color of the ireanaal branches in the posterior part of the piece gradually fades. In six to eight days after section (Fig. 7), 7.e., after the 372 C. M. Child development of the new pharynx is well advanced, the intestinal branches posterior to the new pharynx are seen to be degenerating. The course of degeneration differs somewhat from that described above for normal animals. ‘The branches appear broader and further apart as if this part of the intestine had been stretched longitudinally, and in all probability a mechanical elongation of this part does occur in consequence of its function as a posterior end and region of attachment. A few days later the branches disintegrate completely and the débris, appearing as dark masses ge sec artytromne F- iota aks Heeb goad yr myn Saas een g« a a $ Co : Prom ethan Nn altaya any ey Du Se ie ST ad — \ ie Pe. Co Fics. 7, 8, 9 AND 10 and granules on either side of the slender axial intestine, gradually undergoes resorption until after two weeks or more (Fig. 8) scarcely any traces remain. A slender axial intestine still persists, however. The difference between the postpharyngeal region and the remainder of the body is striking (Fig. 8) for in other regions intestinal reduction has as yet scarcely begun. ‘The sharp limita- tion of this peculiar. process to the postpharyngeal region of the piece makes it certain that the disappearance of the lateral intesti- nal branches is correlated in some manner with the “ redifferentia- tion” of this region from a prepharyngeal to a postpharyngeal region. Studies on Regulation 373 Intestinal reduction in other portions of the body goes on in the same manner as in normal animals, though somewhat more rapidly. But, meanwhile, short and slender new lateral branches develop on the postpharyngeal intestine in many cases. These never attained full development in the specimens observed, but there is no doubt that if the animals had been fed they would have developed and reached normal conditions. Fig. g shows the con- dition seventy days after section of the piece from which Figs. 7 and 8 were drawn. Intestinal reduction in the pharyngeal and prepharyngeal regions has followed the typical course but in the postpharyngeal region new branches have developed. In still later stages without food reduction of all parts of the intestine takes place almost equally until only very short lateral branches remain (Fig. 10, 143 days after section). How much longer such pieces may live it is impossible to say, for my observa- tions extended over only 143 days and many pieces were alive and active at the end of this time. In shorter pieces the process is essentially the same. ‘Taking, for example, a piece including that part of the body anterior to the line d in Fig. 1, the new postpharyngeal region is short, and the pharynx appears nearer the posterior end and the amount of regeneration is somewhat greater than in a long piece like the pre- Paine Fig. 11 shows eis piece fifteen days after section. All traces of the postpharyngeal lateral intestinal branches have dis- appeared, only a very slender axial intestine remaining, which, however, extends a short distance into the regenerated tip. In the short pieces degeneration may begin in the redifferentiating region within four days after section, but in the long pieces does not usually appear for a week or more. As regards later stages a similar difference exists. In these short pieces intestinal reduction in other regions 1s always much more rapid than in longer pieces. In Fig. 11, for example, a stage fifteen days after section, reduction is far advanced in the pharyngeal and prepharyngeal regions, and in Fig. 12, forty-five days after section, scarcely any traces of lateral branches exist in any part of the intestine. [hese short pieces usually die from forty to sixty days after section, 7. ¢., much 374 C. M. Child earlier than the longer pieces. In consequence of the more rapid intestinal reduction and earlier death of these short pieces lateral intestinal branches never develop in the postpharyngeal region. These two pieces represent the two extremes as regards intes- tinal regulation in prepharyngeal pieces. The results in other pieces fall between these two extremes and differ in detail accord- ing to the part of the prepharyngeal region included in the piece. In every case, however, and my observations include some fifty cases, very rapid disintegration of the lateral intestinal branches took place in the region posterior to the new pharynx and in the larger pieces a new system of lateral branches developed later. tee Fics. 11 AND 12 In prepharyngeal pieces from which the head-region and the cephalic ganglia have been removed the formation of a new post- pharyngeal region and pharynx takes place in the same manner as when the ganglia are present (Child ’o5c), the only difference be- ing that the new postpharyngeal regionis longer, the new pharynx farther from the posterior end and the process of degeneration somewhat less rapid than in pieces with posterior ends at the same level but containing the ganglia. As was pointed out in my earlier paper (Child ’o5c), the only ground which suggests itself for this difference is the functional relation between prepharyngeal and postpharyngeal regions. Removal of the ganglia reduces the functional activity of the prepharyngeal region very greatly, but affects the activity of the postpharyngeal region to a less extent, hence in regulation the reaction to the altered conditions at the posterior end involves more of the posterior region of the piece than in cases where the ganglia are present, since the energy of Studies on Regulation 375 reaction is greater in proportion to that of the prepharyngeal reac- tion, when the ganglia are absent, than when they are present. In such prepharyngeal pieces without the ganglia the lateral intestinal branches in the region posterior to the new pharynx dis- appear in exactly the same manner as in the pieces already de- scribed, though apparently somewhat more slowly. The absence of the ganglia, therefore, does not affect intestinal regulation in these pieces, except somewhat as regards rapidity. The products of degeneration of intestinal cells never accumu- late to any great extent in these prepharyngeal pieces. Appar- ently they undergo resorption almost as fast as they are formed, serving, doubtless as nutritive material for the various regulatory processes, and all parts of the intestine become more and more slender and delicate as time goes on. b In the Postpharyngeal Region The character of intestinal regulation after removal of a part of the postpharyngeal region differs according to the relative length of the part removed. If the level of section lies only a short dis- tance posterior to the old pharynx, e. g., at the line g, Fig. 1, the intestinal changes which occur in the region posterior to the old pharynx are essentially identical in character with those described for prepharyngeal pieces and shown in Fig. 7 and 8. ‘This region, originally the anterior end of the postpharyngeal region, rediffer- entiates into a whole postpharyngeal region, and the lateral intes- tinal branches disappear in the same manner as in pieces where the postpharyngeal region is formed from a part of the prepharyngeal region. One important difference exists, however; the degener- ation is always less rapid in these than in prepharyngeal pieces, from three to four weeks being necessary for the disappearance of the branches. Similar changes occur in pieces with posterior ends at levels somewhat posterior to g in Fig. 1, but with increasing length of the old postpharyngeal region in the piece, the degeneration of the lateral branches becomes slower and less complete, until, when half (4, Fig. 1) or more of the old postpharyngeal region remains, the lateral branches do not disappear early as in the 376 C. M. Child pieces described above, but simply undergo reduction as in normal animals. c Discussion In the case of the formation of a new postpharyngeal region from a part of the old prepharyngeal region, or from the most anterior part of the old postpharyngeal region, all parts of the intestine except the longitudinal axial intestine degenerate com- pletely in much less time than that required for reduction in nor- mal animals. These cases present certain peculiar features: here the intes- tinal material is present, but apparently for some reason the lateral branches are unable to persist in the region which undergoes redifferentiation. In later stages in the longer pieces small new intestinal branches usually develop from the axial intestine in the redifferentiated region. When the new postpharyngeal region is formed from the anterior half or more of the old postpharyngeal region, no such intestinal degeneration takes place. How are these peculiar phenomena to be interpreted? ‘The hypothesis that the intestinal material of the redifferentiating region is used up as nutrition for the growth of this region may serve to account for the rapid disappearance of the products of degeneration, but it does not serve to account for the degeneration of the lateral intestinal branches alone, while the axial intestine persists. Moreover, the process cannot be regarded in the light of an adaptation, for it is certainly not economical of material and energy, neither does it fit the animal better in any way for contin- ued existence. On the contrary, it appears to be a useless destruc- tion of structures of great importance, a waste of energy, and in every way a process which must result to the disadvantage of the animal. But when we consider these cases from the functional stand- point they appear in an entirely different light. In the functional redifferentiation of a part of the prepharyngeal region into a whole postpharyngeal region certain changes in the mechanical condi- tions must occur. After such redifferentiation contraction of the body forces the intestinal contents in this region in the anterior Studies on Regulation SHE direction and extension in the posterior direction, whereas the reverse was originally the case. ‘The intestinal contents now tend to enter the more anterior branches of the region during contrac- tion and the more posterior during extension, but the branches were previously subjected to conditions the reverse of these. These altered conditions must bring about a very different distri- bution of the pressures and strains on the various parts of the intes- tine in this region. If the outline, arrangement and direction of the intestinal branches is determined in any marked degree by mechanical factors connected with the presence and movements of fluid contents, it seems impossible to doubt that such an extreme change in these factors must result either in a transformation of the original structures or in their disappearance, for they are the product of conditions the reverse of those now existing. Appar- ently the change is too great to permit transformation and the old structures disappear. Moreover, if these mechanical conditions determine the intes- tinal changes in these cases, the persistence of the axial intestine is to be expected, for the functional conditions in it remain essen- tially as before, the direction of movement of the contents being merely reversed in each particular instance. Only slight quanti- tative changes, if any, are to be expected, therefore, in the axial intestine. As a matter of fact, the only change observed in the axial intestine in these pieces is a change in diameter in different regions. Instead of remaining larger as originally in case the piece was prepharyngeal, the posterior part becomes smaller than the anterior, a change which is doubtless correlated with the new functional conditions. But the fact that the pieces in which the new postpharyngeal region redifferentiates from a short anterior portion of the old postpharyngeal region show the same rapid disappearance of the intestinal branches may perhaps be regarded as an objection to this hypothesis. It may be said that in these cases the mechanical conditions are not altered in the same manner and degree as 1n the prepharyngeal pieces and that the intestinal degeneration cannot, therefore, be due to such alteration. This objection cannot hold, however, as a moment’s consideration will show. In these cases 378 C. M. Child a short anterior portion of the postpharyngeal region becomes functionally a whole postpharyngeal region and de change in mechanical conditions, although not a reversal as in prepharyngeal pieces, is without doubt great. ‘The fact that when half or more of the original postpharyngeal region remains no degeneration, or practically none, except the usual slow process of reduction common to all specimens without food occurs, points in the same direction. The larger the part of the postpharyngeal region from which the new whole region is formed, the less the change in functional conditions associated with the functional regulation and the less the degeneration. The facts as to rapidity of degeneration also support the func- tional hypothesis. The lateral intestinal branches of the rediffer- entiating region disappear most rapidly in short prepharyngeal pieces, where the new postpharyngeal region is formed from a region not far posterior to the cephalic ganglia. ‘The rapidity of degeneration decreases as the level of the region from which the new postpharyngeal region is formed approaches the old pharynx. In pieces without the cephalic ganglia the rapidity of degeneration is somewhat less than in pieces with the ganglia. In those pieces in which the new postpharyngeal region redifferentiates from a short anterior portion of the old postpharyngeal region the dis- appearance of the intestinal branches 1s still less rapid than in the longer prepharyngeal pieces and, as noted above, in those cases where half or more of the old postpharyngeal region remains, the branches persist and undergo reduction in the usual manner. It is not difficult to understand from the functional standpoint why these differences in rapidity of degeneration should occur. The change in the mechanical conditions in the intestine must be greatest when a region originally just posterior to the cephalic ganglia redifferentiates into a postpharyngeal region and least when the new posterior end is formed from a large part of the old postpharyngeal region. Between these two extremes the change is intermediate in degree. Evidently then the rapidity of degener- ation in these cases is, as might be expected, parallel to the degree of change in the mechanical functional conditions. In the pieces without the ganglia movement is somewhat less energetic and less Studies on Regulation 379 frequent, hence the change in conditions in the region undergoing regulation is less extreme than when the ganglia are present, and degeneration is therefore somewhat less rapid than in pieces with ganglia. As will appear below, however, this is true only for headless prepharyngeal pieces of considerable length in which but little of the anterior end posterior to the ganglia has been removed. In short pieces neither a new postpharyngeal region nor a new pharynx is formed and the intestinal changes are very different from those described above. The development of new short and slender intestinal branches in the postpharyngeal region after redifferentiation in the longer pieces is 1n all probability also a response to a functional stimulus. These branches correspond in arrangement and direction to the branches in a normal postpharyngeal region (Fig. g). Their failure to appear in the shorter anterior prepharyngeal pieces 1s undoubtedly due to the fact that in these pieces the intestinal con- tents are used up more rapidly than in longer pieces, probably in consequence of the extreme activity which is characteristic of the short pieces: perhaps also the terminal region of the intestine con- tains less reserve material than other parts. ‘Thus the intestine becomes almost completely empty and very thin-walled after about two months in pieces including only the anterior fourth of the pre- pharyngeal region, and the pieces die, while in pieces including the anterior three-fourths of this region this condition is not reached after about five months. ‘Thus in the short pieces there is proba- bly neither sufficient nutritive material available nor sufhcient intestinal contents to furnish a stimulus to the formation of new intestinal branches in the redifferentiated postpharyngeal region. It is of interest also to note that when new intestinal branches appear in the redifferentiated region they never develop to larger size than the intestinal branches of other regions which are under- going reduction. It seems difficult to account for this early ces- sation of development on any other than a functional basis, but according to this hypothesis it is difficult to see why development should proceed farther, for the functional conditions connected with the presence of fluid contents are similar, quantitatively, in this region as elsewhere. 380 C. M. Child In short, when we consider the various features of this peculiar regulation as primarily “functional adaptations”’ or better as func- tional regulations, the morphological changes and results are readily interpreted. Moreover, I fail to see any other possible basis for interpretation. That additional factors may be involved, which have not been recognized, is extremely probable, but the facts themselves seem to me to indicate that mechanical conditions play a large part in determining the character of the functional regulation, which, in my opinion, is the basis of the morphological changes. Undoubtedly the process is, at least in large part, a complex physiological reaction, not a simple mechanical distortion. 2 Intestinal Regulation in Correlation with Anterior Form- Regulation Anterior regulation is complete only at levels anterior to, through, and immediately posterior to the cephalic ganglia. The parts replaced are replaced chiefly by regeneration from the cut surface. In the case of regeneration of the head the ingrowth of the intestine into the new tissue requires no special consideration here, since it is similar to intestinal regeneration in various other species of turbellaria. In certain cases, however, in which the anterior end has been removed posterior to the ganglia and no new head 1s formed, cer- tain features of interest appear and these are considered briefly below. Pieces from which the anterior end has been removed at a level not far posterior to the cephalic ganglia (a, Fig. 1) behave and react more like normal animals than pieces from which more of the anterior end has been removed: they are more active and react to slighter stimuli than the other headless pieces, but do not regen- erate heads, although they produce more new tissue anteriorly than the others (Child ’o5a, ’05c). In such pieces the only visible changes in the intestine consist in reduction of a type resembling that observed in normal animals. Fig. 13 shows a piece from which the anterior end was removed at a level corresponding to a, Fig. 1. The specimen was originally somewhat smaller than that Studies on Regulation 381 drawn in Fig. 1, so that the difference in size of the piece in Figs, 1 and 13 1s not wholly due to reduction. Fig. 13 repre- sents a stage 143 days after section. Comparison with Fig. 6, a normal specimen kept for about the same length of time without food shows that reduction is somewhat more advanced in the head- less piece than in the normal animal. In such pieces, however, the axial intestine, especially in the prepharyngeal region, appears to be more or less distended by the dark colored products of degen- ; fehl hehe Rehatete Py fovr 43 Fics. 13 AND 14 eration whose movements can readily be followed. This sub- stance accumulates in headless pieces to a greater extent than in pieces with heads, undoubtedly because of the fact that these pieces, being less active than normal animals and pieces with heads, require less nutritive material and so do not use up the prod- ucts of intestinal degeneration as rapidly as do the other pieces. Consequently the products accumulate in the intestine and, since the movements do not force the intestinal contents into the lateral 282 C. M. Child >) branches as frequently nor as strongly as in cases where the head is present, their effect appears chiefly in distension of the axial intestine. In the case of the pieces shown in Fig. 13 the pre- pharyngeal axial intestine 1s about the same diameter throughout its length, while the postpharyngeal axial intestine decreases in diameter posteriorly, because with the loss of the head the char- acteristic movements of the anterior end disappear to a large extent and conditions are much the same throughout its length, while the postpharyngeal region still exhibits the same regional functional differences as before, though its activity is somewhat decreased. In cases where a somewhat longer portion of the prepharyngeal region is removed a second smaller pharynx appears (Child ’o5c). The pharynx varies in position according to the level of section. Where most of the prepharyngeal region remains it is usually a considerable distance posterior to: the old pharynx, but in cases where most of the prepharyngeal region is removed it may be almost identical in position with the old pharynx and in such cases its formation involves the degeneration of the old pharynx. For the discussion of these cases in relation to functional regulation the reader is referred to my earlier paper (Child ’o5c). The point of importance for the present consideration is that a new pharyngeal region is formed in these cases and a new pharynx arises init. Butin many cases the old pharynx persists, at least for a time, so that conditions are different from those in other pieces. In Fig. 14 a piece of this kind is shown at a stage 143 days after section. The level of section corresponds to c in Fig. 1. The old pharynx lies some distance anterior to the new and the intestine shows certain features of interest. Anterior to the old pharynx the axial intestine is large and much distended with the products of degeneration, but only short stumps of the lateral branches remain. Between the two pharynges, however, lateral branches are present, but both these and the axial intestine are very slender. For a short distance posterior to the second pharynx both axial intestine and lateral branches are large and filled with the dark substance, while in the more posterior regions they show the usual Studies on Regulation 383 features of reduction. ‘The condition, of the intestine and the visible movements of the intestinal contents in these and similar pieces in the later stages indicates that in the course of reduction in size the intestine sooner or later becomes occluded in the region of the pharynx or pharynges. During reduction in size the old pharynx is not reduced proportionally and in pieces which have been without food for several months it is often so large in propor- tion to other parts as to cause a bulging of the body- al dorsally and ventrally in its region. It is “eo belill that in such cases the pressure upon the intestine in the pharyngeal region is sufficient to prevent to a large extent the passage of intestinal contents through it. Similarly the development of a new pharynx, posterior to the old, as in Fig. 14, may likewise sooner or later occlude the slender axial intestine in this region. ‘This being the case, the intestinal contents in_ the saee herrea and postpharyngeal regions do not enter the “interpharyngeal” region to any great extent, if at all. Consequently the course of intestinal manletis in this region is largely independent of that in other parts of the body. In the case shown in Fig. 14 this region contains but little fluid and both axial intestine and branches are slender, but since removal of the anterior end does not modify the muscular activities in this region except quantitatively to some extent, the branches have not entirely disappeared. . In the prepharyngeal region (Fig. 14), on the other hand, con- ditions are widely different. Here the axial intestine is greatly distended with a large amount of the dark substance. If Fig. 14 be compared with that portion of Fig. 1 posterior to the level c, which represents approximately the proportions of the piece at the time of section, it will be observed that the prepharyngeal region of the piece has decreased in size much more than the post- pharyngeal region, doubtless, as was suggested in an earlier paper (Child ’o5c), because the energy of functional conditions in this region underwent a greater decrease with the loss of the head than in the postpharyngeal region, and so the former region has served in part as nutritive material for the latter. But the effect of this reduction in size on the intestine has been to hasten degeneration of the lateral branches in this region, since the movements of intes- 384 C. M. Child tinal contents have decreased in frequency and strength with the similar decrease in muscular activity. Moreover, the reduction in length of this region has resulted in confining the fluid contents which remain within a smaller space and so in filling this portion of the intestine more completely, since the products of degeneration form more rapidly than they undergo resorption. ‘This pre- pharyngeal region of the body after removal of the head shows little differentiation of function, 7. ¢., the anterior end retains only in slight degree the characteristic motor reactions, hence the func- tional conditions are very similar throughout as regards the intes- tine, so that intestinal reduction shows no marked regional differ- ences. In the region posterior to the second pharynx, however, func- tional conditions remain much the same as in the normal animal (Child ’o5c), for the removal of the head affects the activities of the posterior end but little. Consequently contraction forces the intestinal contents anteriorly until they reach the region of the second pharynx, which they cannot pass, and so are forced into the lateral branches of this region and distend these. “The second pharynx appears rather late and before its development the intes- tinal banches just posterior to it often undergo more or less re- duction and after it appears enlarge again, very evidently in response to the altered functional conditions. ‘The movements of the dark substance can be observed very clearly 1 in this part of the intestine and the distension, accompanying contraction of the body, of the lateral branches just posterior to the second pharynx is very evident, The fact that the products of degeneration accumulate in the intestine to a much greater extent in headless pieces than in normal animals and pieces with heads is a point of considerable interest. This accumulation cannot be simply the consequence of greater degeneration in these pieces, since in many cases it occurs in stages where degeneration is less advanced than in pieces with heads where no such accumulation exists. In normal animals and pieces with heads these products undoubtedly undergo more rapid resorp- tion than in other pieces and are used to a greater or less extent as nutritive material for other parts, as has already been noted. Studies on Regulation 385 Their accumulation in headless pieces must be due to the fact that these pieces use the nutritive material less rapidly than those where the head is present. ‘This is to be expected from the differences in activity between headless pieces and others. Moreover, it will be shown in the following section that these products accumu- late more rapidly and to a greater extent in the intestine as the activity of the piece decreases. The absence of correlation between intestinal degeneration and the accumulation of the prod- ucts of degeneration within the intestine indicates very clearly that the degeneration or persistence of the intestinal branches does not depend primarily on nutritive conditions. If movements are slight, intestinal degeneration may proceed more rapidly in pieces where a considerable quantity of the detritus, which undoubtedly possesses nutritive value, is present than in cases where the intes- tine is almost empty. Thus, for example, in the case just dis- cussed (Fig. 14) the intestinal branches in the prepharyngeal region have undergone much more complete degeneration in 143 days than in a normal animal (Fig. 2), although the products of degener- ation have accumulated in the headless piece to a much greater extent than in the other. In this case then, as in those discussed above, the visible regula- tory changes in the intestine are very evidently primarily func- tional regulations and are much more closely associated with the mechanical than with the nutritive conditions, 7.e., they are func- tional regulations in response to mechanical stimuli. When a larger portion of the prepharyngeal region is removed, the second pharynx appears nearer the old pharynx, until in cases where the level of section is not far anterior to the old pharynx, this may persist, or it may degenerate and a small pharynx appear in approximately the same position (Child osc). ‘These cases present no new features of special importance as regards intestinal regulation. In some pieces the occlusion of the intestine by the pharynx or pharynges appears to be less complete than in others, and in such cases the peculiar conditions shown in Fig. 14 are less marked. When the level of section lies immediately posterior to the old pharynx, a new pharynx is often formed at the anterior end of the - 386 C. M.-Ghild piece (Child ’o5c): in such cases the intestinal changes are essen- tially similar to, though more rapid than those in the postpharyn- geal region of normal animals. Headless pieces entirely without a pharynx are discussed in the following section. IV INTESTINAL REGULATION IN PIECES WITHOUT A PHARYNX As was pointed out in an earlier paper (Child ’o5c), the isolated postpharyngeal region possesses the power of functional regulation only in slight degree. When the plane of section is immediately posterior to or near the old pharynx a new pharynx 1s often formed at the anterior end of the piece, but there is no visible rediffer- entiation of a part of the piece into a prepharyngeal region. In many such pieces, however, and in all postpharyngeal pieces in which the level of section is any considerable distance posterior to the old pharynx, a new pharynx does not appear, /. ¢., these pieces do not possess sufficient power of functional regulation to give rise to any of the other regions of the body. The same is true of pieces below a certain length from’ the prepharyngeal region pos- terior to the ganglia. But these pieces, although they remain wholly without a pharynx and show practically no regeneration beyond wound-closure and no regulatory formation of other regions by redifferentiation, do present certain remarkable fea- tures as regards intestinal regulation. Such pieces show few of the typical reactions (Child ’o§a, ’o5c): they do not usually attach themselves to the substratum, but are merely propelled through the water by their cilia: they rarely extend to full length and in course of time become greatly short- ened and rounded and show almost no muscular activity beyond slight contractions and extensions and peristaltic waves which pass from one end of the body to the other. Many such pieces, however, were kept under observation during 143 days and the experiments were concluded only because of my departure from Naples. Two such pieces are selected for description: all others observed are essentially similar. The first of these was a short prepharyn- geal piece, including approximately the region between the levels Studies on Regulation 387 = d and e in Fig. 1. Fig. 15 shows the piece twenty-six days after section. It 1s much reduced in size and degeneration of the intes- _tinal branches has been very rapid. At this stage the axial intes- tine was distended by a large quantity of the products of degener- ation. This rapid degeneration and the accumulation of the prod- ucts of degeneration certainly cannot be interpreted as the result of lack of food. ‘The intestine was well filled at the time of section and by no means all of its contents were lost through the wound; moreover, the motor activity of the piece is very slight and its need for food is therefore less than that of more active pieces; it has not formed extensive new parts either by redifferentiation or regener- ation. If nutrition is the primary factor in determining the degen- eration or persistence of the intestinal branches, we should cer- tainly expect that they would degenerate very slowly in such pieces. Yet they degenerate more rapidly than in any other case except Fics. 15, 16 AND 17 those in which a part of the prepharyngeal region redifferentiates into a postpharyngeal region. About forty days after section no trace of the intestinal branches remained. If the experiment were carried no further, such pieces might be regarded as cases of “reversal of development,” similar to those described by various authors as occurring during star- vation. But after sixty-five days the piece presented the appear- ance shown in Fig. 16. A complete new set of very slender and delicate intestinal branches had developed and these persisted as long as the piece was kept under observation, but underwent gradual reduction as the size of the piece decreased in size (Fig. ¥7, 1g0:days). The history of postpharyngeal pieces without a pharynx is essen- tially similar. Figs. 18 to 20 give three stages in the history of a piece corresponding to that portion of the body posterior to the 388 C. M. Child level g in Fig. 1. Fig. 18 shows the condition of the piece forty- five and Fig. 19, sixty-five days after section. At the latter stage all traces of intestinal branches have disappeared and the wall of the axial intestine, which is greatly distended with the products of degeneration, is as smooth as that of a rhabdoccel intestine. But forty days later, 7. ¢., 105 days after section (Fig. 2c), new slender branches had developed along the whole length of the axial intestine, and these underwent gradual reduction, but were still visible when the experiment was concluded at 143 days. The development of the new intestinal branches in these pieces usually requires some twenty days or more. ‘The various stages were examined with great care, in many cases under pressure, and there Fics. 18, 19 AND 20 is no possibility of error. Their development follows in every case the great distension of the axial intestine with the products of degeneration. Similar intestinal changes were observed in every piece incapable of forming a new pharynx, provided it did not die too early. In general the rapidity with which the changes occur increases with decrease in the length of the piece. This difference in rapidity is well shown in the two pieces selected for descrip- tion. In the first, which is considerably shorter than the other, all traces of the intestinal branches were lost within forty days and the new branches were present after sixty-five days, while in the second, 7. e., the longer piece, the degeneration of the branches required sixty-five days and the development of the new branches, forty days more, in all 105 days. Studies on Regulation 389 In these pieces then, although there 1s no other visible regula- tion except wound-closure, the original intestinal branches undergo complete degeneration and a new sét of shorter and more slender branches develops in their place. Moreover, the development of new intestinal branches occurs only after two or three months, during which time the pieces have undergone considerable decrease in size. Evidently it is not correlated with other processes of form- regulation in these pieces. These remarkable phenomena seem to me to constitute a prac- tical demonstration of the hypothesis which has served as the basis for interpretation of the other phenomena of intestinal regulation in this species. As a matter of fact they play an important part in the development of the hypothesis in my mind. A brief dis- cussion will serve to show clearly their correlation with mechanical conditions. In the first place muscular activity is relatively very slight in these pieces, consequently the movements of the intestinal contents are also relatively slight. Under these conditions the intestinal contents accumulate, as can readily be observed, in the axial intestine and enter the branches but little, except when more powerful contractions are induced by artificial stimulation. ‘The old intestinal branches, not being adapted to these conditions, undergo very rapid degeneration and only the axial intestine remains. ‘The persistence of this part of the intestine is to be expected, since all muscular contractions cause movements of its contents, and since these are accumulating as time goes on. These pieces require little nutrition in consequence of their rel- atively slight activity, hence the products of degeneration do not undergo resorption as rapidly as they are formed, but accumulate in the intestine to such an extent that they distend it greatly, and finally bring about the formation of a new set of intestinal branches, which are adapted to the new conditions, and which undergo gradual reduction as these conditions change in following stages. Only in pieces where the axial intestine becomes distended with the products of degeneration do these new branches appear. In general, as the length of the piece decreases, the rapidity of degen- eration of the old branches increases and the distension of the 390 C. M. Child axial intestine and the development of the new branches occur earlier. ‘This difference in the rapidity of change is also exactly what might be expected according to our hypothesis, for the char- acteristic muscular activity and consequently the movements of the intestinal contents into the branches decrease as the length of the piece decreases, hence the shorter the piece, the greater the change in mechanical conditions affecting the intestinal branches. If the mechanical conditions are the determining factors, it follows that the rapidity of degeneration must increase with decreasing length of the piece. But increased rapidity of degeneration results in more rapid accumulation of the products of degeneration in the axial intestine and so in earlier development of the new branches. This interpretation seems to me the only one possible. ‘These cases show very clearly that the factors which determine the degeneration or the development of a structure are not necessarily associated primarily with nutrition or its absence. Development without energy is of course impossible and this energy must come from nutritive material of some sort. But the mere presence of the material does not necessarily determine that a given structure shall develop. ‘That, as I have endeavored to show in most of the papers of this series and in others as well, is determined by func- tional conditions in the widest sense. V THE RAPIDITY OF GENERAL INTESTINAL REDUCTION UNDER DIFFERENT CONDITIONS Intestinal reduction in the whole body or piece proceeds with very different rapidity in different cases. The rapidity of reduc- tion in certain special cases has already been discussed in the pre- ceding sections, but a general comparison of the various cases presents certain features of interest since it shows very clearly that nutritive conditions are, at least in certain cases, not the only, nor even the most important, factors in determining the rate of intestinal reduction. In the first place intestinal reduction proceeds more slowly in the normal animal without food than in headless pieces of any size. This is evident from a comparison of the figures. Fig. 6 repre- Studies on Regulation 391 sents the condition of a normal animal after about four and one- half months without food; Fig. 13 shows a piece including almost the whole body except the head after the same length of time without food. Shorter headless pieces differ still more widely from the normal animal; Fig. 15 shows a short, headless, pre- pharyngeal piece after twenty-six days of starvation and Fig. 18 a headless postpharyngeal piece without pharynx after sixty-five days. In these pieces total, or almost total, disappearance of the intestinal branches has occurred in a period of time from less than one-sixth to about one-half that necessary for intestinal reduction in the normal animal to the condition shown in Fig. 6. Such differences as these cannot be due to differences in nutri- tive conditions. “The normal animal is more active and must use a greater amount of nutritive material in proportion to its size in a given time than a headless piece such as that shown in Fig. 13, and its activity and nutritive requirements must be many times greater than those of the headless pieces shown in Figs. 15 to 17 and 18 to 20, in which movement ts slight, yet in all of these pieces intestinal degeneration is more rapid than in the normal animal. Moreover, 1m headless pieces the rapidity of intestinal reduction increases with decrease in the length of the piece. ‘The long piece in Fig. 13 (all that part of the body posterior to a in Fig. 1) reaches the condition shown in 143 days, the piece shown in Figs. 18 to 20 (that part of the body posterior to g in Fig. 1) loses all traces of intestinal branches in sixty-five days, and the piece shown in Figs. 15 to 17 (that part of the body between d and ¢ in Fig. 1) loses all traces of intestinal branches in less than forty days (Fig. 15, twenty-six days). In these pieces, and in all similar pieces observed, the rapidity of degeneration of intestinal branches 1s in general inversely proportional to the length of the piece. At the time of section the amount of nutritive material in these various pieces must be about the same in proportion to their size, Of course some differences exist in this respect and there is more loss from the wound in some cases than in others, but contraction is usually so rapid that loss from the wound is slight. The two pieces whose history is given in Figs. 15 to 17 and Figs. 18 to 20 were taken from the same worm; nutritive conditions must there- 392 C. M. Child fore have been very similar in both at the time of section. But the muscular activity of the headless pieces decreases in general with decrease in length. The longer pieces must therefore use up nutritive supplies more rapidly than the shorter pieces and if degeneration were due to lack of nutrition it must occur earlier and proceed more rapidly in the longer than in the shorter pieces. But exactly the reverse is the case. Moreover, the formation of new intestinal branches several months after section and after the old branches have undergone complete degeneration shows very clearly that sufficient nutritive material is present to allow the development and maintenance of the intestinal branches, when the proper stimulus is present. It seems impossible, therefore, to escape the conclusion that nutritive factors are not the most important in determining the rapidity of intestinal degeneration. But when we consider the dynamic conditions resulting from the presence and movements of the intestinal contents, it at once becomes evident that the rapidity of degeneration is in general proportional to the change in these conditions. In the normal animal these conditions remain most nearly normal and after the food taken from without has disappeared from the intestine it still contains a certain quantity of fluid, which moves about in the char- acteristic manner, though its effect must be quantitatively less than when the intestine is well filled. In headless pieces the movements differ more or less from those of the normal animal and are always less energetic and less frequent, hence the functional stimulus from the contents must be less than in normal animals and intes- tinal degeneration must occur more rapidly in such pieces than in normal animals if it 1s correlated with decrease or absence of these stimuli. Moreover, motor activity of all kinds decreases with decreasing length of the headless pieces and the mechanical stimuli arising from the intestinal contents must decrease similarly, espe- cially in the lateral branches, since the less powerful the muscu- lar contractions, the less frequently do the intestinal contents enter the branches. Consequently degeneration of the intestinal branches must occur with increasing rapidity as the length of the piece decreases, if it is connected with these conditions. As shown above, the facts correspond exactly with the require- Studies on Regulation 393 ments of the hypothesis and | fail to see how any other interpre- tation of them is possible. ‘They all indicate that the rapidity of degeneration of the intestinal branches is dependent, at least in large measure, on the degree of change in the mechanical func- tional conditions connected with the presence and movements of the intestinal contents, irrespective of their nutritive value. But when we compare normal animals with pieces which pos- sess heads the case 1s somewhat different. ‘The very rapid intes- tinal degeneration in the redifferentiating regions of such pieces has already been discussed in Section III c, and does not concern, us here, but the rapidity of intestinal degeneration in other parts of the body differs from that in normal animals and also differs according to the length of the pieces. In such pieces the activity remains the same as in normal animals, or in short pieces includ- ing little besides the head- “region, is apparently even greater than feel. Consequently these pieces must require as men nutri- tive material in proportion to their size as do normal animals, or probably even more in the case of short pieces. Moreover, these pieces undergo qualitatively complete form-regulation, producing a new postpharyngeal and pharyngeal region with a new pharynx. These changes must also require nutritive material. In such pieces the intestinal contents decrease rapidly in amount—the shorter the piece, the more rapid the decrease—and those portions of the intestine remaining never contain any considerable amount of the products of degeneration as do those of the shorter headless pieces. These products appear to undergo resorption almost as rapidly as they are formed. Consequently the quantity of intestinal con- tents becomes very small and the axial intestine and all other parts become very slender (Compare for example Figs. g and to with Figs. 15 and Figs. 18 and 19). As noted above this difference indicates that the products of degeneration serve as nutritive mate- rial. Since this does not accumulate to any extent in the pieces with heads the intestine becomes almost empty and, notwithstand- ing the normal movements of these pieces, the mechanical stimu- lation of the intestinal walls must be very slight and must decrease centrifugally. Consequently the branches disappear and the rapidity of degeneration is determined, at least in part, by the 3094 C. M. Child rapidity with which the intestine 1s emptied of lis contents in con- sequence of the demand for nutritive material. ‘Therefore intes- tinal degeneration increases in rapidity with decreasing length of the pieces, as is the case in the headless pieces. According to this interpretation the increasing rapidity of degen- eration with decreasing length is due in headless pieces primarily to decreasing movement of the intestinal contents in consequence of decreasing muscular activity, while in pieces with heads it is due primarily to decreasing quantity of intestinal contents in conse- quence of the great demand for nutritive material. In the headless pieces nutritive material arising from the degeneration of the intes- tinal branches accumulates in the remaining portions of the intes- tine, but the branches disappear in spite of its presence. In the pieces with heads, on the other hand, this material is used up almost as rapidly as it is formed and the branches disappear because the intestine is nearly empty. In the first case the degen- eration 1s apparently due largely to lack of movement of the intes- tinal contents, in the other to lack of intestinal contents to be moved. Thus the data concerning the rapidity of intestinal degeneration serve still further to support and confirm the conclusion that intes- tinal regulation in this species is in large part a functional regula- tion in response to mechanical stimult. VI CONCLUSION AND SUMMARY The phenomena of intestinal regulation certainly afford strong support to a dynamic or functional hypothesis of regulation and in this respect are in accord with various other phenomena in this and other species, which I have described and discussed in previous papers. ‘he intestine retains its typical form, or returns to it, only when dynamic conditions are, or become similar to those which give rise to the typical form. Extensive intestinal regulation may Occur in the absence of other form-regulation, or intestinal regulation may fail to occur, while other parts undergo more or less complete regulation. ‘The results in each case are correlated with the dynamic conditions in the intestine, particularly the mechani- Studies on Regulation 395 cal conditions, and can be interpreted only on the basis of this correlation. ‘There can be little doubt that intestinal regulation in various other species of turbellaria will prove to be similarly dependent on mechanical conditions. The history of the pieces without pharynges shows how little significance there is in description or discussion of ‘‘reversal of pe rclnpatent” without consideration of the dynamic factors in- volved. When these dynamic factors act in reverse sequence and direction from that typical of normal development, then, and not otherwise, does reversal of development occur. ‘There is no law, such as certain authors seem to postulate, that causes an organism to return more or less completely to an earlier stage of develop- ment if deprived of food, or under other changed conditions. The so-called “return” usually consists simply of the loss of previ- ous differentiation, but this does not necessarily constitute a rever- sal, for the method of loss may be very different from the method of acquirement, as in the present case. Moreover, the loss of the original structure or differentiation may be merely the first step in the development of something new in response to altered con- ditions, as is the case in the pieces without pharynges. hese pieces are in no sense returning to an earlier stage of development or “embryonic condition,” because they lose the old intestinal branches, but are merely undergoing a process of functional adap- tation or regulation. The gradual simplification in intestinal structure, which occurs in various planarians in the course of starvation and reduction in size, is undoubtedly essentially a func- tional regulation just as truly as is the appearance of new branches under other conditions. - Objection to my interpretation of the facts may perhaps be made on the ground that the recent experiments of Babak (’06) with amphibia indicate that chemical factors are much more important than mechanical in determining intestinal regulation. It can scarcely be doubted, however, that the amphibian intestine differs greatly from the turbellarian in function. As I have pointed out in Section I, the turbellarian intestine is much more than a digestive organ, being both a storage-reservoir for excess of undigested food- material and to a considerable extent a circulatory system. It 306 C. M. Child would be remarkable if mechanical factors were not much more important functionally in the turbellaria than in the amphibia. In order to interpret regulatory phenomena it 1s of the utmost importance to consider all the functions of an organ or structure and not merely one, or the most conspicuous. Only in this way shall we attain complete interpretation. It must be borne in mind that the name assigned to a part does not always indicate fully or exactly its functions, nor is the function commonly assigned to it necessarily its only function: in most cases it is merely a small part of the actual function. In the present case the changes in mechanical conditions are to a certain extent visible and accessible to experimental methods and, as I have endeavored to show, the processes of regulation in the intestine are evidently closely correlated with them; indeed it is impossible to account in any other way for certain of the changes, such as the rapid degeneration in a postpharyngeal region formed by redifferentiation and the development of new intestinal branches in pieces without pharynges after months of starvation. More- over, while other factors, such as the character of the food and the digestive activity, doubtless affect the structure of the cells and very probably their number, it is difficult to understand how fac- tors of this kind alone can determine the form, arrangement and direction of intestinal branches. ‘hese elements of the intestinal form must, it seems to me, be determined mechanically, at least in large part, and it 1s with these that the present paper is primarily concerned. The most important results are briefly stated in the following summary: 1 In normal animals kept for several months without food extensive intestinal degeneration occurs, beginning in the peripheral regions and proceeding toward the pharynx. ‘This degeneration involves chiefly the lateral branches and affects the axial intestine only in the terminal regions. 2 In pieces undergoing regulation without food in which a postpharyngeal region is formed by redifferentiation from a part of the old prepharyngeal or the anterior part of the old postpharyn- geal region, the old lateral branches of the intestine undergo rapid Studtes on Regiianan 397 and complete degeneration in the redifferentiating region and are replaced in the longer pieces by new branches, corresponding in arrangement with those of a normal postpharyngeal region. 3 In headless pieces which have not sufficient regulatory capac- ity to give rise to a new pharynx (short prepharyngeal pieces and most postpharyngeal pieces) the old intestinal branches undergo rapid and complete degeneration, but after two months or more a new set of short and slender intestinal branches arise, which per- sists, but undergoes gradual reduction as time goes on. 4 In all other pieces undergoing regulation without food intes- tinal reduction occurs and usually proceeds from the peripheral towards the middle regions, though special modifications occur with special conditions. 5 The intestine of polyclad and triclad turbellaria is not merely a digestive organ, but functions also as a reservoir for the tempo- rary accumulation and storage of undigested food-material and also, to a considerable extent, as a circulatory system. Its con- tents are largely fluid and undergo movement in consequence of the muscular contractions of the body-wall. The presence and movements of these contents must produce characteristic mechan- ical effects upon the intestinal wall. 6 ‘The facts of intestinal regulation indicate that these mechan- ical conditions play an important role in determining the outline of the intestine and the direction and arrangement of the branches. Total disappearance of the old branches occurs when the mechan- ical conditions are widely altered, even though nutritive material be present in excess. The rapidity of degeneration depends on the degree of change in the mechanical conditions. The develop- ment of new branches after degeneration of the old is determined primarily, not by the presence of nutrition, but by mechanical con- ditions, though of course nutritive material is necessary for such development. Undoubtedly certain features of intestinal regulation are deter- mined by other functional factors, but the general outline and the arrangement and direction of branches are very evidently closely correlated with mechanical factors. 398 C. M. Child BIBLIOGRAPHY BasAk, E., ’06—Experimentelle Untersuchungen tber die Variabilitat der Ver- dauungsrohre. Arch. f. Entw-mech., Bd. xxi, H. 4, 1906. BarDEEN, C. R., ’or—On the Physiology of the Planaria maculata with Especial Reference to the Phenomena of Regeneration. Am. Jour. Physiol., vol. v, IgOl. °o2—Embryonic and Regenerative Development in Planarians. Biol. Bull., vol. ii, no. 6, 1902. ’03—F actors in Heteromorphosis in Planarians. Arch. f. Entw-mech. Bd. xvi, H. 1, 1903. Cuitp, C. M., ’o4a—Studies on Regulation. V. The Relation Between the Central Nervous System and Regeneration in Leptoplana: Pos- terior Regeneration. Journ. Exp. Zool., vol. i, no. 3, 1904. ’°ogb—Studies on Regulation. VI. The Relation Between the Cen- tral Nervous System and Regeneration in Leptoplana: Anterior Lateral Regeneration. Journ. Exp. Zool., vol. 1, no. 4, 1904. ’o5a—Studies on Regulation. VIII. Functional Regulation and Re- generation in Cestoplana. Arch. f. Entw-mech. Bd. xix, H. 3, 1905. ’o5b—Studies on Regulation. IX. The Positions and Proportions of Parts During Regulation in Cestoplana in the Presence of the Cephalic Ganglia.. Arch. f. Entw-mech., Bd. xx, H. 1, 1905. ‘o5c—Studies on Regulation. X. The Positions and Proportions of Parts During Regulation in Cestoplana in the Absence of the Cephalic Ganglia. Arch. f. Entw-mech., Bd. xx, H. 2, 1905. ’°o6a—Contributions Toward a Theory of Regulation. I. The Signif- cance of the Different Methods of Regulation in Turbellaria. Arch. f. Entw-mech., xx, H. 3, 1906. °o6b—The Relation Between Functional Regulation and Form-Regula- tion. Journ. Exp. Zodl., vol. i, no. 4, 1906. Lituir, F. R., ’o1—Notes on Regeneration and Regulation in Planarians. II. Am. Journ. Physiol., vol. vi, 1got. PoE BenAVIOR OF LOXOPHYLLUM: AND: ITs RELA- TION TO REGENERATION BY S. J. HOLMES Witu Seven Ficures GENERAL CHARACTERISTICS OF THE SPECIES ‘The general form of Loxophyllum meleagris, the species studied, is flattened and leaf-like, and tapering toward the anterior end which is turned toward the dorsal margin. ‘The anterior third or fourth of the body is flatter and less granular than the hinder portion and the margins of the body are thinned out, especially along the oral side. “The middle and posterior regions are more convex and may be considerably distended when gorged with food. ‘The body is ciliated on the right side on which the ani- mal usually glides. ‘Vhe cilia are arranged in rows which extend in a longitudinal direction except near the anterior end of the body where they curve toward the dorsal side. ‘The whole oral margin is also furnished with cilia, but none could be detected on the left side of the body. The body wall is traversed with myonemes, both on the right and the left side, which extend longitudinally for the most part, but curve dorsally, like the rows of cilia, near the anterior end. ‘They are more conspicuous and apparently thicker near the anterior end of the body, and they are especially well developed near the oral side. ‘lrichocysts are abundant along the entire oral margin and around the anterior end of the body, forming a uniform series closely set at right angles to the surface. On the dorsal side the trichocysts are mainly confined to the small promi- nences, a dozen or more in number, which give that side its crenu- lated contour. Numerous trichocysts occur also on the right or ciliated side. 400 S. Ff. Holmes The contractile vacuole is nearly spherical in form and is situ- ated near the dorsal side of the body a little in front of the posterior. end. ‘There is a fine canal extending from it anteriorly along almost the entire dorsal side. A short canal may lead into it from behind. The meganucleus is composed of numerous rounded masses (over twenty in some individuals) scattered through the larger part of the body. ‘The anterior third or fourth of the body, how- ever, is usually free from nuclear material. Although the mouth of Loxophyllum is an inconspicuous slit near the edge of the body the animal is nevertheless able to ingest comparatively large forms. Rotifers form a common article of diet. I have often seen Loxophylla containing specimens of Anurea cochlearis and other rotifers of as large size, the body being thereby much distorted in shape. In ejecting the lorica after the rotifer had been digested the body is much lacerated, but its power of rapid regeneration soon causes it to assume its nor- mal outlines. I have never been able to obtain Loxophyllum in abundance. Like many other predatory infusoria it thrives only in compara- tively pure water and quickly disappears 1 in the presence of putre- fying material. It is found on aquatic vegetation, and sometimes appears on the walls of aquaria, especially those supplied with run- ning water. A favorite situation is on the side of an aquarium just below the surface of the water. NORMAL MOVEMENTS The normal movements of Loxophyllum, compared with those of most infusoria, are sluggish, a circumstance which makes it easy to study the precise way in which they are performed. The creature glides along the substrate on its right side, moving its anterior end about slowly as if feeling its way. Its usual mode of locomotion 1s as follows: It elongates the body, swims forward a short distance, then contracts, swims backward, turns toward the oral side, and then elongates, and swims forward in a new direc- tion. As it generally swims but a short distance before jerking back, the organism circles about toward the oral side in nearly the The Behavior of Loxophyllum 401 same situation. I have often observed specimens on the side of an aquarium that remained over an hour within a few millimeters of their original position, although continually moving about. It is an interesting fact that the motor reflex or avoiding reaction in Loxophyllum takes place in a direction just the reverse of that of Paramecium and many other infusoria; the turning is always to the oral instead of the aboral side. Most of the turning, how- ever, occurs after the infusorian has ceased to swim backward which makes it probable that the anterior cilia are relatively more active at this time. While swimming forward the direction of movement is at the same time more or less toward the aboral side, and the backward movements are more or less toward the oral side, but the principal change of direction occurs at the close of each backward movement. ‘The infusorian thus continually cir- cles about to the oral side. When swimming backward the body is generally bent over to the oral side, often throwing the oral margin into one or more folds. The movements of the body vary considerably in rapidity accord- ing to the degree of excitement of the animal, but I have never seen an individualin a state of absolute quiet. There isa certain regu- larity or rhythm of the forward and backward movements which is fairly constant for a long period. Most of the individuals in a dish move at a tolerably similar rate if some of them have not been more disturbed than others. At times Loxophyllum may glide forward for a considerable distance without reversing its direction, but it does this, I believe, only when in a comparatively high degree of excitement. Its body is then strongly elongated and nearly straight. In the short forward movements which are performed in its usual circling about near one place the body is not so greatly elongated and it 1s bent over more strongly toward the aboral side. ‘The extension and straightening of the body during its more rapid gliding aid the animal in maintaining a more direct course, although it com- monly veers around somewhat to the aboral side. The body of Loxophyllum is very mobile and it is able to change its shape in many ways by contracting locally in different regions. It may contract to half its maximum length, bend up or down or 402 S. fF. Holmes to either side, or twist about on its long axis. It is almost con- stantly bending and writhing about in various ways. ‘The ante- rior extremity is the most active as well as the most sensitive part of the organism. It is continually executing small movements, bending back and forth or up and down as if attempting to explore its environment. ‘The oral margin of the body at times performs a sort of undulating movement, usually when it is lifted up free from the surface. This motion when the animal is largely free from contact with the substrate may become a vigorous and rapid one and serves to turn the body about in the water. When slight the fluttering movements are confined to near the anterior end of the body but when more decided they involve a considerable part of the oral margin. When turned over so as to lie on its left or unciliated side Loxophyllum may right itself in several ways. At first it writhes about for a little while, but it is usually only a short time before one of its methods of turning over is hit upon. One common method is to raise up the ends of the body more or less, twisting about the anterior end until its right side touches the bottom. The rest of the body is then pulled over much as in the common righting movements of a planarian. Often, but not always, this is accompanied by a rapid undulation of the oral margin which apparently aids the turning. Generally Loxophyllum raises the oral side and twists about aborally, but it not infrequently turns over in just the reverse direction. Frequently the body is twisted about when the two ends are free in the water, the turning begin- ning at the anterior end and continuing until the whole body 1s twisted about. When placed on its left side Loxophyllum sometimes bends the anterior end of its body upward at right angles to the long axis, raising it until it stands erect, and then toppling over upon the opposite side. In one instance I saw both ends raised up to about the same extent until they nearly met, forming a sortof hoop; then the animal rolled over, through the force of ciliary action until the anterior end touched the bottom, when it attached and glided ahead, thus straightening out the body into its normal position. Loxophyllum is, as a rule, rather reluctant to swim through the The Behavior of Loxophyllum 403 water, but when in a state of unusual excitement it may do so quite readily. It swims in a spiral course like most infusoria, circling about in a clockwise manner and at the same time rotat- ing on its long axis in the same direction. ‘The spiral course is maintained not so much through the natural asymmetry of the body, as by the fact that the body is curved toward the inner side of the spiral and held at a slight twist. By means of its spiral movement Loxophyllum is able to travel in a nearly straight gen- eral course for a considerable distance. REACTIONS TO STIMULI Mechanical. In experimenting on the reactions of Loxophyl- lum to mechanical stimuli a glass rod was used which was drawn out into an exceedingly fine thread at the tip. By using a Braus- Driner binocular microscope it was possible to apply stimuli of various degrees of intensity to any part of the body and readily observe the result. When the anterior part of the body is stimu- lated the animal contracts longitudinally, swims backward and to the oral side, and bends its body orally at the same time. After this it extends the body again and swims forward. Stimuli applied to the tip of the body most readily produce this reaction. It may be produced even without contact by moving the rod about a short distance in front of the anterior end. Stimulating either side of the body back to a considerable dis- tance produces the same reaction. It is obvious, therefore, that when the animal is stimulated on the oral side it turns directly toward the stimulus instead of away from it. Repeated appli- cations of the stimulus to the oral side will cause the animal to keep turning toward the stimulus, notwithstanding the unadaptive, or even injurious, nature of the response. The facility with which a stimulus evokes a response diminishes toward the posterior end of the body. Stimuli applied between the middle of the body and the posterior third, especially after the second or third trial, frequently produce no response, even when quite strong. ‘The animal may often be poked about in this way, almost to the point of producing mutilation, without suffering any interruption of its usual activities. 404 . S. fe Holmes If a stimulus is applied to the posterior end of the body, or a short distance in front of this on either side, the usual motor reflex is not produced. The animal swims directly forward. With repeated stimulation of the posterior end it may be kept swimming forward for a long time. If the stimulus is applied during the progress of the animal the rate of movement isaccelerated. Essen- tially the same behavior has been found by Jennings to occur in Paramecium in response to weak stimuli, and I have often ob- served the same phenomenon in this and several other infusoria. It indicates the first step toward reacting in a specific manner to the localization of a stimulus. The reactions of Loxophyllum are quickly modified by suc- cessive responses to stimulation. ‘This, I believe, is in large part due to the dulling of the sensitiveness of the organism through the repetition of stimuli. A very slight stimulus to the anterior end of the body suffices at first to produce a reaction. With repeated poking the anterior end becomes so dulled that the organ- ism may continue to swim forward in spite of frequent stimulation at this point. Recovery, however, is quick, for in a few minutes the responsiveness is as great as ever. A similar result is more quickly reached by the application of stimuli to the sides. ‘The motor reflex may be elicited very readily for a few times, but it soon requires much stronger stimuli to bring it about. If the stimuli are applied quite far back it requires fewer stimulations before the animal refuses to respond at all. There seems to be a tendency for the organism to resume its usual activity which asserts itself when its sensitiveness becomes dulled so that it does not react so readily to stimuli. It may be kept from going forward, for instance, by repeatedly stimulating the anterior end of the body. But sooner or later the tendency to normal activity predominates and the animal may go forward in spite of considerable stimulation. Notwithstanding its rapid habituation to stimulation Loxo- phyllum exhibits certain features of behavior that seem referable to the summation of stimuli. Repeated stimulation may induce a condition of unusual excitement which may be manifested in con- tinual and quite rapid swimming, increased writhing movements, The Behavior of Loxophyllum 4.05 or in increased rapidity of its ordinary back and forth movements. The reactions may be less easily evoked, but the spontaneous activity of the organism 1s increased. ieee: Owing to the small number of specimens available few experiments on the reactions of Loxophyllum to chemicals were tried, since they frequently produced fatal effects. Several drops of water containing specimens of Loxophyllum were spread out on a slide and a minute grain of salt or a small drop of weak acid was placed at one edge. The Loxophylla showed no ten- dency to go directly away from the diffusing chemical. Some- times they would go toward it. In many cases they would move about irregularly until overcome by the chemical in case it was strong enough to be injurious. ‘The majority of the individuals, however, usually succeeded, sooner or later, in getting away to a safe distance. When swimming toward the chemical the anterior end is more strongly stimulated and the animal swims backward, turns, and goes in some other direction. ‘The length of the back- ward course being dependent on the strength of the stimulus received, the animal is apt to go back further when pointed to- ward the stimulus than when pointed away from it. Also excur- sions toward the stimulus are more quickly checked than those in other directions. In consequence of these reactions the animal works its way farther from the stimulating substance. ‘The proc- ess 1s a slow one, especially since, owing to its natural rhythm of movement, Loxophyllum frequently changes the direction of its locomotion. Even when pointed directly away from the chemical it does not usually go very far before backing up and turning in another direction, and thus much of what was gained is lost. The whole process of negative chemotaxis in this form is a very slow, uncertain, and bungling one. In one experiment I placed a lot of Paramecia with several Loxophylla, and a drop of acid was introduced at the edge of the liquid containing them. ‘The Paramecia showed a very quick and marked negative reaction. “The Loxophylla were incompar- ably longer in getting away from the chemical. Some went toward it and were killed, while practically all of the Paramecia got safely away from the injurious substance. 406 S. J. Holmes THE BEHAVIOR OF PIECES) OB LOXOPHYLEUM It has been shown by Jennings and others that pieces of infu- soria react in much the same way as the entire organism so far as this is rendered possible by the shape of the parts concerned. The observations which I have made on the behavior of pieces of Loxophyllum confirm the general results obtained by other observers and add a few points of interest, especially in relation to the subject of regeneration, which is treated in a subsequent section. A specimen was cut transversely in two at about the anterior third. ‘lhe two pieces swam rapidly apart, the anterior one going forward, the posterior one backward. For some time the anterior piece swam about in a circle toward the aboral side. After this it began to move alternately forward and aborally, and then back- ward and orally. At times the oral margin would be raised up and moved in an undulating manner like the edge of a flag, and some- times the piece would turn completely over on its left side, but it usually glided along on its right side with little or no marginal motion. After about six minutes its backward and forward excur- sions became limited to about the length of its body. In going ahead there was a slight extension of the body, while in going backward the body was always widened. A little later its motions became confined to nearly the same spot. It would go forward, then backward, turning through twenty or thirty degrees, and then go forward again. Its behavior had become, therefore, much like that of the normal animal under usual conditions. When the anterior end of the piece was stimulated by contact with a fine capillary glass rod it would swim backward and turn toward the oral side. When the posterior end of this piece was stimulated it would not react nearly so readily, and often quite strong stimuli produced no effect. When the response did occur, however, it was manifested in two different ways. At times the piece would flatten and swim backward, especially if the stimulus were strong. At other times the body would elongate and swim forward. In the first case it is probable that the animal was pushed ahead so that the more sensitive anterior end was stimu- The Behavior of Loxophyllum 407 lated, and this would naturally produce a backward movement. The second response is like that of the normal animal when irri- tated at the posterior end. When the piece was swimming through the water stimulation of the posterior end frequently resulted in a marked acceleration of its speed. The righting movements of the anterior piece were much like those of the entire animal. Generally the oral margin would be raised up and waved rapidly back and forth, a movement which probably causes the oral side to be elevated until the piece topples over upon its right side. Considerable variation occurs in the method of turning over in the pieces as in the whole organism. In another experiment in which attention was paid mainly to the movements of the posterior piece a Loxophyllum was cut in two near the middle. ‘The posterior piece swam backward quite rapidly for about three minutes. After this its movements became slower and it would swim forward occasionally. In a few min- utes the forward movements began to increase, and after a while the infusorian settled down to moving forward and backward to about its own length. Sometimes it would raise itself from the bottom and tumble over on the other side only to quickly turn again into its normal position. At one time it left the bottom and swam in a spiral course for a considerable distance. When moving backward the piece would widen out, especially at the anterior end. When moving forward, on the other hand, the piece would become elongated and strongly drawn together at the anterior end. These changes of shape were invariably associated with the different directions of movement. Stimulation of the posterior end of the piece causes it to pinch together in front, elongate, and swim forward. When the anterior cut end is stimulated the piece spreads out and swims backward. Comparatively strong stimuli are required, however, in this case as the cut end 1s considerably less sensitive than other regions of the body. By stimulating either end of the piece it may be caused to swim continually forward or backward as the case may be. In about thirty minutes after the cut was made the regeneration of the anterior end of the piece was well under way. ‘The slight waving motions of the anterior border were visible, but not so 408 S: fe Holmes apparent as in the entire organism. ‘The piece kept going for- ward and aborally a very short distance, then backward and orally, circling about in nearly the same spot, with a regular, incessant, rhythmical movement. In going forward the body became not only elongated but curved toward the aboral side. When con- traction occurred during its reversed movement, it was greater on the oral than the aboral side as it is in the entire individual. The behavior of several other pieces taken from the two ends of the body was essentially like those described. Sometimes the pieces would swim about in a spiral course through the water for several minutes, but eventually they all settled down to the same regular back and forth movements. Pieces cut from the middle of the body showed the same rhythmical movement, extending and bending slightly aborally as they went forward, and contracting more on the oral side as their motion was reversed. RHYTHMICAL ACTIVITY OF LOXOPHYLLUM When first observing the activities of Loxophyllum I came to the conclusion that the frequent reversals in the direction of its move- ments were due to reactions caused by minute objects with which the sensitive anterior end of the body came into contact. But further observation showed that these reversals were due to inter- nal rather than external causes. When specimens of Loxophyl- lum were placed in water as free as possible from small particles the same regularity of reaction was found to continue. When gliding on the upper side of the surface film of a drop of clear water Loxophyllum reverses its movement about as often as when in the midst of objects with which it is continually colliding. But any doubt concerning the inherent rhythmicality of its movements is removed when we consider that the pieces into which the body is cut move back and forth at about the same rate as the whole animal. The cut anterior ends of the pieces of Loxophyllum are com- paratively insensitive to mechanical stimuli, and there can be no doubt that the movement of these pieces is a manifestation of rhythmic activity comparable to the beating of the heart muscle of higher animals. There seems to be no constant difference in The Behavior of Loxophyllum 409 the rate of the back and forth movements between a piece contain- ing the sensitive anterior end of the body and a piece from any other region. Even very small pieces show the same rhythm. On account of its rhythmical activity Loxophyllum does not have to wait for something to turn up in order to acquire new experiences. Its life is one of continual trial. Only to a compar- atively slight extent is its activity, under usual circumstances, directed by external conditions at all. It goes forward, back, turns orally, goes forward and back again, and so on, repeatedly, through its own inherent activity. In many of the lower organ- isms behavior mainly consists in more or less direct responses to external stimuli with little spontaneous movement, but unless something unusual affects it Loxophyllum keeps circling about near the same place for a long time. When it meets with a strong or injurious stimulus it has its own methods of getting out of the way, but its ordinary behavior is mainly guided by internally initiated impulses. COMPLEXITY OF BEHAVIOR From the preceding account it 1s evident that the behavior of Loxophyllum is considerably more varied than that of Para- mecium and many other infusoria. Paramecium, for instance, has a yery few stereotyped modes of behavior, such as spiral swim- ming, the motor reflex, acceleration of forward motion when lightly stimulated at the posterior end, the thigmotactic response, and bending the body when crowded among obstacles. Loxo- phyllum has not only all these responses, but several others in addition, 7. e., gliding movements, regular changes in the form of the body accompanying forward and backward movements, small feeling movements of the tip of the body, undulations of the oral margin, twistings, turnings, and contortions of the body under various conditions, special movements involved in swallow- ing large objects, and several kinds of righting movements. ‘This greater complexity of behavior is probably a consequence of the fact that most of the creature’s life is spent in contact with solid objects. It appears to be a general rule that the behavior of the 410 S. EP Holmes free swimming infusoria is more simple than that of the creeping or the permanently attached forms. REGENERATION While studying the behavior of pieces of Loxophyllum I found that regenerative changes set in soon after the animal was cut in two. A good opportunity was thus afforded for watching the regeneration of the animal which takes place so rapidly that one might almost be said to actually see it going on. ‘To determine, so far as possible, the exact method followed in regeneration is always a matter of interest and importance; and a form in which Fig. 1 Showing the course of regeneration of a piece from the posterior end of Loxophyllum. The dotted line in this and the following figures indicates where the cut was made. the process can be watched under the microscope and followed step by step 1s especially favorable for this purpose. A Loxophyllum was cut in two near the middle by a slightly oblique cut (Fig. 1). In the anterior piece the sides near the cut end were drawn inward and soon met, thus closing in the cut portion of the margin and giving the piece much the appearance, except in its relatively greater width, of the normal animal. In the posterior piece to which attention was mainly directed, since much greater modifications were necessary to restore the normal form, the first step in the process of regeneration was the closing in of the sides at the anterior end. ‘The piece continued to swim forward and backward, undergoing the regular changes in form The Behavior of Loxophyllum ALI that accompany these movements which have been previously described. Soon, apparently as a result of these stretching out movements that accompany its swimming forward, the piece began to acquire a more narrow and elongated form. With each for- ward motion the sides would be pushed around more toward the middle of the cut end which gradually became reduced in extent. With each movement ahead it could be seen that not only the body elongated but that it elongated more on the oral than the aboral side, causing it to bend toward the aboral side at each advance. ‘This bend is not the result of the contraction of the aboral side, as one might very naturally suppose, but the exten- ¢ Fig. 2 Showing the regeneration of the posterior part of the body when cut off obliquely. sion of the oral side. Soon the oral side begins to grow longer than the aboral and to become pushed around the anterior end of the body. The striations which originally ran in a longitu- dinal direction are now bent around the anterior end of the body more on the oral side than on the aboral. ‘There is no formation of new tissue here, and no differentiation of new cilia on the cut surface, but the oral margin becomes stretched around the anterior end of the cut piece. Both sides of the body extend and contract, the movements being greater toward the anterior end. ‘his end becomes (in consequence of these movements?) more narrowed and more like that of the normal individual. ‘The cut end of the body is closed in by the gradual extension of the sides which fin- 412 Si 7: Holmes ally meet, the point of union being carried by the greater exten- sion of the oral side so that it finally comes to lie on the aboral side some distance behind the anterior end. The method of regeneration here followed in restoring the exter- nal form of the body is the simplest and most direct that can readily be imagined. The elaboration of new structures is re- duced to a minimum. ‘The part of the infusorian behaves much as an entire individual, narrowing the body as it advances and stretching the oral more than the aboral side; and this behavior seems to help mold the part into the final form. In order to find out how small a part of the differentiated oral margin could be stretched out to form the entire oral margin of the new individual a piece was cut obliquely across the body Fig. 3 Regeneration of small pieces from the anterior half. (Fig. 2), so that the oral side of the posterior portion was consider- ably shorter than the aboral. In this case the general method of regeneration was much as before. Both sides curve in to close the cut end, the piece elongates and becomes narrowed; the oral side in the movements of the animal is extended more than the aboral, and we can see that it is gradually stretched forward; and finally it is pushed around the anterior end. ‘The middle part of the cut surface which is becoming more and more reduced is apparently drawn back, but this appearance is due, I think, not to its being pulled back in the center, but to the extension of the two sides around it. he length of the ciliated oral side when regeneration is complete is considerably greater than at first. There was no extension of cilia in this case upon the cut surface. The anterior limit of the oral margin was very distinct and could «3 The Behavior of Loxophyllum 413 be followed in its course without any difficulty. The short oral margin of the posterior cut piece was simply stretched out to form the whole oral and anterior margins of the regenerated individual. Essentially the same method is followed in the regeneration of comparatively small transverse pieces (see Fig. 3). The experiment was then tried of reducing the oral margin still more. By making a cut across the anterior end and a longitudinal cut near the oral side the whole oral margin was removed except a small part near the posterior end of the body (see Fig. 3). In this case the amount of cut surface exposed was very much greater than in the previous experiments, so much so that it seemed incred- Fig. 4 Regeneration of a specimen from which the anterior end and most of the ciliated oral margin was cut off. ible that the small remaining part of the ciliated margin could be extended so as to stretch over it, especially since the part remain- ing is one that gets little stretching during the usual activities of the animal. As might be expected, although it was stretched around the oral side to a certain extent, this part failed to give rise to any but a small part of the new oral side. ‘The new oral margin with its differentiated structures had therefore to be pro- duced by anew method. ‘The piece began to elongate and become narrowed and rounded in front. Owing to the lack of the con- tractile and extensile elements of the oral margin the character- A414 S. : Holmes istic pushing ahead of the oral side did not occur. The aboral side in fact began to be pushed ahead of the oral which accounts for the form of the pieces shown in the figures. After several hours the oral margin became thinner and clearer and the gran- ules of the endoplasm came to lie further from the edge. A slight transverse striation could be detected in it such as occurs more plainly in the normal individual, and soon short cilia began to be put out here and there chiefly toward the posterior end. As the clear margin became broader it showed a longitudinal striation and soon began to extend and contract more during the move- ments of the body. As the oral margin slowly acquired its char- acteristic differentiation it began to push ahead and extend around the anterior end where its striations assumed the usual bend. Fig. 5 Regeneration of a piece from the middle of the body from which the oral margin was removed. In this case regeneration was very slow compared with the two preceding experiments. The piece was larger in size than the ~ others but more differentiation had to be accomplished. Not until the oral margin became furnished with its cilia and its differ- entiated contractile elements so that it was capable of performing its usual role in the movements of the animal was there any marked progress in molding the body into its final shape. ‘The anterior end, although it had become narrowed and rounded soon after the operation, did not take on any of its characteristic structural features until the oral margin became differentiated and began to be stretched around the front as in the cases of regeneration just described. The development of the new cilia extended gradually forward from the small part of the ciliated margin that remained and the The Behavior of Loxophyllum 415 possibility suggested itself that the new cilia which were devel- oped arose through the influence of the old ciliated margin or of material which were developed arose through the influence of the old ciliated margin or of material which might be derived from it. ‘To test this possibility a specimen was cut as is shown in Fig. 5 so as to leave no part of the cialiated oral margin remain- ing. ‘The general course of regeneration is indicated by the Figs. 2-6. It will be seen that the aboral side, as before, extends at first more than the oral, but after the oral margin becomes dif- ferentiated in its characteristic fashion it pushes around more than the aboral and produces the usual curvature at the anterior end of the body. The cilia made their appearance in scattered groups about fourteen hours after the cut was made. Fig. 6 Regeneration after removal of the dorsal half. When Loxophyllum is cut in two longitudinally the process of *® regeneration is comparatively slow. The usual form of the body may be approximately reached in a comparatively short time, but the differentiation of the structures characteristic of either margin requires several hours. Fig. 6 represents a specimen from which the aboral half was removed by a longitudinal cut. The posterior end of the first became bent aborally and was brought forward so that the two parts of the cut margin met and fused together. ‘The body as a whole shortened and widened; the injuries that were incidentally made near the anterior end of the body were repaired, and while the cut margin so far as could be ascertained seemed to close by the approximation of the upper and lower edges it was over twelve hours before the groups of trichocysts character- istic of the aboral margin made their appearance. 416 S. Ff. Holmes In a specimen cut longitudinally part way through the body (Fig. 7), regulation was effected by the meeting and fusion of the cut surfaces. [he movements of this specimen were of interest. As it swam forward the two sides became crossed. During back- ward swimming on the other hand, they diverged very widely. This is doubtless due to the fact that in the lengthening and short- ening that respectively accompany the forward and backward movements of the organism the marginal regions of the body are more active than the middle. The mechanism of the extension of the sides I have not ascertained. THE ROLE OF MOVEMENTS IN REGENERATION The foregoing experiments make it probable, as Child has attempted to show in other forms, that the role of the movements / Fig. 7 Regulation in a specimen cut longitudinally as shown in 7. In 2is shown the shape assumed when the animal is swimming forward; 37 shows the form during backward swimming. of the organism in bringing about the normal form of the body» is an important one. ‘There are, in fact, few cases in which the efhicacy of the factor of movements seems more manifest. To a considerable extent at least the organism seems to pull itself into shape. It has certain ways of acting which, as observations on the behavior of the parts have shown, are characteristic of the behavior of even quite small parts in much the same way as they are of the whole. A small piece cut from almost any region of the body shows the same rhythm of back and forth movements, the same correlation of extension with forward movement and of con- traction with backward movement, and to a certain extent the same oral and aboral bendings as the entire animal. And when one carefully follows the course of regeneration it seems evident The Behavior of Loxophyllum 417 that these movements are gradually working the part into the form of the whole. In the regeneration of the posterior half, for instance, one may see the oral margin extending and extending, growing a little longer with successive stretchings, until it curves about the anterior end of the body, and its striations are bent around so as to give the characteristic appearance of that region of the normal animal. The same kind of action is apparently instrumental in producing the same kind of form. But precisely what is the relation of the movements of the organ- ism to its regeneration does not, however, lie on the surface. It seems evident that the movements have an important part in shaping the general outline of the body. But are they the funda- mental causes of this change of shape, or agencies which merely assist or accelerate the action of other formative factors? ‘The experiments performed, while they indicate the importance of behavior in regeneration, show, I believe, that this factor is of a secondary or subordinate nature. It will be instructive to con- sider the course of regeneration in those experiments in which most or all of the oral margin was removed. Here regeneration was forced to follow a very different method from that adopted in the cases first described where a part of the oral margin was stretched out into the whole. “The new margin had to be formed entirely de novo. ‘There were involved the thinning out and clearing up of the oral side, the differentiation of new contractile threads, new trichocysts, new cilia, a complicated ordering of newly differentiating structures. The gross movements of the body could have had very little to do with all this. Until these differentiations were made the movements of this side of the body were not of the usual kind. Commonly the oral side extends and contracts more than the aboral, but when the marginal elements of this side were removed the opposite side was the more active. The oral side did not extend so rapidly as the aboral until the structures characteristic of the oral margin were established. If in the experiments first described the general form of the body seemed to be produced by the characteristic behavior of the animal, the characteristic behavior 1n this case had to wait until its structural basis was established by comparatively slow differ- 418 S. Ff. Holmes entiations of new parts. If form seems in some cases to be molded by function, function in turn is apparently the result of organ- ization. [he modifications of form and function, of course, go on pari passu, and are after all but different aspects of the same process. ‘he gross activities of the organism are largely depend- ent on the finer organization of the animal since they are carried on in a very similar way, even by comparatively small pieces of the body. Where, as in the experiment cited, the animal is cut in such a way as to modify certain of its grosser movements the finer differentiations go on until a structure is produced which then undergoes the movements characteristic of the whole when the external shape is rapidly assumed. The processes of building up the finer structures of the body, the formation of new myonemes, cilia, etc., are really the fundamental features of regeneration. Pulling the body into shape is a sort of secondary matter in which the gross movements play an important part, to be sure, but these are themselves dependent upon the finer differentiations. In certain cases among the infusoria, such as some of the Hypotricha, the comparative rigidity of the body excludes the factor of movement from playing a very important role in shaping the outlines of the regenerating organism: Yet these forms regenerate with great readiness. Where the factor of movement is of importance in the regeneration of the infusoria, it is, | believe, rather in the nature of an aid to other formative factors than an essential and fundamental factor itself. Zoological Laboratory University of Wisconsin REGENERATION AS FUNCTIONAL ADJUSTMENT BY S. J. HOLMES Witu One Ficure In a previous paper' | have ventured to outline a general theory of form regulation, based on the conception of an essentially symbiotic relation between the parts of an organism. ‘The con- ception 1s, of course, nothing new, but, so far as [ am aware, no one has hitherto attempted to deduce from it a theory of regener- ation and other processes of a regulatory nature. The theory may be stated in brief as follows: The various parts of an organism are supposed to stand in such a relation to each other that each part derives some advantage or is helped to perform its normal functions through the materials and stimuli it receives from other and especially the contiguous parts of the organism. Each part in turn contributes something to the normal functioning of the parts surrouningit;the relation is one of mutual dependence. Being mutually dependent, the parts of an organism tend to settle into a condition of functional equilibrium. When a part of the organism is removed and tissue of an undifferentiated nature 1s produced in its place, this new tissue develops in the direction of the missing part because this line of development is favored through the influence of the surrounding parts. Whatever advan- tages accrued to the part formerly in this position from its relations to the parts around it would also accrue to this tissue in so far as it differentiates in the same way as the part removed. ‘The new tissue differentiates according to the functional demands upon it and its line of specialization may be regarded as a case of func- tional hypertrophy. Regeneration of the missing parts, therefore, may be interpreted as an expression and act of getting back into a condition of functional balance. 1 Archiv fiir Entwickelungsmechanick, xvii, Bd., p. 265, 1904. Tue JourNAL or EXPERIMENTAL ZOOLOGY, VOL. IV, NO. 3. 420 S. Ff. Holmes The process might be illustrated by the case of a social organism composed of animal cells and symbiotic alga which I described in my former paper. “‘We may suppose that both animal and plant cells tend to grow and multiply as far as circumstances permit. As these cells depend upon each other to a certain extent, neither kind of cell will tend to preponderate over the other, but they will all adjust themselves to a condition of approximate equilibrium. Now suppose that a considerable number of the alge of this com- posite organism be removed. ‘There is a functional demand by the rest of the organism for the products of the algae and an excess food supply for those which remain. The alga, therefore, are supplied with exceptionally favorable conditions for growth and multiplication, and will be stimulated to regenerate their missing number. By supplying the functional demand of the animal cells they indirectly benefit themselves, because by producing more oxygen they enable the animal cells to produce more of the sub- stances which they utilize as food. If we suppose that in our hypothetical organism there are, in addition to the two kinds of cell mentioned, indifferent cells which are able to develop into either animal cells or alge, it seems probable that, in the event of the removal of the alga, the indifferent cells will differentiate so as to take the place of the missing numbetss © “* * “For the sake of a simple illustration we have described an organism consisting of but two kinds of cells, but there is no reason to doubt that in a complex organism consisting of many varieties of cells standing in a symbiotic relation there would be a similar regeneration of any part that is removed. Let us imagine an or- ganism made up of a number of differentiated cells, each of which derives some advantage from some substances produced by the contiguous cells, and giving out some substance upon which the contiguous cells are more or less dependent. We will suppose that, in addition to these differentiated cells, there are scattered through the body numerous indifferent or embryonic cells whose multiplication is held in check by the others, but which upon the removal of any part respond to the functional disturbance by growth and multiplication near the place of mutilation. We may represent our hypothetical organism graphically by the following Regeneration as Functional Adjustment 421 diagram in which the differentiated cells are represented by the larger circles 4, B, C, etc., and the indifferent cells by the smaller circles between them. Each cell such as 4 contributes something utilized by B, G and F, and derives something in return from each of these sources. Now suppose 4 is removed; the indifferent cell lying near by,.no longer held in check by the same stimuli, begins to grow and develop. What line of differentiation will it most naturally take. Owing to the symbiotic relation subsisting be- tween the cells differentiation in the direction of 4 will be most favored as this secures it the advantages which 4 received. In other words, this will be the direction of development along which social pressure will tend to guide it. And the result will be a re- ErGaet generation of the missing part.” For applications of this theory here set forth in barest outline to morphallaxis, heteromorphosis, physiological regeneration, and other modes of regulation, refer- ence may be made to my former paper. In some recent articles the problem of regulation has been ap- proached from points of view somewhat similar to my own. — Jen- nings” has attempted to show that regulation in behavior is funda- mentally similar to other forms of regulation. ‘The method of trial and error, which is so pronounced a feature of the behavior of lower organisms, and one through which they secure a large part of their adaptations to external conditions, is assumed by Jennings to be followed in the various processes occurring in ? This Journal, vol. 1i, and Behavior of the Lower Organisms. New York, 106. 422 Sia Fa 21 olmes the regulation of organic form. “A disturbance of the physio- logical processes,’ he says, “results in varied growth activities. Some of these will relieve the disturbance; the variations then cease and the processes are continued.” The result of the selec- tion of those growth processes which relieve the disturbance—or we might say make for functional equilibration—1is the restoration of the lost part. Jennings has not attempted to develop a theory of form regulation in detail farther than to show the fundamental similarity a the method of regulation in various fields, but he holds that my own point of view so far as form regulation 1s con- cerned is in “essential agreement” with his. Child® has recently outlined a theory of regulation which, as he states, is “somewhat similar to that adopted by” myself, although differing in certain important particulars. According to both Child and myself, regeneration and other formative processes are the result of functional activity, or more specifically, func- tional equilibration. ‘Tissue differentiates in the direction of the missing part because it takes on the functional activity of the miss- ing part. To cite an illustration by Child, “after removal of the anterior or posterior end in Bipalium or Planaria maculata the terminal regions of the piece remaining are subjected to condi- tions somewhat similar to those existing in the terminal regions of the part removed. ‘The anterior end of the headless piece of Planaria is subjected to external conditions more or less similar to those to which the old head was subjected; moreover, its rela- tion to the other parts of the body is more or less like that of the head. Stimuli resulting from forward movement affect it first, and are transmitted from it to other parts, etc. Functionally speaking, it serves in some degree as a head. ‘The case is similar as regards the posterior end. After removal of the original pos- terior end, the posterior region of the piece functions in some degree as a posterior end, or to put the matter more strictly, its functional relations with other parts are more or less similar to these of a¥ tail” or postenar end, -= ‘“Redifferentiation occurs as a result of a functional substi- ’ Archiy fiir Entwickelungsmechanick, xx Band, 1906; and this Journal, vol. ii, 1906. Regeneration as Functional Adjustment 42.3 tution of a larger or smaller part of the old tissues of the piece for the part removed; the substitution may be imperfect or incom- plete at first, and gradually attain completeness. In consequence of this functional substitution, the structure of the part involved is altered until it comes to resemble more or less closely that of the part removed.” Where regeneration through the formation and differentiation of new tissue occurs, it is this tissue which becomes the func- tional representative of the old part. In the regeneration, for instance, of the arm of a star-fish, or the leg of an arthropod or amphibian, the new tissue ““must be subjected to many condi- tions—internal and sometimes external—similar in a greater or less degree to those to which the part removed or some portion of it was subjected.” The factor of the exercise of a part which Child has regarded in many cases of so much importance, is not always necessary for regeneration. ‘‘ The growth of the new leg is not the result of the attempt to use the leg which is missing. The growing tissue begins to develop into a leg because its relations to the other parts of the system are in some degree similar to those of the leg removed. As it grows, the conditions approach much more and more nearly those to which the normal leg 1s subjected, 7.2., there is a gradual return of the functional conditions to the normal.” The application of Child’s theory to the subject of polarity, heteromorphosis, and many other regulatory phenomena, it will not be necessary for our present purpose, to discuss. The funda- mental idea of the theory is that form regulation is a result of func- tional regulation. In certain cases Child attempts to show that this supposed functional relation actually occurs as a consequence of mechanical conditions. The anterior end of a planarian gets stimulated by the water and by the: impact from foreign bdics much like the head does. The posterior end of Schatten is used by the animal in locomotion much as the tail is. At first the external stimuli affecting the ends are much the same whether they are anterior or posterior. In the movements of the animal both re- ceive frequent impact from contact with various objects. ‘There 424 Sif. Llolmes is a difference perhaps in the mechanical stimuli received, but granting that these start the course of differentiation in different directions, they are entirely inadequate to account for the whole process of differentiation, as Child himself would probably admit. Where the factor of movement is absent, Child has recourse to the supposition of some other form of functional substitution, but he gives no clear account of why the substitution should occur. To say that the end of an arthropod’s appendage is regenerated because of the functional activities that occur within it, that mor- phallaxis occurs when the part readily takes on the function of the whole, and that regeneration takes place because the functions of the missing part are imposed upon the new tissue that is developed in its place, may all be very true so far as it goes, but until some principle for the explanation of this functional adyjust- ment is brought forward the explanation of regeneration 1s far from complete. If form regulation is a consequence of functional regulation, as Child and [| agree that it is, the interpretation of functional regulation is the next obvious step. The inadequacy of Child’s theory is, that it does not contain any general principle of explanation for that functional substitution and equilibration upon which it is assumed that form regulation depends. ‘This, however, is a matter of incompleteness rather than error. But I suspect that when his theory comes to be developed so as to sup- ply this missing element, it will involve, to make it workable, the assumption of some such symbiotic relation between the parts of an organism as I have assumed. It is a strong point in favor of the theory of symbiosis that it affords to a certain degree an expla- nation of physiological adjustment; in fact, it 1s primarily a the- ory of physiological equilibration. ‘This physiological adjustment brought about through the symbiotic relations of the parts may, as | have attempted to show, be explained, or at least much of it may be explained, as the outcome of a tendency toward chemical equilibration. ‘To the extent that this is true, we have an explana- tion of the regulatory activities of an organism in terms of famil- iar chemical phenomena. The conception of something like a symbiotic relation between the parts of an organism which is involved in my own theory eo Regeneration as Functional Adjustment 425 of regulation, Child rejects, but I think on insufhcient grounds. I have assumed that, according to the symbiotic relation of the parts of an organism, upon removal of a part, such as 4, in the figure, the undifferentiated tissue in the region of 4 will differen- tiate in the direction of the missing part because of the functional demands, or for what, for want of a better term, I have called social pressure, upon that tissue. “‘This,’’ according to Child, in refer- ring to the particular case illustrated by the diagram, “‘is exactly what will not occur under these conditions. If all the cells 4—F are symbiotically correlated then removal of one of them, 4, must affect all the others, 7. ¢., the whole complex is altered by removal of one of its members. It is perfectly clear that the ‘social pres- sure’ of the altered complex will not be in the direction of differ- entiation of the indifferent cell into something like 4 but in some other direction, in other words, the indifferent cell cannot replace A but will form something different. Moreover, since all the cells were dependent upon 4 in some degree, the removal of 4 will probably render continued existence impossible for some of them and their place will be taken by the undifferentiated cells, but these will also develop into something different because the ‘social pressure’ is altered. It is perfectly evident that no regulation in the sense of replacement of a missing part could occur in sucha complex.” Now this conclusion may be perfectly clear to Dr. Child, but I must confess—perhaps I am blinded by my bias in this matter— that it is far from being so to me. According to Child, since the removal of 4 would alter B, G, F, etc., not only something differ- ent would be developed in place of 4, but the whole complex, according to my theory, would be profoundly altered. Now, I admit that the removal of 4 tends to alter B, Gand F, etc. How far this tendency will result in a modification of these cells depends on the plasticity of the organism and the degree of mutual depend- ence of the parts—factors of course which vary in different organ- isms. But Child overlooks the fact that according to the sym- biotic relation assumed, the other cells C, D, E, etc., tend to keep B, F, G in their original condition. In so far as these remain in their original state, their influence on the indifferent tissues in the 426 B ae Holmes region of 4 will tend to mold it in the direction of the missing parts. In so far as B, G and F are modified through the loss of the missing part, their influence on the tissue in the region of 4 will come to be modified, and they will, in turn, modify the cells lying next to them. But, as there is a tendency for the modifica- tion produced by the loss of 4, to spread successively to other parts, there is also:a tendency, according to my theory, toward the checking and reversal of this process. If the loss of 4 tends to modify B, F and G, the presence of EF, C and D tends to hold them in place, and in so far as these are maintained through this influence they tend to mold the tissue in the position of 4 into the form of the missing part; and in so far as this 1s so molded its modifying influence on B, F and G is diminished. How the process works out depends naturally on the degree of specification of the parts, whether or not new tissue is formed in the place of the missing part, and perhaps other factors. If the organism is plastic and its parts have not acquired an irretrievable set which prevents further modification, it may be entirely worked over in consequence of the disturbance of its social pressure in the vicinity of the missing part, thus leading to redifferentiation, or morphallaxis. Whether we have morphallaxis or regeneration in a narrower sense may depend, among other things, upon the degree of specification of the parts. As I have suggested in my for- mer paper (p. 288), and as Child has maintained more at length, regeneration as opposed to “redifferentiation, increases as func- tional specification of the tissues increases or, in other words, the greater the degree of differentiation—the visible result of functional specification—the less likely is extensive functional substitution and consequent redifferentiation. ” This, it seems to me, is very much what one might expect ac- cording to the theory of regeneration | have outlined. Replace- ment of 4, according to Child, “can occur only when the relation is largely one-sided, 7.e., when 4 is dependent on B—-F, but these latter are not to any marked degree dependent on 4. In this case, and in this case only, will the social pressure force the undiffer- entiated cell to differentiate into something like 4.” Where redifferentiation from new tissue is concerned, as in the present = Regeneration as Functional Adjustment 427 case, it is not the relation of 4 to B—F, that should be more or less one-sided, but the relation of the tissue in place of 4 to this complex. This is an important distinction which Child does not seem to have considered. -—F are relatively fixed, the tissue in place of 4 is young and plastic, and more dependent so far as the direction of its differentiation is concerned, upon B—F, than these are upon it. We may grant that, when regeneration occurs, the relation of depend- ence between the old parts and the new tissue is more or less one- sided, although the relations of the part removed may not have been. ‘This would naturally result if the parts were relatively stable. ‘They may be in a symbiotic relation, nevertheless, each part contributing in some way to the normal functioning of the others, and dependent to the extent that the removal of one part may alter only to a certain degree the quality and quantity of the activity of the surrounding parts, without producing extensive modifications of structure or function. If the parts B—F were more plastic, absence of 4 would natur- ally tend to cause greater changes in them, especially if new tissue were not produced in place of 4, which would come to assume some of the missing functions before the modification extended very far. ‘There would be a progressive modification extending from the region of 4, which would tend to become less the farther it extended, but eventually perhaps affecting more or less the entire organism. Functional equilibrium would then be maintained by working over the organism so that all the parts were adjusted to functioning on a smaller scale. ‘The different methods of regu- lation, through morphallaxis, regeneration and the various com- binations/of these processes are, | believe, interpretable according to the symbiotic theory, and the relations of regeneration and morphallaxis to the degree of specialization of the parts which Child has elaborated, are, in fact, exactly what the theory would lead us to expect. The difficulty pointed out by Child that the process of differen- tiation of a new part seems to begin at the tip and work back to the base is one which gave me some concern when developing my theory, but I think the difficulty is by no means a fatal one. When a developing limb shows first those structures character- 428 S. Ff. Holmes istic of its distal end we should bear in mind the possibility that the differentiation which first appears is not necessarily that which first occurs. What we know of developmental processes renders it very probable that a great deal of differentiation is going on in the rudiment of the limb before it is manifested by any external signs. Between maturation and cleavage an ovum may show no outward sign of differentiation, but experiment shows that this period is one in which developmental processes are rapidly tak- ing place. Before any external features are produced in the devel- opment of a limb, the main outlines of its differentiation may have been established through influences proceeding from its basal part, after which the tip might differentiate more rapidly than the intervening portion, and the other visible features of structure appear successively toward the base. [| do not suggest this merely to save my contention by a retreat into the invisible, but there are certain considerations that make such an interpretation more or less probable. In many cases the visible differentiation is cen- trifugal rather than centripetal. ‘The tail of a tadpole cannot be said to differentiate from the tip toward the body as it regenerates. Zeleny has shown that in the early regeneration, as in the embry- onic development of the antennz of Mancasellus the formation of segments proceeds at first from the base to the tip; later new seg- ments are formed in the reverse direction. But granting that, in many cases, differentiation actually begins at the extremity and works toward the base of the regenerating organ, the process is not inconsistent with the point of view here set forth. We may suppose that the influence of the environment causes the extremity of an organ to begin to differentiate like that of the missing part. ‘That is only one step. We have then to account for the numerous coordinated differentiations that take place as the part develops toward the base. In my illustrations of the course of differentiation under the guidance of social pres- sure, I have taken the old part as a starting point, but if we have an undifferentiated mass of cells, it is conceivable that, if, for any reason, differentiation should start at the distal extremity of the mass, 1t might work back under the guidance of social pressure toward the base. The distal differentiation would have to get Regeneration as Functional Adjustment 429 started in the right direction, or something else than the missing organ would be produced. ‘That cases of heteromorphosis some- times occur might be interpreted as the result of such failures. But the comparative rareness of heteromorphosis makes me sus- pect that the beginnings of visible differentiation that frequently appear at the tips of regenerating organs do not occur without any relation to the basal part. “The fact that, with few exceptions, such as the failure to regenerate the intermediate segments of the appendages, etc., the whole organ, nothing more nor less, 1s regen- erated, and forms a congruent union with the basal part, is indica- tive of close interaction of the various parts of developing organ with the body of the organism at all stages of the process. I am inclined to think that neither centrifugal nor centripetal differentiation, expresses the entire truth of the matter, but that the new part differentiates as awhole, much as organs do in embry- onic development, and at all times in intimate functional relations with the old part, differentiation becoming accelerated in one part or another, according to special conditions. If differentiation began at the tip of the rudiment of an organ, and proceeded cen- trally, the whole might be differentiated before the body was reached, leaving a mass of unused tissue between; or differentia- tion might reach the body before all the immediate parts were produced. If differentiation proceeded in the reverse direction, similar imperfections might arise. We must look upon a regen- erating mass of tissue as one in which incipient developmental tendencies are proceeding in various ways, modifying each other, and gradually working into a condition of physiological equilib- rium with the basal part and with the environment before much outward evidence of differentiation makes its appearance. It is probable that the main elements of a regenerating appendage of an arthropod, for instance, are blocked out before any external marks become visible. Even during the early stages of prolifera- tion of the cells of the regenerating appendages, it 1s not improb- able that incipient differentiations are becoming established. And the basal part notwithstanding the fact that the visible differen- tiation may take place in a centripetal direction, may exercise a guiding influence at all times over the regeneration of the part, 430 Si vis Holmes and determine that it forms in harmony with the rest of the organ- ism. Such a conception is entirely congruous with the symbiotic theory, and is, I believe, consistent with the various observed facts of regeneration. If we explain form regulation as an outcome of functional regu- lation, we make little progress until we have some interpretation of the latter process, and any theory of form regulation which offers nothing in this ciecuon makes no more than the first step toward an explanation of the phenomenon. In his criticism of the theory of form regulation which I have outlined, Child has advanced arguments which are, I believe, by no means fatal to it, and he has not brought forward any other explanation of func- tional equilibration, which both of us regard as the basis of form regulation. Perhaps this may be supplied in further developments of his theory which Child hints are to be made in the future. While functional adaptation may occur independently of any sym- biotic relations, especially in the direct adaptation of parts to exter- nal conditions, the mutual adaptation of parts which forms so important an element in formative processes are, | believe, for the most part, dependent on symbiotic relations. At present [ am unable to see how any general explanation of functional equili- bration among the parts of an organism can be reached unless we assume that the parts are, to a considerable degree, interdependent. Perhaps some other interpretation of functional regulation may be advanced which does not make use of this idea. “That remains, of course, to be seen. But the theory of the symbiotic relation of the parts of an organism has the merit of enabling us to interpret form regulation and functional regulation as the outcome of ordin- ary physiological activities, and hence to give, in a measure, a causal explanation of the teleological behavior which is manifested in so striking a degree by formative processes, and which forms the strongest support of some recent vitalistic theories. So far, at least, I hope it is in the line of progress. Zodlogical Laboratory University of Wisconsin From the Wistar Institute of Anatomy and Biology, Philadelphia SOME FACTORS IN THE DEVELOPMENT OF THE AMPHIBIAN EAR VESICLE AND FURTHER EXPERI- MENTS ON EQUILIBRATION BY GHORG EE so) RE EASE Re MeD: Associate Professor of Neurology at the Wistar Institute Witu Srx Ficures In a previous paper concerning experiments on the developing ear vesicle! it was shown that the group of cells forming the primi- tive epithelial ear cup or ear vesicle of the tadpole is specialized to that degree that although removed to an abnormal environ- ment the cells still continue to differentiate themselves into a struc- ture possessing many of the features of a normal labyrinth. Re- cently it has been shown by Lewis’ that even earlier, while still an uninvaginated plate, the ear anlage is already capable of a cer- tain degree of independent differentiation. In the following paper additional evidence will be given of the high degree of develop- mental independence possessed by the early labyrinth cells. It will be pointed out that individual parts of the vesicle may develop independently of the rest of the vesicle. It will also be shown that the process of differentiation extends to the difference existing between a right and left-sided organ. A left ear vesicle trans- planted into the empty pocket left by the removal of the right ear vesicle develops into a labyrinth that is perfect in general form and in its relations to the brain, with the exception that it main- tains its left-sided character; the anterior semicircular canal is found on the caudal side toward the vagus group, while the pos- terior canal lies toward the eye, and likewise the lagena which 1 Streeter, G. L., 06: Some experiments on the developing ear vesicle of the tadpole with relation to equilibration. Jour. of Experimental Zodl., vol. in. 2 Lewis, W. H., ’07: On the origin and differentiation of the otic vesicle in amphibian embryos. Anatomical Record, No. 6, Amer. Jour. of Anat., vol. vii. Tue JouRNAL oF EXPERIMENTAL ZOOLOGY, VOL. IV, NO. 3. 432 George L. Streeter, M.D. normally buds out from the caudal border of the saccule in these cases is found extending forward toward the proodtic ganglion. The ear vesicle, however, is not in all respects independent of the surrounding structures. Some experiments which are reported below, indicate that its position in reference to the brain, ganglion masses and the surface of the body is determined by the environ- ment itself; it may be rotated in any direction, and nevertheless it eventually develops in the normal attitude, with the saccule toward the ventral surface, the semicircular canals toward the dorsal surface, the lateral semicircular canal being toward the lateral surface, and the endolymphatic appendage toward the brain. The experiments were carried out on larve of Rana sylvatica and Rana pipiens, and the operating stage was the same that was used in previous experiments.* ‘The time is just at the close of the non-motile stage, and the epithelial ear consists of an invag- inated cup-shaped mass of cells just in the process of being pinched off from the deeper layer of the skin, with the edges turning in to form a closed vesicle. For simplicity the term “ear vesicle’ will be used even though the closure is not yet complete; the attempt to distinguish between auditory cup and auditory vesicle does not seem to be justified for the present purposes. ‘The technique of the operations was also the same as that described in the previous paper. Notes were made on the behavior of the animals, and at the end of from four to six weeks the specimens were preserved in a chrome-acetic mixture, cut 1n serial sections, and stained with hamatoxylin and congo red. With certain specimens the ear vesicle, adjacent ganglia, and a portion of the central nervous sys- tem were reconstructed after the Born wax plate method. Eleven such models were made, and photographs of some of them are reproduced in Figs. 2, 3 and 6. With the aid of these models it was possible to identify relations and detailed features of the laby- rinths that otherwise could not have been recognized. The morphological features of the experiments will be first con- sidered, and the behavior of the animals and its relation to equilib- rium will be treated separately in the latter part of the paper. +)StreeterOOs) w.rGe tig agenp iG 47.= i x“ Development of Amphibian Ear Vesicle 433 DETERMINATION OF POSITION OF THE EAR VESICLE The conclusion that the attitude of the developed labyrinth, the position of its canals and various chambers, is determined by its environment is based on seventeen experiments in which the ear vesicle was loosened from its normal situation and placed in an abnormal attitude, and the specimen then allowed to continue in its development. At the end of a month examination showed that the labyrinth had become differentiated with varying degrees of completeness, and in each instance had developed in normal rela- tion to the surrounding structures. Rotation in Two Directions. In eight of these experiments the ear vesicle was rotated 180° around both its vertical and transverse axes, so that it was turned face inward and upside down; or, inother words, its lateral or invaginated surface was toward the brain and its ventral border was where the dorsal border should be, the max1- mum displacement. After this procedure the wounds healed within a few hours, and the larva were reared up to the fourth or hfth week, when they were killed and cut in serial sections. ‘The labyrinths of fve specimens were reconstructed. Before describ- ing them reference should be made to the normal condition of the labyrinth at this age. A reconstruction of a normal one with its adjacent structures is shown in Fig. 1. From the reconstruction of a normal specimen it can be seen that the three semicircular canals have individual characteristics by which they can be separately identified; such as the Y-shaped union of the anterior and lateral canals, and the overlapping of the caudal end of the lateral canal by the posterior canal, and the junction of the posterior and anterior canals to form the crus com- mune. ‘The differentiation between utricle and saccule is not yet complete, but the part that 1s to become saccule is so labeled. From the caudal border of the saccule can be seen a small pocket budding out which constitutes the lagena or primitive cochlea. Directly median to the crus commune ts the endolymphatic append- age, consisting of a small duct leading from the main labyrinth chamber up between the labyrinth and brain to a rounded pouch, the saccus endolymphaticus. In their histology, as well as in 434 George L. Streeter, M.D. their general form, the various parts of the labyrinth exhibit at this time individuality. (See Fig. 4.) “The ventro-median portion of the vestibular sac and the ampullar ends of the semicircular canals possess high columnar cells forming the neuro-epithelial macula which are supplied with fibers from the acoustic ganglion, lying against the medial wall of the labyrinth. ‘The endolymphatic sac has cuboidal cells, and the lagena has intensely staining col- umnar cells like those seen in the macular regions. ‘The lagena is further characterized by its sharply rounded outline, and by the fact of its being compactly surrounded by ganglion cells and fibers, and cartilage forming cells. These features are so definite that Crus commune C ose. anterior Sac.endol.., C.sc. posterior \ Rhomben nN ele : ‘Gang. prootic. ‘SSacculus C.sc. lateralis 4 Lagena ’ Fig. 1 Reconstruction showing the form and relations of the membranous labyrinth of a normal tadpole (Rana pipiens) one month old. The labyrinth, adjacent ganglia and part of the brain wire reconstructed after the Born method, and the remainder of the figure was drawn from a dissec- tion of a tadpole of the same age. Enlarged 35 diameters. the various parts of the labyrinth can be recognized without difh- culty, even though they happen to be incomplete, or out of their normal relations. Now if one examines the models of the operated specimens, photographs of three of which are reproduced in Fig. 2, it 1s seen that the individuality of the semicircular canals can at once be identified. In model a, the canals are practically normal; in model b, the anterior canal is small, and the lateral canal consists only of a pouch which has not been pinched off from the main cavity; in model c, the posterior canal remains a simple pouch, while the Development of Amphibian Ear Vesicle 435 anterior and lateral canals arenormal. In considering the posture of the canals it is to be noted that the surrounding structures have been left out in Fig. 2, to avoid unnecessary duplication; the three models are all represented in the same relative position as that of the labyrinth in Fig. 1, 7. ¢., the cephalic end is on the right, the caudal end is on the left, the ventral surface 1s below, and the dor- sal surface is above. ‘Thus it will be seen that the lateral canals in all three models are in the same plane; likewise the posterior canals all form the dorso-caudal border of the labyrinth, and the anterior canals form the dorso-cephalic border. ‘The fact that the anterior canal is small in model b,‘ and the posterior canal is small in model c, gives rise to a false impression of a backward sac. endolymph. lagena a ya b C Fig.2 Reconstructions showing the form and posture developed by three labyrinths one month old» which while primitive ear vesicles were rotated from their normal position so as to lie face inward and upside down. The models are placed so that their planes are parallel with those in Fig. 1. Thus they present a lateral view with the cephalic end toward the right, caudal end toward the left, dorsal surface above, and ventral surface below. Enlarged 50 diameters. and forward tilting of the vesicle. “The saccule and lagena have the same position as in Fig. 1, and the lagena points caudally as it should do. ‘lhe endolymphatic appendage lies on the median side of the crus commune; in models b and c it is small, but the tip 4 This may be due to injury received at the time of operation. Such localized defects are frequently seen. They may involve any part of the labyrinth, and they vary greatly in the extent of the labyrinth wall affected. In one case the entire labyrinth was defective, with the exception of the endolymphatic appendage, which was normal in structure and position, and presented a curious appearance, being attached to the small irregular vesicle representing the labyrinth. Such localization of abnormal development is evidence of the high degree of specialization of the cells forming the primitive ear vesicle. 436 George L. Streeter, M.D. of it can be seen in model a. The acoustic nerve and ganglion are attached to the median and ventral surfaces of the labyrinth, and the nerve connection with the brain appears to be normal. The conditions found in the three specimens pictured in Fig. 2 are typical of what is found in the other five specimens examined. They vary in the completeness of their differentiation, some of them consisting of only a vesicle with perhaps a single canal pouch, but in all cases the acoustic ganglion is present on the ventro medial surface, and the macular areas can be recognized. The lagena is present in seven out of eight cases. “The endolymphatic appendage developed in six out of eight cases. As regards posture, the rule is that the more perfectly the labyrinth is developed the more accurately its posture corresponds to the normal relations. But even in the most imperfect specimens when the endolym- phatic appendage appears it 1s on the medial surface, and the ten- dency to canal formation is always on the dorso-lateral surface, and the saccule and lagena appear on the ventral surface. “This condition of course applies only to vesicles that have been im- planted in the acoustic region as was done in all the above cases. Rotation in One Dtrection. In four experiments the ear vesi- cle was rotated 180° around its vertical axis, 7.e, turned face inward. These specimens were then reared as in the preceding instance, and eventually cut in serial sections. A reconstruction model of one of them is reproduced in Fig. 3, and if it 1s compared with Fig. 1 it will be seen that although the vesicle was started in its development with invaginated side toward the brain yet the com- pleted labyrinth has the normal posture. A section of the same specimen is reproduced in Fig. 4, showing the labyrinth surrounded by developing cartilage. The acoustic ganglion is connected in normal manner with the brain and sends peripheral fibers to the thickened floor of the saccule. “The endolymphatic sac is in its normal position, and the narrow duct can be seen connecting it with the main chamber of the labyrinth directly median to the crus commune. ‘The series through this specimen show that his- tologically it is practically perfect. Of the other three specimens one was almost equally perfect, another showed some abnor- malities in the formation of the canals and the lagena, and the - Development of Amphibian Ear Vesicle 437 fourth was quite imperfect, consisting of only a large vesicle with a thickened epithelial floor connected by a few nerve cells and fibers with the brain. Fig. 3 Reconstruction of a tadpole labyrinth one month old, which when a primitive ear vesicle was rotated from the normal position 180° in one direction, so as to lie with invaginated side toward the brain. A section through the same labyrinth is shown in Fig. 4. Enlarged 55 diameters. * : : A. 26 “Gang. acus! * Fig. 4 Section through the membranous labyrinth shown in Fig. 3. turned face inward it has developed in the normal attitude. It shows that though originally e, endolymphatic appendage; c.c., crus com- mune}; sacc., saccule; c. Jat., lateral semicircular canal; gang. acust., acoustic ganglion. Enlarged 55 diameters. Transplanted Specimens. ‘Vhe irregularity of form of the six specimens transplanted to the region between the eye and nostril, previously reported,’ is so great that they give no assistance in solv- Streeter 706, /. c., p. 557. 438 George L. Streeter, M.D. ing the question of posture. However, in five cases, which will be presently described, where the ear vesicle was transplanted from the left side to the right side into the place made vacant by the removal of the right ear vesicle, in spite of the fact that these ear vesicles were implanted with haphazard attitude toward the adjacent structures, they nevertheless in each instance developed right-side up, and with the median surface toward the brain, as can be seen in Figs. 5 and 6. Mesenceph. Sac.endolymph } Dienceph. 1 ! ‘ ' 1 ' ' ' Khombenceph. Med.spinalis -angprootics. © na Lagena Fig. 5 Reconstruction showing the form and relations developed by a left ear vesicle when trans- planted to the right side; it shows that under such circumstances the ear vesicle retains its left-sided characteristics, though it otherwise normally adapts itself to its new situation. A photograph of the same specimen is shown in Fig. 6, c. DETERMINATION OF THE DEXTRAL AND SINISTRAL CHARACTER OF THE EAR VESICLE The question as to whether the right or left-sidedness of the ear labyrinth is controlled by the environment, or is determined by some intrinsic character of its own constituent cells, is answered in favor of the latter by the fact that if the left primitive ear vesicle, before the time of its complete closure, is transplanted to the oppo- site side of the embryo it retains its original left-sidedness. In five specimens, at the usual operating stage, the right ear vesicle was removed, and at the same time the left ear vesicle was uncovered and lifted from its natural bed and then placed into the pocket Development of Amphibian Ear Vesicle 439 from which the right vesicle had been taken and allowed to heal. In making the transplantation no effort was made to place the ear vesicles in any particular posture. After keeping the speci- mens alive for one month they were sectioned and from three of them reconstructions were made of the transplanted ear vesicle together with the adjacent structures. ‘The three labyrinths are shown in Fig. 6, and model c is again shown in Fig. 5, with the brain included. It will be seen that in developing they have assumed the normal attitude toward the brain. ‘The endolym- sac. endolymph. lagena a b c Fig. 6 Reconstructions of three labyrinths which while primitive ear vesicles were transplanted from the left to the right side. They are all represented in the same position as the models in Fig. 2. The model c is the same that is shown in Fig. 5. Enlarged 50 diameters. phatic appendage and the median side of the labyrinth 1s toward the brain, the semicircular canals are toward the dorso-lateral sur- face, and the saccule and lagena are toward the ventral surface. But it can at once be recognized that the saccule and lagena point forward toward the eye, and that the anterior and posterior canals are in reversed positions. We thus have a complete mirror image of the right labyrinth, 7. ¢., a left labyrinth. Model a possesses three semicircular canals and is almost a normally formed left labyrinth. In model 4 the lateral canal consists of a pouch whose walls did not undergo the customary approximation and central absorption. In model c the posterior canal is not pinched off. In each of the models the lagena, saccule, and endolymphatic 440 George L. Streeter, M.D. appendage are typical, and there is establishment of normal appearing nerve and ganglion connections. EQUILIBRATION It was found in the experiments performed a year ago that removal of one or both ear vesicles, just after they are pinched off from the skin, produces in the tadpoles definite disturbances in the development of their power of equilibration. It was found that when a tadpole is deprived of but one ear vesicle he is by virtue of the remaining one able to develop practically normal swimming abilities; but when both ear vesicles are removed the results are more serious, and in that case the tadpole never de- velops any sense of equilibrium and is never able to swim. ‘The loss 1s not compensated for by any other organ and the animal lies helpless on the bottom of the dish. With one ear vesicle the tadpole swims practically in normal fashion, and with no ear vesi- cle he cannot swim at all. The fact that one ear vesicle is sufficient for the maintenance of equilibrium greatly simplifies the study of this mechanism; it means that one side can be immediately eliminated, and the prob- lem is reduced from a bilateral one to a unilateral one. A series of experiments at once suggested themselves, in which the ear vesi- cle of one side was to be removed, and then various operative procedures undertaken upon the ear vesicle of the opposite side, and the test of its consequent functional ability was to be the very decisive one of whether the animal could swim properly, or whether it could not swim at all. In the paper referred to there is described the experiment of transplanting the ear vesicle into a subdermal pocket in front of the eye. When this was done the transplanted ear vesicle continued in its development, and in some instances established a nerve-gan- glion connection with the forebrain; but such specimens never gave evidence of functional activity. The failure to functionate was not unexpected, inasmuch as the connections established were at an abnormal situation, and furthermore the vesicles though having developed many essential features of the normal labyrinth Development of Amphibian Ear Vesicle AAT were still quite imperfect in the formation of the separate cham- bers and thesemicircular canals. Sothis year in carrying out the experiments described in the first part of the present paper the behavior of the specimens was eagerly watched, and the endeavor was made to determine the amount of alteration in position and defectiveness in form that is compatible with functional activity, involving the problem of the correlation between function and morphology. ‘The observations made in the different experiments have been arranged and condensed as follows: a Left ear vesicle removed; right ear vesicle loosened from skin and rotated, in six specimens around the vertical axis 180° and in eight specimens around both the vertical and transverse axis 180°. As has already been shown these ear vesicles developed into labyrinths of varying degrees of perfection, some being com- pletely normal in form and having apparently normal ganglion and nerve connection with the brain wall. (See Figs. 2, 3 and 4.) The behavior of all the specimens was uniform, both where the ear vesicle was rotated in one plane and where rotated in two planes; at the end of a week after the operation, when with a nor- mally functionating labyrinth they should be able to swim freely and directly, they instead exhibit only irregular movements or spin around in spirals or circles. “‘Uheir incoordinate movements con- tinue, and at the end of a month there is no improvement; 1.c, they behave exactly like specimens with both ear vesicles removed. Evidently ear vesicles thus treated do not perform their natural function. ; b Left ear vesicle removed; right ear vesicle fragmented by teasing between the points of two needles, the fragments left in place. ‘Ten specimens were treated in this way, and were kept under observation four weeks, during which time they gave no evidence of any sense of equilibrium. c Right ear vesicle removed; left ear vesicle transplanted to the empty pocket on the right side. Five specimens were oper- ated upon and observed for one month, at the end of which time they were cut in serial sections, and it was found that the ear vesi- cles had developed into fairly complete labyrinths, but had main- tained the characteristics of a left-sided organ. (Figs. 5 and 6.) 442 George L. Streeter, M.D. Throughout the whole period of observation they had exhibited incoordinate movements, and at the end of that time they were unable to swim. ‘This and the two previous operations indicated that rotation of an ear vesicle, or transplanting it from one side to the other, or fragmenting it was not compatible with the devel- opment of its function, in spite of the fact that the ear vesicle pro- ceeded in its development and had become to all appearances almost a perfect labyrinth. In the next experiments less severe treatment was tried. d Left ear vesicle removed; right ear vesicle uncovered and carefully lifted out and then immediately placed back in its orig- inal position, the effort being made to do a minimum amount of injury. Of six specimens all exhibited symptoms of the absence of all sense of equilibrium. In the experiments a, b, c and d there was the possibility of injury to both the nerve-ganglion connection and the ear vesicle. In the following experiments the effort was made to restrict the injury to one or the other. e Left ear vesicle removed; right ear vesicle uncovered and a fragment cut from the cephalic portion of its wall, care being used not to otherwise disturb the vesicle. Eight such specimens were kept five weeks, and none of them developed any sense of equilib- rium, or were able to swim. } Left ear vesicle removed; right ear vesicle uncovered and a small piece cut from its caudal border, any further disturbance being avoided as in e. Eight specimens were operated upon, and after keeping them four weeks none of them could swim properly. g Left ear vesicle removed; longitudinal incision made through skin on right side just dorsal to ear vesicle, and needle passed down between the neural tube and ear vesicle and moved back- ward and forward so as to sever its nervous connection without otherwise disturbing the ear vesicle or loosening it from the skin. None of the four specimens studied swimmed properly, though one of them could swim somewhat, but was easily confused by any excitement and then made wild and ill directed movements. It was thought that the ear vesicles in these cases would escape injury; but examination of the specimens when cut in serial sections Development of Amphibian Ear Vesicle 443 showed that they were not perfectly normal. This experiment might be repeated on a larger number of specimens and still greater care used in severing the nerve connection, in which case a perfect labyrinth could doubtless be obtained. h (Rana catesbiana) Left ear vesicle transplanted into an- other specimen, in a subdermal pocket in the region of the pro- otic ganglion between the right eye and ear vesicle, thus the host had three ear vesicles, two being on the right side. “Twelve days after the operation three out of four specimens so treated exhib- ited incodrdinate movements. Here we have to consider the crowding out of position of the normal right ear vesicle by the one transplanted near it. 1 Left ear vesicle removed; fine needle passed through the skin so as to make a small puncture in the right ear vesicle; on with- drawal of the needle the edges of the wound immediately close and there 1s no lossof cells from underneath or from the skin itself. Of four specimens at the end of one month three were able to swim, and this demonstrated the functional ability of an ear vesicle thus treated. 1 Left ear vesicle removed; small section of the covering skin removed so as to expose the right ear vesicle, but otherwise it is not disturbed and the nerve ganglion connection is left intact. Five specimens were kept under observation for one month, and four of them behaved throughout like those possessing one untouched normal ear vesicle; except for slight incodrdination brought out by excitement they could swim properly. On bringing together the results of these experiments, it becomes immediately apparent that almost any operative procedure car- ried out on young larve in the region of the ear vesicle seriously interferes with the development of the function of that organ. It is possible to lift a skin flap and expose it, and to make a needle puncture in it without destroying its subsequent usefulness; but any operation involving a loss of part of its wall or disturbing its position and nerve-connection with the brain causes apparently complete loss of function. “The functional disturbance is out of al) proportion to the histological condition. ‘There may be a laby- rinth that to all appearances is perfectly formed and that seems to 444 George L. Streeter, M.D. have a normal nerve ganglion connection with the brain at the proper place, and yet the specimen may not have given signs of any functional activity on the part of that organ. Spemann’ is doubtless mistaken in attributing the disturbance in equilibrium simply to the alteration in the planes of the canals. He reports some experiments 1n which at an early stage a skin flap was turned back, and the ear vesicle taken out and replaced in various positions; and in such specimens he observed faulty equi- librium, and on sectioning his material the vesicle seemed to lie in an abnormal position, and this he assumes to be the cause of the abnormal movements observed. On the one hand, wax plate reconstructions of misplaced ear vesicles show that in my cases they regain their proper position, and the canals eventually lie in their normal planes; the specimens nevertheless continue to make inco- ordinate movements. On the other hand, in those experiments where the normal position of the vesicle, as regards the planes of space, was undisturbed the results were equally serious. My own experiments suggest that the difficulty lies not so much with the end organ as with the central connections, and perhaps further experiments in that direction would furnish additional infor- mation upon this subject. CONCLUSIONS The primitive ear vesicle of the tadpole may be loosened from its normal position and rotated in various directions, so that its axes lie in abnormal planes, and notwithstanding such interfer- ence it eventually develops into a labyrinth which is right side up and exhibits the normal relations to the brain and the surrounding structures. When transplanted to the opposite side of the body, if placed in the acoustic region, it likewise assumes a normal posture. Judging from these facts, the posture of the labyrinth is controlled by its environment. ‘The “laterality” of the labyrinth is determined before the clo- sure of the ear vesicle. When the left ear vesicle is transplanted 6 Spemann, H.,’06: Ueber embryonale Transplantation. Verhandl. der Gesell. Deutscher Naturf. u. Aerzte. 78 Vers. Stuttgart. Development of Amphibian Ear Vesicle 44.5 to the right side it retains its characteristics as a left-sided organ, though it otherwise adapts itself to its new position in a normal manner. The functional disturbance, in experiments on the ear vesicle, is out of all proportion to the histological appearances; any opera- tion carried out in the acoustic region involving a loss of part of the wall of the ear vesicle, or disturbing its position, or nerve con- nection with the brain results in faulty equilibrium; absence of function was observed in cases where the labyrinth and its nerve connections seemed to have attained perfect histological develop- ment. COMPENSATORY MOTIONS AND THE SEMI- CIRCULAK CANALS BY BENJ. C. GRUENBERG Witu Two Ficures HeeNeactions oftheiror ta movements onrotatoneryye tetris oe ete ilo steiee ties hilt tiered 447 2eebheones ofthe sunctionobthesemicirculanicanals mitre jects i et Petey ateeiteieteheetl teres 448 3 Theoretical objections to the semicircular canal hypothesis... ..............0+...205-++e00e 450 ApoE xperimentalobyectionsito the hypothesis eal-\-ajet aye) lere- etsiats este tet ohed totetotenee peel tore ete 452 Keeanalysis and newiexperiments eis erie ee ee 1 = oe wat Sesaralers ataepelteral ote tokels ctv eteraber ey eP Pate repae 453 Gay SUITTIIN ATV pore sas evh foyss oteNate ar abo fe foe eka 7= Stee cw eke pe Fale olay aradalhs okey seaei ca tage svoxenelePel eV epae atevedens tetekerate oeeretete 462 Fae lxeterencesrandibibliggrap hye rreve skater toe ols eles hols = oe op ustoreieie’« ebeieat enero aavtstevalele tehsyoresytaperay ieee 463 I REACTIONS OF THE FROG TO MOVEMENTS OF ROTATION When a frog is slowly turned in a horizontal plane by moving or rotating the vessel in which the frog is at rest the animal turns its head in a direction opposite to that of the rotation. When the rotation has proceeded beyond a certain point, the frog will jerk the head back into alinement with the body, and then again turn it in the opposite direction, and so on, as long as the rotation is continued. ‘The existence of this back-jerk or “ nystagmus” is specifically denied by von Cyon (’97, pp. 45, 73) and by Lyon (1899, p. 86), and has been overlooked by other observers. When the frog is restless or active, it will frequently jump or walk in a direction opposite to that of the rotation, bringing the head and body into alinement; then turn the head again and follow this movement with a jump, and so on, while the rotation is continued. But when the frog is fairly quiet, there is always a back-jerk. When the vessel containing the frog is tilted on a transverse horizontal axis, the animal nods its head up or down, according as the rotation is upward or downward anteriorly. When the base upon which the frog rests is tilted on the longitudinal (hori- zontal) axis-—too slowly to dislodge the animal—the movements THe JourNAL or EXPERIMENTAL ZOOLOGY, VOL. IV, NO. 3. 448 Benj. C. Gruenberg of the head are such as tend to keep the plane of the mouth horizontal; that is, the animal raises the head on the side that is lowered, and vice versa. (At the same time there is a contraction in the limbs on the ascending side and a corresponding flexion on the descending side.) ‘There can, of course, be no question of a nystagmus in these two cases, since the rotation cannot be con- tinued beyond a small arc of a circle without dislodging the animal or causing it to make definite efforts to hold its own, and in neither case are there compensatory responses to rotation. Nor are the responses normal if the animal is fastened to the support. When a frog is moved about in a circle having a diameter of two to three meters (by walking about with a jar containing the animal), the animal turns its head away from the center if the frog faces the direction of motion, and toward the center if the animal is carried facing in the direction opposite to that of the movement, that is, backward. ‘This movement and the response are virtually the same as in the case of rotation about a vertical axis, in a large circle and at a slow rate. ‘The movement of the head in the cases referred to is in general in a direction opposite to that of the displacement of the body. Such responses have long been known in the frog! as well as in other animals, and are frequently spoken of as “compensatory movements.’ ‘The implication of this designation, as well as the expressed belief of many physiologists, is that the movements in question are in some way related to the orientation of the ani- mal with regard to gravity, or, what is mechanically equivalent, to acceleration of motion in some direction. “The movements have been regarded as reflexes set up by the sensation of the semicir- cular canals. 2 THEORIES OF THE FUNCTION OF THE SEMICIRCULAR CANALS The oldest theory as to the function of the semicircular canals was that they were concerned in the perception of the direction of sound, and was deduced from their intimate anatomical asso- 1 Some of these responses seem first to have been described by Goltz (’69, p. 71), and many of them have been shown by Steiner (’85, p.126) to take place in frogs whose fore- and mid-brain have been removed. Compensato ry Motions 449 ciation with the other auditory organs and the fact that the three canals of each side lie in planes almost exactly at right angles to one another; that is, in planes corresponding to the three dimensions of space. It can be readily shown that the perception of the direction of sounds is actually accomplished otherwise; and this theory as to the function of these organs has been long abandoned. In 1828 Flourens (1 and 2) made the observation that, as a result of cutting one of the membranous canals in a pigeon, the bird moved about an axis at a right angle to the plane of the injured canal; that is to say, the movement was in the plane of the divided canal; the sense of hearing, however, was in nowise affected. ‘The disturbed movements were so much like those resulting from inju- ries to the cerebellum that Flourens concluded that the canals were concerned in the coordination of movements; but he made no attempt to explain the method of their operation. In 1870 Goltz offered an explanation of the method of the work- ing of the semicircular canals. According to this theory it is the downward pressure of the endolymph on the various parts of the sensitive lining of the membranous canal, according to the posi- tion of the head, that gives rise to the corresponding sensations. This theory has been called the “hydrostatic theory.” A few years later, Breuer (’74, ’75) and Mach (’75) proposed hydrodynamic theories of the action of the canals. According to Mach the sensations in the canals are aroused by variations 1n the pressure of the endolymph in the ampulla, and the varia- tions in pressure result from the streaming of the endolymph, which is caused by movements of the head. According to Breuer it is the movements of the endolymph, resulting from the move- ments of the head, that arouse the corresponding sensations, through their pressure on or movement of the lining hairs of the canals. About the same time Crum-Brown (’74, 1, 2 and 3) urged the view that the movement or pressure of the perilymph was as much concerned in the production of the sensation as the disturbance of the endolymph; and he also pointed out that the canals operate in pairs. 450 Benj. C. Gruenberg In 1878 von Cyon rejected both the static and the dynamic theories of the workings of the semicircular canals. After draw- ing off the endolymph and replacing 1 it with gelatin, and after the introduction of pieces of laminaria into the canals, thus producing great changes in pressure, there were none of the disturbances of equilibrium that had been observed by Flourens as resulting from divisions of the canals. Without advancing any other explana- tion of how these peripheral organs are excited, von Cyon main- tained that the canals assist but indirectly in giving the organism a knowledge of space relations; the sensations in the canals set up reflexes in the eye muscles, and it is from the sensations of the eye muscles and the retinal images that the notion of spatial rela- tions of the head and of the body are obtained. In 1883 Sewall (83) from experiments on skates and sharks concluded that the results were not sufhcient to warrant the opin- ion that the semicircular canals are the organs of equilibration 3 THEORETICAL OBJECTIONS TO THE SEMICIRCULAR CANAL HYPOTHESIS Whatever the real manner of operation of the semicircular canals may be, there have appeared certain theoretical objections to Goltz’s static theory as well as to the various dynamic theories; these explanations seem to be out of harmony with the observed fact that the responses of the frog’s head to rotation are not coordi- nated with the position of the animal in relation to the axis of rota- tion. ‘hus, in Fig. 1, a frog in any one of the four positions on the turntable, will “aleers turn his head to the left if the table is turned to the right (clockwise), and vice versa, as indicated by the dotted outlines and the peripheral arrows. ‘The action of gravity, or acceleration, in relation to the frog, or whatever dynamic prin- ciple it may be that does act, seems to work in a different direction in each of the four cases. ‘The special sense of the action in each case 1s indicated in Fig. 1 by a small arrow. In other words, the animal responds uniformly te what is apparently a variety of stim- uli; the stimuli in these cases are the same in kind and in degree, but differ in sense or direction, or incidence in the animal’s body: Compensatory Motions 451 but the response to every stimulus is the same in degree and in direction. Lyon (99, p. 89) has already called attention to this anomalous appearance, and the facts had been observed much earlier (Cyon 97, p. 42; Ewald ’go, Fig. 51; Schafer ’87), and had caused great confusion largely because the earlier writers described the re- sponses with reference to the periphery and the axis instead of Mog with relation to the body of the animal. The anomalies and confusions did not lead directly to a rejection of the hypothesis that the semicircular canals were the peripheral organs for per- ceiving acceleration or spatial relations. Yet these theoretical considerations would seem to make the semicircular canal hypotie- sis of equilibration untenable without radical modification. ‘The experimental evidence 1s conflicting and inconclusive. 452 Benj. C. Gruenberg 4 EXPERIMENTAL OBJECTIONS TO THE HYPOTHESIS On removing the semicircular canals entirely, or on cutting the » acoustic nerve, Girard (’92) and Schrader (’87, p. 87) report com- plete loss of compensatory motions; while Tomaszewitz (’77), Breuer (’75, p. 99), Baginsky (81), Cyon (88), Kreidl (792, 2), Ewald (’92), Strehl (’95, p. 216), Bechterew (’96), and others, found all the phenomena in response to rotation continued as in the normal animal; and Steiner (’85, 89) concludes that there is not complete loss of compensatory movements. Breuer (’91), Ewald and Delage-Aubert (’88) observe, however, that the compensa- tory movements disappear after the operation if care is taken to exclude the use of the eyes also, and therefore they do not abandon the theory; but Cyon (’97) speaks sarcastically of the logic of this argument. When single canals only were operated upon, Hensen (’79) found the movements disturbed in the plane of the canal in ques- tion; whereas Ewald (’g0, Experiment 66) cauterized the mem- branous canals in pigeons without in any way affecting the move- ments of the animals. According to Girard and Ewald, the destruction or removal of the labyrinth on one side of the head caused the frog to take on an unsymmetrical attitude, the head and body being inclined toward the operated side; Ewald found that this new attitude was main- tained in one case for a year after the operation. According to Loeb (’91, 2) cutting the acoustic nerve brings about a perma- nent tendency in the shark to turn toward the injured side. On the other hand, Cyon, Steiner (’89), Baginsky (’85) and Bech- terew ('96) found on cutting the acoustic nerve that “all the phe- nomena that served to support the assumption of the sensory function [of the semicircular canals] continued to appear’’ (Cyon, 97), while Mach found that under these conditions the eye- and head-nystagmus appeared as in normal animals. Breuer (’8g) found that mechanical, thermal and electrical (galvanic) stimulations of separate canals set up head-turnings in the corresponding planes; the movements are in response to the streaming of the lymph, and are in the same direction as the stream- Compensatory Motions 453 ing. “Galvanicdizziness * * * iscaused by irritation of the vestibular nerve endings, as galvanic phosphorescence is pro- duced by irritation of the retina.” Lee (’93, ’94, ’98) found the responses to stimulation of the canals and ampulla such as to lead to the conclusion that the canals are directly concerned in equilibration. Mach’s theory that compensating movements are set up by variations in the pressure of the endolymph seems to be disproved by the experiments of Cyon (’88, pp. 294-297), Spamer (80) and Ewald (90, Experiment 42), who secured normal reactions after producing permanent changes in the pres- sure of the endolymph by removing the liquid entirely, by replacing it with gelatine or with amalgam, and by inserting into the canals dry bits of laminaria, which swelled up on absorbing moisture, and so increased the pressure. On the one hand Ayers (’9g2) draws from his morphological studies the conclusion that the canals are specially modified lateral- canal organs, that have no relation whatever to equilibration. On the other hand Schaeffer (’94) tells us that whirling produces no effect upon tadpoles until after the semicircular canals become developed. But Streeter (’06) has succeeded in separating the action of the canals from that of the rest of the ear vesicle and concludes that while the ear vesicles are essential to the develop- ment of the power of equilibrium in tadpoles, the canals are not. Schafer (87) succeeded in demonstrating to his own satis- faction that the responses to rotation are due solely to the inertia of the loosely jointed head; he made a wooden model that behaved on the turntable just like a frog or a pigeon, with a few exceptions to be noticed later. Other minor experiments have been reported by various inves- tigators, with equally definite but conflicting and inconclusive results. 5 ANALYSIS AND NEW EXPERIMENTS This then appears to be the situation: 1 From the structure of the semicircular canals it was inferred that they were somehow related to the perception of space or direc- tion. 454 Benj. Ff. Gruenberg 2 The manner in which the semicircular canals operate to bring about perception of space has been variously explained as resulting from (a) static inertia of the endolymph; ()) variation in pressure of the endolymph; and (c) movements or acceleration of the endolymph, or of the perilymph, or of both; brought about by movements of the head. 3 Theoretical considerations seem to show that this function cannot be ascribed to the canals for the reason that identical reac- tions are produced under conditions in which the sense of the acceleration may be opposite. (As explained in connection with ia) 4 But operations to remove or destroy the semicircular canals, or to sever the connections of the VIII nerve, show (a) in some cases that the movements of the animals are seriously affected, and (b) in other cases that the animals continue to respond to rotation as do the normal animals. 5 Mechanical, thermal and electrical stimulations of the single canals show (a) in some cases decided disturbances of movement related to the planes of the respective canals, and (b) in other cases the absence of related responses. On examining again the movements represented in Fig. 1, it will be seen that a given rotation will produce for the frog a dis- placement of the retinal image or “view,” and always in the same direction without regard to the position of the animal on the turn- table. Whereas the actions of centrifugal force and of acceleration depend upon the position of the animal with relation to the axis of rotation, the sense of displacement of the field of vision does not so depend, and it may therefore be supposed that the uniform turnings of the head are in response to the changing view; the frog seems to be trying to keep the same view before him. To test the responsiveness of the frog to the apparent displace- ment of his surroundings, a “revolving environment” was ar- ranged, consisting of a cylinder of stout paper about 60 cm. in diameter and about 35 cm. high, attached to a wooden hoop which was suspended so that it could be readily rotated in either direc- tion. A portion of the cylinder consisted of light colored material bearing black vertical stripes about 5 mm. wide and from 2 to 5 Compensatory Motions 455 cm. apart. Another portion of the cylinder was of black paper in which had been cut numerous holes of various shapes and sizes. When in use the cylinder was always part striped and part fenestrated, or part open and part one or the other of the described surfaces. (Ihe proportions were varied, but the character of the surface did not seem to make a constant difference. ) Frogs placed in the middle of this “circus” arrangement could be made to turn their heads and to give the nystagmus or back- jerk by revolving the cylinder, the same as when the animals them- selves were rotated on the turntable. ‘The response was not, how- ever, equally marked in all cases, nor was it in any case as quick as in the actual rotation of the animal. When the revolving of the cylinder was very rapid or very slow, there was no response at all; but when the optimum rate was found, the responses were well marked and continuous. ‘hese experiments with the movy- ing environment would indicate that the visual impressions do, or may, play an important role in setting up compensatory move- ments; and in the case of the animal rotated on the turntable one might conclude that it is the displacement of the retinal image that is the constant, and therefore the determinant factor. But such a conclusion would be false, and for the following reasons: If a frog is placed on the turntable, in every possible position with relation to the pivot, and the table is turned to the right, (that is, clockwise) the frog’s head will always turn to the (ani- mal’s) left, and the animal PT seek, humanly speaking, to keep the same view before his eyes. But if now the vessel containing the animal is completely surrounded by some opake material, the frog will respond in precisely the same way. If the frog is taken into a room almost dark—one barely light enough to permit the observer to discern the outlines of the animal against a white background—the animal will respond in the same way. If the animal is placed upon the turntable together with the source of illumination, and completely cut off from the sight of external objects, rotation will result in the same reactions. If, finally, the animal’s eyes are covered with a mixture of vaseline and lamp- black (which will entirely exclude vision without in the least irri- tating the frog) the responses to rotation are still the same. 456 Benj. C. Gruenberg It may accordingly be safe to conclude that while the turnings of the head on rotation may be responses to visual impressions, they may also be quite independent of visual impressions. One is therefore driven back to a reéxamination of the semicircular canal theory, or to search for some other means of perceiving movement or acceleration. It had already been found that there is not complete loss of the compensatory movements on cutting the acoustic nerves,’ or on destroying the semicircular canals. ‘his is comprehensible, since the eyes are also capable of leading to similar results. If a frog that has been operated upon is rotated with the eyes covered, or surrounded by some opake medium that rotates with him, there is no response. ‘This excluding of visual impressions is not, as Cyon supposed, eliminating the determining factor, since the nor- mal frog under the same conditions will continue to react, though Cyon failed to observe this. ‘The results referred to in this para- graph I have verified experimentally. The following is the record of one frog whose semicircular canals had been destroyed by piercing into the capsule from the dorsal side. The animal was etherized; there was no bleeding. 1 Immediately after the operation (the animal recovered con- sciousness and began to move about sluggishly within two or three minutes after the operation): a Animal lies on back quietly over one-half minute at a time without making efforts to right itself. b Rights itself only with great difficulty and after making many awkward movements. c Limbs not correlated in crawling about; does not hop. d On turntable, no response. e In swimming, rolls from side to side. 2 After thirty minutes: a Lies on back for short intervals, but not quietly. b Rights itself with difficulty, but more quickly than at first. c Moves about awkwardly, but better than at first; can hop, but in jumping frequently lands on back. *Vide supra. Compensatory Motions 45 “SI d On turntable responds as normal, but more slowly. e Responds as normal to “revolving environment.” Be Aiter One week: a Lies on back indefinitely, quietly. b Rights itself more easily than before, but still awkwardly. c Walks about unsteadily; leaps awkwardly, falling on side. d On turntable, responds normally. Another frog, with both sets of semicircular canals destroyed by boring into capsule from the dorsal side, showed after three weeks a marked lack of coordination of movements, though not as great as at first; this was evident in swimming as well as in walk- ing, and in both swimming and in jumping the animal frequently turned over on its back; it righted itself rather quickly, but move- ments still showed awkwardness. On turntable, responses were as in the normal animal. As has been pointed out above, the summation of mechanical disturbances or accelerations on the rotation of an animal upon the turntable in a given direction seems to depend upon the posi- tion of the animal with reference to the axis of rotation, whereas the sense of the response in relation to the animal’s own axis 1s constant. Thus, the rotation being to the right (clockwise), the factor of wind, or resistance of the air, acts upon the right side of the animal if the animal faces the periphery, but on the /e/t side of body if the animal faces the pivot; but in any case the response is to the /eft. The same apparent contradiction 1s observed if we consider the direction of the centrifugal force of the rotation; the direction of the centrifugal pressure of the viscera or other loose parts, of the strains on the skeletal articulations, and of the friction of the body on the supporting surface, 1s toward the periphery, however the animal may be placed; but the response to a given rotation is constant with reference to the axis of the animal. The same apparent contradiction is found when the attention is directed to the inertia of the viscera or of the contents of the semi- circular canals. In addition to these contradictions is the fur- ther fact pointed out by Schafer in 1887 and by others, and re- ferred to above (§4), that the inertia of the head, because of its loose articulation to the trunk, is sufficient to account for the 458 Benj. C. Gruenberg “responses’’ to rotation even in a wooden frog; and these re- sponses agree in sense with those described for the live frog. That is to say, with a given rotation of the turntable the “turning” of the head is constant, without regard to the position of the body in relation to the axis of rotation. ‘These considerations in detail have led many physiologists to abandon the theory that the canals are the organs for the perception of movement or acceleration, since they so obviously arouse the same response to opposite sets of stimuli, and since the responses can be obtained from wooden animals as well as from Nature’s own. But there is one element in the mechanical theory that seems to have been overlooked as a constant source of rotation stimult. ‘The inertia of a loosely jointed head as an explanation of the phenomena may be left entirely out of account because in the first place i it cannot account for the back-jerk, in the second place the inertia 1s overbalanced by the centrifugal force when a certain rate of rotation 1s reached, whereas the responses do not dis- appear at this point, and in the third place the responses can be inhibited by stimuli that do not seem to affect the freedom of the head articulation. Steiner (785) has described the reactions of the frog in response to rotation essentially as given here, and analyzed the movements in terms of tangential force; but he concludes from the persistence of the reactions after the division of the eighth nerve, that the canals are not concerned in the matter at all. It is to be noticed that when the animal is rotated on a turntable, the posterior end of the body is constantly changing its position with reference to the anterior end, as is the right side with reference to the left, ete. ‘This movement is constant in direction, and parallel (in direction) to the rotation. ‘Uhat this is a real motion and quite distinct from the motion of translation or rotation, is known to every physicist and to some laymen; but the physicist has a name for it. Prof. Albert P. Wills, of the Department of: Physics of Colum- bia University, has kindly helped me to get the matter clear by assuring me that this kind of motion is well recognized in mechan- ics, and by giving me the technical name for it. It is known as the “spin.” ‘This spzn it is that remains constant in direction on Compensatory Motions 459 a given rotation, unaffected by the position of the animal on the turntable; when the rotation is clockwise, the spin is clockwise, and vice versa; when the spin is to the right, the head of the frog turns to the left, and vice versa. It is the “‘spin,”’ therefore, that determines the compensatory movement. To test the validity of this interpretation, it is necessary to eliminate this factor from the rotation of the animal. For this purpose an eccentric was arranged on the turntable in such a FiGeZ manner that the animal could be moved in a circle with its long axis always parallel to its first position (Fig. 2). When the frog is rotated on this eccentric turntable, the re- sponse takes the form of a pendulous movement of the head; during one portion of the revolution the head is turned to the right, and then the head is turned in the opposite direction. “Uhis back and forth swinging of the head continues as long as the rotation 460 Benj. C. Gruenberg is continued, and there is no nystagmus whatever. The head turns to the left in that part of the rotation which carries the ani- mal’s body to its right, and vice versa, whether the rotation be clockwise or the reverse, and without regard to the size of the circle described; that 1s, without regard to the proximity of the animal to one of the pivots. If the animal is placed in an opake vessel and the rotation on the eccentric set up, there is no response whatever. To show the effects of the various factors that have been con- sidered as having a possible relation to the “compensatory move- ments,’’ the following comparative table may be helpful. In the experiments whose results are given under I[ the animal was I II i ae i. yaa Cc D A B Facing Periphery Facing Pivot Facing Periphery Facing Pivot > I Air pressure 2 Inertia (of viscera, lymph, etc.) Centrifugal action w 5 Displacement of retinal image 6 “Spin” eS zee x de smacdion bits plod mee <— © < PST eae <-- a ‘b Vv ae <= Paes O co = ae rotated to the right (clockwise) on the ordinary turntable; in those represented under II the eccentric arrangement was used and the results given are for a portion of the revolution only, since a continuation of the rotation beyond 180° is virtually equivalent to a reversal of the motion. (It is of course understood that the rotations in the reverse direction gave corresponding results but in the opposite sense.) “lhe arrows indicate the directions in which the respective factors are supposed to act. Response 7 Compensatory Motions 461 On comparing the arrows in the four columns it will be seen that whereas all in II seem to be related to the direction of the head turning, none in I are so except 5 and 6 (retinal and spin impres- sions). In experiments on the turntable (1) the factor 5 could be elim- inated in a variety of ways: By surrounding the vessel containing the frog with some opake material, or placing it in a tall opake cylinder; by covering the eyes with the opake non-irritant mix- ture already referred to, or with a pad of absorbent cotton mixed with vaseline and lampblack; by placing the source of light on the turntable with the animal. In all cases the turning of the head in response to rotation was the same as in the usual rotation as to direction; but frequently it was less in degree. In other words, while the displacement of the retinal image can and does set up the compensatory response, the eye is not the sole sense organ through which such movements can be initiated. This leaves the spin as the only other factor to be further con- sidered. According to the results indicated in columns A and B of the table, the spin is the only factor (of those considered) in addition to vision that can constantly set up the head turning. In the experiments on the eccentric (columns C and D) where the spin is already eliminated, the further elimination of sight results in a total loss of the response. The slightest amount of spin is sufficient to set up a perceptible amount of head turning; considerable displacement of the retinal image is required to bring about the same amount of response. It is possible to move the frog in a right line without the animal giving any response whatever; but if the movement is not smooth, that is, if there is vibration, or very slight turning in a horizontal plane, the head responds at once. ‘That the response to the spin is quicker and greater in amount is also certain; the two factors may be caused to operate in opposite directions in the following manner: A dish holding the frog on a horizontal plane and facing the observer, is swung about slowly by the observer at arm’s length. The head will be seen to turn in the same direction as the move- ment of translation; that is, in a direction opposite to what we 462 Benj. C. Gruenberg should expect on the anthropomorphic view of the animal “seek- ing to keep the same vision in sight.” But the turning of the head is Opposite to the direction of the spin that the observer uncon- sciously imparts to the dish in moving his arm outstretched, which is thus in the radius of a horizontal rotation. That the perception of spin or rotation is located in the organs of the inner ear seems likely from the fact that the response is eliminated when the semicircular canals are destroyed or removed, or when the acoustic nerve is cut. “That the sensation concerned involves a factor of rotation or turning is indicated by the fact that rectilinear acceleration does not yield the same constant response. It may, therefore, be concluded that the compensatory move- ments of the frog’s head set up by rotation arise in response to two distinct sets of stimuli, visual and dynamic; that the response to the visual stimulus is relatively feebler and slower than that to the dynamic stimulus; that the organ for the perception of the dynamic factor 1s probably located in the internal ear; and that the dynamic perception involves a rotation or turning element in the stimulus, as distinguished from an acceleration or movement in a single direction. 6 SUMMARY 1 There is apparent contradiction between the various re- sponses of the frog to rotation on the turntable and any theory of mechanical stimulation of peripheral organ as the origin of the responses. 2 There is considerable contradiction among various experi- ments that have been made in connection with the relation of the semicircular canals and compensatory movements. 3 A reéxamination of the compensatory movements and of the conditions under which they arise shows the presence of a mechanical factor, the “spin,”’ the significance of which in this connection seems not to have been considered before. 4 From an examination of the results obtained by earlier observers, a repetition of some of their experiments, and new experiments made in the course of the study, the following con- clusions are drawn: Compensatory Motions 463 a ‘Lhe compensatory movements of the frog’s head set up by rotation arise in response to two distinct sets of stimuli, visual and dynamic. b ‘The response to the visual stimulus is relatively feebler and slower than that to the dynamic stimulus. c The organ for the perception of the dynamic factor 1s prob- ably located in the internal ear. d ‘The dynamic perception involves a rotation or turning element in the stimulus, as distinguished from an acceleration or movement in a simple direction. I wish to express my sincere thanks to Prof. TV. H. Morgan for the helpful suggestions and encouragement that have made this study possible; and to Prof. A. P. Wills for his assistance in eluci- dating to me the mechanics of rotation. 7 REFERENCES AND BIBLIOGRAPHY 1828 FLourens, Marre JEAN PrerRE—(1) Expériences sur les canaux semicir- culaires de l’oreilles dans les oiseaux. Memoirs de |’acad. roy. des sci., 9:455-466. 1830. Read before the Academy, 11 Ag., 1828. (2) Expériences sur les canaux semicirculaires de |’oreilles dans les mammi- feres. Memoirs de l’acad. roy. des sci. 9:467-477. 1830. Read before the Academy, 13 Oc., 1828. 1842 FLourens, M. J. P.—Recherches expérimentales sur les propriétés du systeme nerveux. Paris. p. 438. 1846 PRECHTL, JOHANN JosEF—Untersuchungen tiber den Flug der Vogel. Wien, Carl Gerold. p. 212. 1866 Manoyer, M.—Recherches expérimentales sur la locomotion chez les poissons. Ann.desSci.natur. Ser. V. 6:5-15. (Function of the fins and of the air bladder in locomotion and in equilibration; specific gray- ity relations.) 1868 Lerypic, F.—Ueber Organe eines sechsten Sinnes. Dresden. 1869 Go1tz, Fr—Beitrage zur Lehre von den Functionen der Nervencentren des Frosches. Berlin, Hirschwald. pp. 130. 1870 Gottz, FrrepRicH—Ueber die physiologische Bedeutung des Ohrlaby- rinths. Pfluger’s Archiv. 3:172-193. SCHULZE, Franz E1tHarD—Ueber Sinnesorgane der Seitenlinie bei Fischen und Amphibien. Arch. Mikr. Anat. 6:62-88. Taf.4-6 (The sense 464 1873 1874 1875 1876 Benj. Ce Gruenberg organs of the lateral line serve for the perception of water waves of lower frequency than those perceived by the ears.) Cyon, E. von—Cours de physiologie. St. Petersburg. vol. ii. Breuer, J.—Ueber die Funktion der Bogengange des Ohrlabyrinths. Med. Jahrb. pp. 72-124. Brown, ALEXANDER CruM—(1) Preliminary Note on the Sense of Rotation and the Function of the Semicircular Canals of the Internal Ear. Proc. Roy. Soc. Edin. 8:255. 19 Ja., 1874. (2) On the Semicircular Canals of the Internal Ear. Proc. Roy. Soc. Edin. 8 :370-371. (3) On the Sense of Rotation and the Anatomy and Physiology of the Semi- circular Canals of the Internal Ear. J]. Anat. and Phys. 8:327-331. Cyon, Ernst von—Ueber die Function der halbcirkelférmigen Canile. Pfliiger’s Archiv. 8:306—326. Breuer, J.—Beitrage zur Lehre vom Statischen Sinne (Gleichgewichtorgan, Vestibulapparat des Ohrlabyrinths.) Med. Jahrb. pp. 87-156. Macu, Ernst—Grundlinien der Lehre von den Bewegungsempfindungen. Leipzig. Cyon, Ek. von—(1) Methodik der physiologischen Experimente. St. Petersburg. pp. 540-547. (2) Rapports physiologique entre le nerf acoustique. Compt. Rend. Acad. Sci. 82:856. Cyon, E. von—Les organs periphériques du sens de l’espace. Compt. Rend. 85:1284. Tomaszewicz, ANNA—Beitrage zur Physiologie des Ohrlabyrinths. Zurich. Inaug. Diss. Brown, ALEXANDER CrumM—Cyon’s Researches on the Ear. Nature. 18 :633-635; 657-659. (Historical summary and abstract of Cyon, 1878). Cyon, E. von—Recherches expérimentales sur les fonctions des canaux semicirculaires et sur leur réle dans la formation de la notion de |’espace. Thesis. Hensen, V.—Physiologie des Gehérs. Hermann’s Handbuch der Physiol- ogie. III (2). SpAMER, C.—Experimenteller und kritischer Beitrag zur Physiologie der halbkreisformigen Kanale. Pfliiger’s Archiv. 21:479-590. Bacinsky, BENNo—(1) Ueber die Schwindelerscheinungen nach Ohrver- letzungen. Monatsb. Berl. Acad. (2) Die Funktion der Bogengange des Ohrlabyrinthes. Biol. Zentralb. 1 :438-446. 1882 1883 1884 1885 1886 1887 1888 1889 1890 Compensatory Motions 465 KiesseLtpacH—Zur Funktion der halbzirkelformigen Kanale. Arch. fiir Ohrenheilkunde, 18. Srwa._, H.—Experiments upon the Ears of Fishes with Reference'to the Function of Equilibrium. Jl. Phys. 4:339-349. Buppe, E.—Ueber metakinetische Scheinbewegungen und tiber die Wahr- nehmung der Bewegung. Arch. Phys. 126-152. Bacinsky, Benno—Zur Physiologie der Bogengange. Arch. Anat. u. Phys. 253-266. STEINER, J.—Untersuchungen tber die Physiologie des Froschhirns. Braunschweig, Vieweg. pp.,143, 32 cuts. Macu, E.—Beitrage zur Analyse der Empfindungen. Jena. STEINER, Is.—Ueber das Central nervensystem des Haifisches und des Amphioxus, und uber die halbcirkelformigen Canale des Haifisches. Sitzungsb. Berl. Akad. 20 My., 1886. pp. 495-499. DeELaGeE, YvEs—Sur une fonction nouvelle des otocystes comme organes d’orientation locomotrice. Arch. Zoél. Exper. et Gen. Ser. II. 5:1- 26. (Function of otocysts in orientation of invertebrates. ) ENGELMANN, TH. W.—Ueber die Funktion der Otolithen. Zool. Anzeiger. 439-444. ScHAFER, Karit—Ueber die Wahrnehmung einiger passiven Bewegungen . durch den Muskelsinn. Pfliiger’s Archiv. 41:566-640 (30 wood cuts). ScuraDER, Max FE. G.—Zur Physiologie des Froschhirns. Pfliiger’s Archiv. 41:75-90. Cyon, E. von—Gesammelte physiologische Arbeiten. Berlin, August Hirschwald. p. 338. DELAGEs, YVES AND AuBERT—Physiologische Studien tiber die Orientirung. Tubingen. Logs, JAcgues—Die Orientierung der Tiere gegen die Schwerkraft der Erde. Sitzb. phys-med. Gesel., Wiirzburg, Jahrg. 1888, 5—10. Breuer. ].—Neue Versuche an den Ohrbogengangen. Pfliiger’s Archiv. 44.:135-152. Brown, ALEXANDER CruM—Our Sensation of Motion. Armitstead lec- ture, in Dundee. Nature, 40:449-453. (The mechanics of possible stimulation through motion). STEINER, J.—Der Meniere’sche Schwindel und die halbzirkelformigen Canale. Deutsche med. Wochensch. No. 47. 958-960. Leipzig. Ewa.p, R.—Physiologische Untersuchungen tiber das Endorgan des Ner- vus Octavus. Berl. Klin. Wochensch., Nr. 32. Reprinted, Wiesbaden, 1892. Marey, E. J.—Le vol des oiseaux. Paris, G. Mason. 466 1891 1892 1893 1894 1895 1896 Benj. C. Gruenberg Breuer, J.—Ueber die Funktion der Otolithen Apparate. Pfliiger’s Archiv. 50:195-306, Pl. III-V. KreipL, ALors—Beitrage zur Physiologie des Ohrlabyrinths auf Grund von Versuchen an Taubstummen.- Pfliiger’s Archiv. 51:119g~-150, Pl. VII. Lors, Jacgues—(1) Ueber Geotropismus bei Tieren. Pfliiger’s Archiv. 49:175. (Studies in General Physiology, Chicago, 1905, pp. 176-I9go.) (2) Ueber den Antheil des Hohrnerven an den nach Gehirnverletzung auf- tretenden Zwangsbewegungen, Zwangslagen und assoziirten Stellungs- anderungen der Bulbi und Extremitaten. Pfliger’s Archiv. 50:66-83. Ayers, Howarp—A Contribution to the Morphology of the Vertebrate Ear with a Reconsideration of its Functions. Jl. Morph. 6:1-360. (With bibliography of 295 numbers.) (Author’s abstract, as lecture,in Wood’s Hole Biol. Lect. for 1890.) Ewatp, R.—Bedeutung des Ohres fur normale Muskelkontraktionen. Zentralbl. fiir Phys. 5:4-6. Grrarp, H.—Recherches sur la fonction des canaux semicirculaires de loreille interne chez le grenouille. Arch. de Physiol. 24:353-365. Kreip_, ALors—(1) Zur physiologischen Bedeutung des Ohrlabyrinthes. Neurol. Centralbl. 11:222-223. (2) Weitere Beitrage zur Physiologie des Ohrlabyrinthes. Sitzungsb. Wien Akad. Wiss. Bd. 101, Abt. III, 469-480. 1893. 102 (I1):149-174. Ler, FREDERICK S.—A Study of the Sense of Equilibrium in Fishes. Jl. Phys: 15-201: Breer, THEODORE—Die Accomodation des Fischauges. Pfliiger’s Archiv. 58:523-650; 35 fg. 3 pl. (With bibliography of 65 numbers.) Berue, ALBrEcHtT—Ueber die Erhaltung des Gleichgewichts. Biol. Zen- tralb. 14:95-114, 563-582. Leg, F. S.—A Study of the Sense of Equilibrium in Fishes. Jl. Phys. 17:192-210. SCHAFFER, Kart L.—Funktion und Funktionsentwickelung der Bogen- gange. Zeistchr. f. Psych. 7:1-9. (Whirling produces no effect upon tadpoles; but after the canals become developed, they show signs of dizziness. ) STREHL, H.—Beitrage zur Physiologie des inneren Ohres. Pfliger’s Archiv. 61:205-234. Warren, H. C.—Sensations of Rotation. Psych. Rev. 2:273-276. BECHTEREW, W. von—Ueber die Empfindungen welche mittels der soge- nannten Gleichgewichtsorgane wahrgenommen werden, und uber die Bedeutung dieser Empfindungen in Bezug auf die Entwickelung un- serer Raumvorstellungen. Arch. f. Phys. 105-141. 1896 1897 1898 1899 19OI SNES 1904 1905 1906 EOS 7, Compensatory Motions 467 Ciark, GayLorp P.—On the Relation of the Otocysts to Equilibrium Phe- nomena in Gelasimus pugilator and Platyonichus ocellatus. Jl. Phys. 19 :327-343. Cyon, E. von—Bogengange und Raumsinn. Arch. f. Phys. 29-111. Ler, F. S—The Function of the Ear and the Lateral Line in Fishes. Am. Jl. Phys. 1:128-144. (Reviewed in Zool. Zentrabl. 6:409-411.) Devirz, J.—Ueber den Rheotropismus bei Tieren. Arch. f. Phys. (Suppl.) 231-244. WHEELER, WiLtt1AmM Morton—Anemotropisms and Other Tropisms in Insects. Arch. fiir Entwickelungsmechanik. 8:373-381. Lyon, E. P.—A Contribution to the Comparative Physiology of Compen- satory Motions. Am. Jl. Phys. 3:86—114. Prentiss, C. W.—The Otocyst of Decapod Crustacea; its Structure, De- velopment and Functions. Bull. Mus. Comp. Zool. Harvard. 306. Tu sere, T.—Das Labyrinth der Fische, ein Organ zur Empfindung der Wasserbewegung. Beihang: K. Svenska Vet-Akad. Handlingar, Stock- holm, vol. 28, no. 15. 25 pp. Lyon, E. P.—On Rheotropism in Fishes, I. Am. Jl. Phys. 12:149-161. YERKES, Ropert M.—Inhibition and Reinforcement of Reaction in the Frog. Jl. Comp. Morph. and Psych. 14:124. PARKER, Ae he eh veralsline Orns andiBarae@reacene eanile bration. Proc. Am. Zool. Soc. Science, 20. Haptey, Puitie B.—The Relation of Optical Stimuli to Rheotaxis in the American Lobster. Am. Jl. Phys. 17:326-343. STREETER, GEorGE L.—Experiments on the Developing Ear-vesicle of the Tadpole with Relation to Equilibration. Jl. Exp. Zool. 3:543-558. Logs, Jacgues—Ueber die Summation heliotropischer und geotropischer Wirkungen bei den auf der Drehscheibe ausgelosten compensatorischen Kopfbewegungen. Pfliiger’s Archiv. 116:368-374. A STUDY OF THE SPERMATOGENESIS OF TWENTY- TWO SPECIES OF THE MEMBRACIDZ, JASSIDA, CERCOPIDA AND FULGORIDA, WITH ESPECIAL REFERENCE TO THE BEHAVIOR. OF THE: ODD CHROMOSOME! ALICE M. BORING Witn Nine Pirates IARC CH ONG wo asooaddds FOPETUUO DDE OoDeURUGLEUpe Ue oo pisdzovo ae oHuooceNandconsonosasoae 470 IFistonicalirevieniecs savers ciissreis ave s\sinyeaiafsla ei situa acme Poors letey cauptics ooktee ts eheletnate cca tore tetera. Mave terstane 470 iraterialwandbmetho dsm sianicte. cise siarcte¥s aya eves s 7ene cre, sieiate ls lets: eunusieysteasiouster sie ee eaee area een ea eee trae 478 (Cleisanvetntns cs aaeaan Sent pic in aie ROI a ee MS ae rain Rass Roath nia ticmis nice or 480 Wieralbracidacr te wcv ey trevor Sas eitecess ee susie alae Sala sie sister arate GTS ier ea lato erie 480 Bn tildisiitta tasyereetare craters + Gicheie «srertea ate ess topiskers « ayntanads tetete Sys,cietis ete er eRe eter Reet 481 Wan duzeatarcu ata ste crtetscrsercts revels athe nicole lens eonttveua, are felstccr overcome cteyaras arsvenehe coemeie sree roars eres 486 Geresatauninac ss steve ycteye eevee) ocvey arate se bwrouel ses saya e 7S 005, c oteeevane ap tvcheiegscassberatneberapeletera eT eracrameects 487 Geresanbubyal us eccscricte cose cht iat seater ol Sroiao vio aeeIaie tare SEE TE IO aa: COS aa 488 Ger esa dicerosneriss JARRE ee Oe Cot ROE a one ae HOen te emak aa noina ce bierichae A 506 Bibliography emer etre terekaraes eka cut cioig sees siege Eee catera a ae ee Ne eR aite eee 509 DESI PLIOM OMe lates mays tek pnee. a cet cts ica tic Meare Se yee cases eee ees atk eres a eet acter re 513 ‘A dissertation presented to the Faculty of Bryn Mawr College for the degree of Doctor of Philosophy. Tue JourRNAL or ExpERIMENTAL ZOOLOGY, VOL. IV, NO. 4. 470 Alice M. Boring INTRODUCTION The purpose of this investigation is to extend, to some families of the Hemiptera Homoptera, the studies of McClung, Stevens, Wilson and others on the relation of the accessory or odd chromo- some to sex determination. Except for the aphids, which have been extensively worked out by Stevens (’o5a, ’o6a), Cicada tibi- cens (Wilcox ’95) and Aphrophora quadrangularis (Stevens ’o6b) are the only species of this group whose spermatogenesis has been previously described. This study covers eight species of the Membracidz, six of the Jassida, four of the Cercopide and four of the Fulgoridz. My work was begun at the suggestion of Dr. N. M. Stevens at Woods Hole in the summer of 1905, continued under Prof. E. G. Conklin, at the University of Pennsylvania, in the year 1905-06, and completed under Dr. Stevens, at Bryn Mawr College, in the year 1906-07. To both Dr. Stevens and Professor Conklin I wish to express my appreciation of their valuable suggestions and constant help and inspiration. I wish also to thank Dr. Herbert Osborn of Columbus, Ohio; Mr. E. P. Van Duzee, of Buffalo; Mr. H. C. Barber, of New York City, and Dr. H. Skinner, of Philadelphia, for the identification of material. HISTORICAL REVIEW Most of the work on the spermatogenesis of the tracheate arthro- pods has been done since 1890. Such studies as those of Bitschli (71), La Valette St. George (’85), Platner (’86), Verson (’89), and Sabatier (’85) were concerned only with the formation of the spermatozoa, the arrangement of the cells of the testis into cysts, and the general mechanics of karyokinesis. The work of van Beneden (84), Boveri (’87) and O. Hertwig (’90) on Ascaris, and Mark (’8r) on Limax, turned the interest in the study of the sex cells to the chromosomes, while Weismann’s daring hypothesis (87) as to equational and reducing divisions added to the interest. By 1890, practically all investigations on spermatogenesis centered around the chromosomes in the spermatocyte divisions, and in S permatogenests 471 that year we find the first statement that one chromosome behaves differently from the others (Henking ’90). Unfortunately there 1s the greatest confusion in the results for the next decade; but since Montgomery’s suggestion (’o1a) that synapsis means the conjuga- tion of homologous maternal and paternal chromosomes, and its confirmation by Sutton’s work on Brachystola (’00, 02, ’03), there has been greater accord. As a consequence of this, certain funda- mental theories are coming to rest on a firm foundauon. ‘The chro- mosomes are shown to keep their individuality from one cell gener- ation to another. The real reduction in number is proved to be brought about by the joining of each paternal to a corresponding maternal chromosome in synapsis. It is found to make no differ- ence whether the reducing or equational division comes first, but the distinction between these two divisions is constant, the one being the separating of the individual spermatogonial chromo- somes, the other a simple splitting of these univalent chromosomes. In addition to this, recent work indicates that there is usually present throughout the Tracheata an odd chromosome in the spermatogonia, which behaves differently from the other chromo- somes throughout its history. Still later work seems to establish the fact that this chromosome has no paternal mate, does not join any other chromosome in synapsis, divides in only one sperma- tocyte division, and enters only half of the spermatozoa. In some forms, a small chromosome is present as the paternal mate of this odd chromosome, but dimorphism of the spermatozoa results in either case. The following review takes up the different observations on the Tracheata since 1890, and attempts to show how each helps to establish, or differs from, the above mentioned theories. Arachnida Wallace (’05) finds an even number of spermatogonial chromo- somes, 40, two of these being larger than the others and different in behavior. They are condensed in the spermatogonial rest stage, and take an eccentric position in the equatorial plate. They remain separate from each other in the spermatocyte growth 472 Alice M. Boring period and do not divide in either spermatocyte division, as the other 19 chromosomes do, thus appearing in only one quarter of the spermatozoa. Wallace concludes that all the spermatozoa degenerate except those with the two odd chromosomes. Montgomery in Lycosa (’05) finds an even number of chromo- somes in the spermatogonia. ‘['wo of these he calls heterochromo- somes, although the only characteristic that justifies this name is that they remain condensed in the growth period. ‘They conju- gate like the other chromosomes and divide in both divisions, all of the spermatozoa receiving one-fourth of the heterochromosome tetrad. The results of neither of these investigators agree with the more recent work on the odd chromosome in spiders and other forms. If, as Wallace states, no spermatozoa develop except those con- taining the two odd chromosomes and the nineteen ordinary chromosomes, the eggs must all contain only 19 chromosomes, as the spermatogonial number is 40. Suppose each egg to have 19 chromosomes; fertilization by a spermatozoOon with 19 +2 chromo- somes would give all the offspring 38 +2 (19 +2 in the reduced number), whether male or female; but according to Wallace’s con- tention, the egg can have only 19; therefore it is impossible that all the spermatozoa, except those with the two odd chromosomes, degenerate. According to Montgomery, the heterochromosome in the spermatocyte is bivalent and divides in both divisions. Berry’s work (’06) brings the odd chromosome inthe spider into line with the odd chromosomes in other forms; it is a single chromo- some in the spermatogonia, and divides in only the second divi- sion of the spermatocytes, resulting in dimorphism of the sper- matozoa. M yriapoda Blackman (’o5a, ’o5b) finds in Scolopendra heros and S. sub- spinipes an uneven number of spermatogonial chromosomes. Synapsis takes place in the late anaphase of the last spermatogonial division, all of the chromosomes uniting in pairs except the odd one. The odd chromosome divides only in the second spermato- cyte division. ‘The peculiarity here is that the other chromosomes S permatogenesis 473 seem to undergo their reducing division when the odd chromosome is dividing equationally, but this is only a further mark of the indi- viduality of the chromosomes, and does not furnish any evidence against Montgomery’s theory of synapsis. Medes (’05) finds a similar condition in Scutigera forceps. Orthoptera Neither vom Rath (91, ’92) nor Wilcox (’95) noticed an odd chromosome in Gryllotalpa or Caloptenus, although both mention a nucleolus in the spermatocyte growth period which may be the same structure. [hey both insist that there are two reducing divisions; that is, two divisions that separate whole chromosomes from each other. ‘This is probably due to a confusion in the use of the word chromosome. If we use the terminology suggested by McClung (’00), univalent chromosome in the spermatogonium, bivalent chromosome in the spermatocyte, and chromatid for each unit of the tetrad, the discrepancies inthe work of vom Rath and Wilcox are cleared up. Vom Rath finds 12 spermatogonial chro- mosomes. In the growth period, the spireme splits into six rods, each of which forms a tetrad, or divides into four “chromosomes,” as he expresses it. As he calls each chromatid a chromosome, he considers that he has two divisions which separate chromosomes from chromosomes; and therefore must be reducing; while in terms of the original spermatogonial chromosomes, one division is reducing and one equational. Wilcox falls into the same diff- culty; he finds 12 spermatogonial chromosomes, and then the spireme divides into 24 “chromosomes,” which form 6 tetrads. He had, in reality, 24 chromatids, and only one reducing divi- sion. McClung (’o00, ’o2a) has described the odd chromosome in the Acrididz and Locustide. He worked on a number of forms and obtained uniform results. In the Orthoptera, this chromosome can be traced back into the spermatogonial rest stages. It divides only in the first spermatocyte division, giving dimorphism of the spermatozoa. In 1901, McClung suggested the theory which has since that time received substantial corroboration, that the dimor- 474 Alice M. Boring phism of the spermatozoa corresponds to the dimorphism of sex. McClung considers that the longitudinal division always precedes the reducing division, and thinks that this 1s important on account of the failure of the second polar body to be extruded in parthe- nogenetic eggs; but the work in the other groups of insects shows that the reducing division probably comes first as often as the equational. Sutton’s careful work (’oo, ’02) on Brachystola magna offers convincing evidence for the individuality of the chromosomes. Each pair of spermatogonial chromosomes becomes enclosed in a separate compartment of the nucleus, while the odd chromosome is in a vesicle shut completely off from the others. He suggests the application of Montgomery’s theory of the union of maternal and paternal chromosomes in synapsis to Mendelian inherit- ance. The observations of de Sinéty (01) on the odd chromosome in one of the Acrididz and in several Phasmidz are entirely in accord with those of McClung; this chromosome divides in only one sper- matocyte division, producing dimorphic spermatozoa. In one of the phasms, he finds a chromosome complex similar to that de- scribed later by McClung (’05) for Hesperotettrix, where the odd chromosome attaches itself to one end of a tetrad, forming a hexad which divides along the transverse axis of the tetrad, thus sending the odd chromosome and two chromatids of the tetrad to one cell, and two to the other. Unfortunately de Sinéty interprets both of the spermatocyte divisions as longitudinal, but on this point he is in the minority among the workers on Orthoptera. Baumgartner (’04), in Gryllus domesticus, finds the odd chromo- some in a separate vesicle as Sutton did for Brachystola, but he finds it dividing in the second division instead of the first. Stevens (o5a) in Stenopelmatus and Blatella germanica, and Otte (’06), in Locusta viridissima, find that the odd chromosome divides in the second division instead of the first. Evidently there is no fixed rule as to where the odd chromosome shall divide. Voinov (’03), Montgomery (’05) and Zweiger (06) all hold a different view as to the valence of the orthopteran odd chromo- some; but as each has studied only one species of the order, while S permato genesis 475 the work of McClung, de Sinéty, Sutton, Baumgartner and Stevens covers numerous species in several families, we have a right to question the views of these other three observers. All three hold that the heterochromosome which they describe is formed from two spermatogonial chromosomes and divides in both spermato- cyte divisions. Moore and Robinson (’05) claim that the odd chromosome in Periplaneta americana is only a plasmosome which dissolves before each division and is reconstructed after it. Odonata The paper of McGill (04) on Anax junius seems to show the same confusion which Wilson has discovered in Paulmier’s work on Anasa tristis. McGill finds an even number of chromosomes in the spermatogonia, two of them small. ‘These she identifies with the chromatin nucleolus of the rest stage and the odd chromo- some, which divides in the first division and not in thesecond. If it could be shown that there are only 27 chromosomes in the sper- matogonial plate, and that the odd chromosome is one of the larger ones, this form would fall into line with other work. Lepidoptera The early investigators in this field, Platner (’86) and Verson (94) paid no attention to the chromosomes. I have not been able to read ‘Toyama’s papers, but the references to them by McClung indicate that the work is not very satisfactory. Stevens (o6b) gives a few figures for two species. ‘There are two con- densed bodies throughout the growth period, which fuse in pro- phase like the m-chromosomes in Alydus (Wilson, ’o5c), and this body divides in both divisions like the equal “idiochromosomes”’ of Nezara. Coleoptera The only work on the Coleoptera which deals with the hetero- chromosomes is that of Stevens (’o5b and ’o6b) and of Nowlin (06). Some of the beetles have an odd chromosome and others have an unequal pair in which the large member of the pair is the 476 Alice M. Boring maternal homologue of the odd chromosome, and the small mem- ber is the paternal mate which is lacking with the odd chromo- some. In the Coleoptera, the reducing division comes first, the equational second. In this order of insects there is substantial » proof of McClung’s sex determination theory, as the odgonial equatorial plates have been shown to have the large chromosome, while the spermatogonial plates have the small one, and there is the same difference between the somatic plates of the males and females. The theoretical bearing of these facts will be discussed later. Hemuptera The chromosomes in this group are so large and few in number that they have attracted many workers, but in spite of this fact, there have been greater discrepancies than in almost any other group. Henking (’90) in working on Pyrrhocoris apterus, was the first to notice that in one spermatocyte division, one chromo- some does not divide, thus causing a dimorphism of spermatozoa. He counted 24 chromosomes in the spermatogonia, and thought that this odd chromosome had the same valence as the others. He observed a large darkly-staining nucleolus in the growth period, although he did not associate a chromatic nature with it, or con- nect it with the odd chromosome of the spermatocyte mitoses. He formulated no theory to account for the dimorphism of the spermatozoa. Wilcox (’95) records that there are 12 spermatogonial chromo- somes in Cicada tibicens, and 24 spheroidal bodies in the sperma- tocytes, instead of a reduced number, results similar to those on Caloptenus femur-rubrum. In Anasa tristis, Paulmier (’99) describes two small sperma- togonial chromosomes, which form first the chromatin nucleolus in the growth period, then a tetrad which divides in the first sperma- tocyte division, and not in the second. Because this chromosome is small and appears in only part of the spermatozoa, he regards it as degenerating chromatin. Wilson (’05c), working over the same field, finds that Paulmier has confused two bodies, inas- much as the two small chromosomes form a tetrad and divide in S permatogenests 477 both divisions, while the odd chromosome, which divides only in the first division, is the chromatin nucleolus of the rest stage and one of the large chromosomes of the spermatogonia. He main- tains that Paulmier made a mistake also in the spermatogonial number, which is always odd. Foot and Strobell (’07), by the use of smear preparations and photo-micrographs, have attempted to show that Wilson is in error in his observations on the spermato- genesis of Anasa. ‘They find that the odd chromosome acts essen- tially like any other chromosome, is made up of two spermatogonial chromosomes and divides in both spermatocyte divisions, its only peculiarities being that it does not appear as a tetrad in prophase and occasionally divides later than the other chromosomes in metaphase. ‘They attempt to show that the chromatin nucleolus of the rest stage is not a chromosome, but dissolves before meta- phase like a plasmosome. Wilson (’07) has carefully gone over his preparations and still thinks that his former conclusions are correct. here is need of more work with smear preparations to test their reliability. Gross (’04), in his work on Syromastes, apparently confuses the m-chromosomes with the odd chromosome much as Paulmier did. In Pyrrhocoris apterus (’06) he finds the odd chromosome bival- ent but dividing in only one spermatocyte division. Montgomery (’ora) calls the odd chromosomes of the Hemip- tera “chromatin nucleoli” and considers that they may vary in number and valence. He explains them as chromosomes on the way to disappearance during progressive evolution. His results show many discrepancies which have since been explained by Wilson (’o5b and ’o5c). Wilson groups the Heteroptera into three classes, those with an unequal pair of heterochromosomes, those with an odd chromo- some and m-chromosomes, those with an equal pair of hetero- chromosomes. In the first class, the chromosome number in the second spermatocyte is one less than in the first spermatocyte. This is due to the fact that the conjugation of the unequal pair does not take place until after the first spermatocyte division. ‘This is the most direct evidence yet found for Montgomery’s synapsis hypothesis, for the small chromosome can be proved to be paternal, 478 Alice M. Boring and the large one, maternal. In the second class, the odd chromo- some is homologous with the large maternal element in the unequal pair. The m-chromosomes are a pair, whose synapsis is delayed until just before the first spermatocyte division. ‘The third class includes forms where there is neither an unequal pair, nor an odd chromosome, and therefore no visible dimorphism of the sperma- tozoa, but the fact that the equal heterochromosomes do not con- jugate until after the first spermatocyte division, relates this class to the first class, and suggests that there may be a masked dimor- phism, the equal heterochromosomes representing different char- acters, possibly, as truly as the unequal heterochromosomes where there is a visible dimorphism. Wilson cites a great deal of evi- dence for the individuality of the chromosomes, finding the same size relations between pairs of spermatogonial chromosomes as there are between single chromosomes in the spermatocytes. He elaborates McClung’s sex determination theory, brings forward much evidence for the dimorphism of the spermatozoa, and shows that there is a corresponding dimorphism in the somatic equatorial plates of the male and female of several species of the Hemiptera heteroptera. MATERIAL AND METHODS My material was collected at Woods Hole in the summer of 1905, at Cold Spring Harbor in the summer of 1906, and at Bryn Mawr in the fall of 1906. The insects were caught in the usual sweep net; and the testes dissected out as soon as possible. [ach testis consists of a group of several follicles, each attached by a separate duct to the vas deferens. ‘The testes from the larvae just ready for metamorphosis, and from the adults soon after meta- morphosis, in most cases give all stages from the spermatogonia to the mature spermatozoa. Before putting up material of any species, Schneider’s aceto- carmine proved to be a quick and efficient reagent for determining whether the testes contained all the important stages. This fixes and stains the material at the same time. ‘The testis is put on a slide in a drop of the stain, and the cells separated by press- ing down the coverglass. The preparation is made air-tight S permatogenesis 479 with vaseline, and in a few minutes, the chromatin is stained a deep carmine. ‘The entire spermatogenesis might be worked out in such preparations, the only disadvantage being that the achro- matic structures are not well fixed, and the preparations are not permanent. Camera drawings made from the aceto-carmine material, compared with those from sections of material fixed in the usual reagents, show the chromosomes in the former much larger in size. (Compare Fig. 198 with Fig. 205, and 201 with 207.) This difference is largely due to shrinkage in the usual fixing fluids and alcohols. The relative sizes and positions of the structures are the same in both kinds of preparations. If the material showed the right stages, it was put up in various fixing fluids: Gilson’s mercuro-nitric, Flemming’s strong chromo- aceto-osmic, Hermann’s platino-aceto-osmic, and Carnoy’s acetic alcohol with sublimate. ‘The dissecting was usually done in the fixing fluid, but the small quantity of material that was dissected in physiological salt solution and immediately transferred to the fixing fluid, showed just as good fixation, as is shown by the clear outlines of all the cell structures. A few cases of poor fixation were apparently due to the long time the insects were kept in captivity, as was sometimes necessary when the material was col- lected several miles from the laboratory, and immediate dissection was impossible. Gilson’s mercuro-nitric was the fixative used most frequently, because it gives excellent fixation of the chromatin and is a very convenient fluid to use, but nearly all material was also put in one or both of the osmic mixtures, as these give better fixation of the achromatic structures. “The Gilson was used for two to six hours, the Flemming and Hermann for twelve to twenty- four hours, followed by the same length of time in running water. The Carnoy was used but little. It does not fix so well as the Gil- son. Its real value is for material where an aqueous fixative can- not be used. After fixation, the material was run through the alcohols, cleared in xylol, and embedded in paraffine with a melting point of 52°C. Most of the sections were cut 5 » thick, a few 34 wand 63 mw. Many stains were tried. The three giving most satisfactory 480 Alice M. Boring results were Heidenhain’s iron hematoxylin, either without a counterstain, or with a slight tinge of orange G, thionin without a counterstain, and Auerbach’s combination of acid fuchsin and methyl green. With iron hematoxylin, the long method gave the best results. Preparations in this stain furnish the best outlines for camera drawings, but for work in spermatogenesis, there is the disadvantage that it often stains plasmosomes and chromosomes alike. Thionin has proved a valuable stain for distinguishing between chromatic material and plasmosomes. With this mate- rial the best results are gained by leaving the slides in the stain from one to five minutes, rinsing off with water, and differen- tiating under the microscope nh 95 per cent alcohol. The basichromatin holds the stain as a navy blue or dark purple, depend- ing upon the material; while the plasmosome and oxychromatin either take a very pale blue, or hold no color at all. “The Auer- bach stain also gives differentiation between basi and oxychro- matin, the odd chromosome standing out bright green in the rest stage against the pink spireme or scattered oxychromatin. OBSERVATIONS Membracide In the Membracidz, the testes are situated ventrally, near the anterior end of the abdomen. ‘They are white in color, and each follicle is round. Such ripe spermatozoa as are present are found near the duct and the spermatogonia are situated on the opposite side. The rest of the follicle is filled with the intermediate stages, grouped into cysts containing cells in about the same stage. The succession of these stages is rather difficult to follow in the Mem- bracidz, because the follicles are spherical and no one longitudinal section gives all of the stages. “The only way to trace the develop- ment is to find cysts with most of the cells in one stage and a few in transition to the next stage. In this way, the (ieee between the stages can be filled in. In the eight species from which my mate- rial was obtained, the general course of development is very simi- lar, with only here and there a striking difference. I shall there- S permatogenesis 481 fore describe in detail one species, Entilia sinuata, and then men- tion the chief points of interest in the other species. Entilia sinuata This form was found in September, at Woods Hole, on the leaves of the Golden Glow, and later near Philadelphia, on the wild sunflower. The resting spermatogonia stain very lightly, as there are only a few basichromatin granules in the midst of much scattered oxy- chromatin (Fig. 1). When the cell is preparing for division, a heavy, rather darkly-staining spireme is formed with the chroma- tin aggregated at regular intervals along the linin (Fig. 2). A longitudinal split appears in this spireme, a slight indication of which can be seen in Fig. 2. The chromatin next becomes con- densed and segmented, but these segments still retain their linin connections. ‘The longitudinal split in each segment is also very conspicuous at this stage (Fig. 3). Condensation of the segments continues, there being first an elimination of the longitudinal split (Fig. 4), and then a shortening of the segments until they are about twice as long as broad, the form which they have as they enter the equatorial plate of the spindle (Fig. 5). “They appear in the plate with their longitudinal axis at right angles to the longitudinal axis of the spindle and with the linin connections still intact. This division, therefore, is a longitudinal division, separating each chro- mosome into two parts along the line of the original longitudinal split, which appeared in prophase. A lateral view of the spindle in metakinesis also shows convincingly that this division is longi- tudinal (Fig. 6). The number of chromosomes in the spermato- gonial division is 21 but it is impossible to pick out the odd chromosome. ‘he chromosomes become so closely massed together in anaphase (Fig. 7) that one cannot tell whether the linin connections still remain intact, or the conjugation of chromo- some pairs takes place here. By the time the cell division is com- pleted, the new nuclear membrane has been formed, possibly as Conklin (’02) has suggested, by the joining together of the linin sheaths of the chromosomes after these have absorbed liquid from the cytoplasm (Fig. 8). A linin connection joining the chromo- 482 Alice M. Boring somes end to end is visible soon after they have lost their smooth contours (Fig. 9g). The last spermatogonial telophase is followed by a dense, darkly- staining contraction stage, which looks like a tightly wound spi- reme. Here the outlines of the chromosomes and their connections are entirely obliterated. “The contracted mass occupies only a part of the nucleus, leaving a large clear space at one side (Fig. 10). ‘This space appears in preparations where the fixation of other parts seems to be perfect, so it can hardly be looked upon as an artefact, as McClung (’oo) at first claimed. I have used Wilson’s (o5b) expression, “contraction stage”’ as simpler than McClung’s “synizesis,’’ for the most condensed period of “‘synapsis’ as Moore used the term. The chromatin now goes through a series of changes comparable to those of Anasa tristis (Wilson ’o5c): (1) an early postsynapsis, with a fine spireme, much twisted on itself, still staining deeply, but filling the nucleus much more com- pletely than in the contraction stage (Fig. 11); (2) a late post- synapsis, with the spireme filling the cell completely, less twisted, and staining unevenly (Fig. 12); (3) an early growth stage, with the spireme thicker, the basichromatin aggregated at regular intervals along the linin (Fig. 13); (4) a rest stage, where the spi- reme scarcely stains at all, and in the midst of the pale nucleus (in iron haematoxylin) there is one lens-shaped black body (Fig. 14), which, following Stevens, I shall call the odd chromosome. It is the “accessory of McClung, the “chromatin nucleolus” or hetero- chromosome” of Montgomery, the “chromosome spéciale” of de Sinéty, or the “heterotropic chromosome” of Wilson. From the action of similar bodies in related species, [ am convinced that it must be present here in the postsynapsis and early growth stages, but the spireme stains so deeply and twists on itself so much that it hides the odd chromosome. In the succeeding stage, where the spireme becomes longitudinally split, the odd chromosome length- ens out and loses the smoothness of its outline, although not the intensity of its staining reaction (Fig. 15). The spireme next divides into ten segments, each retaining its longitudinal split (Fig. 16). Counting the odd chromosome, which remains closely applied to the nuclear membrane, there are now 11 chromatic Spermatogenests 483 elements present in the nucleus. Just before the contraction stage, the spermatogonial chromosomes were joined end to end by linin connections, and out of the contraction stage there came a continuous spireme, which has passed through various stages and finally segmented. If the chromosomes conjugate end to end in the late anaphase (Fig. 8), as Fig. g might suggest, the longitudinal axis of the primary spermatocyte segments, or chromosomes, represents the longitudinal axis of the spermatogonial chromo- somes. [he presence of a massed anaphase and of the contrac- tion stage makes it impossible to prove that this is the case here. It has, however, been proved for other forms (Sutton) and the agreement of all other steps in the process points to a possible similarity in this respect also. The 10 segments next become tetrads by the formation of transverse arms which always remain a little shorter than the longitudinal arms, and thus make it always possible to distinguish between the longitudinal and transverse axes (Figs. 17to 19 ). While the tetrads and dumb-bells are form- ing, the odd chromosome rounds up again and becomes a lens- shaped body, still applied to the nuclear membrane (Fig. 20). It is in the dumb-bell form that the chromosomes usually enter the spindle (Fig. 24), but occasionally they are still in the form of cross- shaped tetrads (Fig. 22). “This shows conclusively that the longi- tudinal axis of the dum>-bell is the same as the longitudinal axis of the tetrad, and that the first spermatocyte mitosis 1s a transverse division. ‘That it is probably a reducing division can be shown by tracing back the development, and working out the corresponding axes: the division between the halves of the dumb-bell (Fig. 24) corresponds to a division along the lateral arms of the tetrad (Fig.. 17), and that to a transverse section of the spireme segment (Fig. 16) and that to the separation of one spermatogonial chromosome from another, if we assume that each spireme segment equals two spermatogonial chromosomes joined end to end. ‘This may be further evidence against McClung’s (’00) contention that the reducing division is always the second. In the equatorial plate of the first spermatocytes the odd chromosome stands a little apart from the other 10 chromosomes, and is smaller in diameter (Fig. 21). It does not divide in the first spermatocyte division, but lags 484 Alice M. Boring behind the others in going toward the spindle pole (Figs. 25 and 27). [he chromosomes mass together in the anaphase, so that as soon as the odd chromosome joins the others, it is no longer possible to distinguish it (Fig. 28). The spindle fibers stand out very clearly, especially in the mate- rial fixed in Flemming or Hermann, and it is noticeable that the odd univalent chromosome is joined to only one pole by its mantle fibers, while the bivalent chromosomes are attached to both. During the telophase the granules of a “Zwischenkorper” can be seen on some (Fig. 25) or all (Fig. 26) the spindle fibers. “These show only in iron hematoxylin preparations which have not been extracted very thoroughly. In such preparations the centrosomes of the first spermatocyte division can also be seen (Fig. 23). “They divide during the anaphase of the first division (Figs. 25 and 27) in readiness for the second division which succeeds the first without - any reconstruction of the nucleus. The chromosomes rearrange themselves (Fig. 29) into a plane at right angles to the plane of the first division, and soon form a regular equatorial plate. Half of the second spermatocytes con- tain 10 chromosomes (Fig. 31) and the other half 11 (Fig. 30), that is, 10 plus the odd chromosome. In the cells containing 11 chro- mosomes, the odd one does not differ enough in size to make it any longer distinguishable. In this division, all the chromosomes in all of the cells divide. The reasons for this conclusion are: (1) the lateral views of the metaphase (Fig. 32) never show one undi- vided chromosome among the other dividing ones, (2) all the chro- mosomes are attached by mantle fibers to both spindle poles, and (3) in the anaphase, there is never a lagging chromosome near one pole without a mate at the other pole (Fig. 33). “That this division of chromosomes is at right angles to the first, thatis, longitudinal and equational, is certainly conditioned by the formation of the spindle which is derived directly from that of the first division. The same fibers between the chromosomes and centrosomes remain intact, and as the centrosome divides, the chromosomes are pulled into an equatorial plate at right angles to the equatorial plate of the first spermatocyte division. ‘This second division therefore corresponds to the preliminary longitudinal splitting of S permatogenests 485 the spireme in the growth period. One spermatocyte division is reducing and the other equational. In the anaphase, the chromo- somes again become massed together (Fig. 34) and the nucleus is reconstructed by the formation of a nuclear membrane (Fig. 35). The “Zwischenk6rper” is again noticeable in this telophase. In the young spermatid (Fig. 36), the chromatin is still massed together and stains deeply. ‘The spindle material remains as the “Nebenkern,” as first described by v. La Valette St. George (’86) for insect spermatids. ‘The chromatin soon scatters through the nucleus in definite clumps and it is evident that half of the sperma- tids contain a smooth round darkly-staining body (Fig. 37), while the other half do not (Fig. 38). Through several succeeding stages, this same fact is noticeable; 7. e., when the chromatin becomes more diffuse (Figs. 39 and 40), when it forms a pale net- work and the axial filament has grown out (Figs. 41 and 42), and even when the chromatin has begun to condense to form the head of the spermatozoon (Figs. 43 and 44). The method of deter- mining whether this body is in only half the cells or in all is as follows: cysts of spermatids in various places were picked out and the number of cells with and without this body were counted in each cyst. In studying sections, it must be remembered that parts of some cells are in another section, so even if this body (x) were actually present in all the cells, it would not appear in all in any one section of a cyst. On the same principle, if it were actually in only half the cells, it would appear in less than half in any one section. In Entilia, this body appears in a few less than half of the sperma- tids. It always takes the chromatin stains, deep blue with thionin, and green with the Auerbach. As it resembles the odd chromo- some of the first spermatocyte rest stages in staining reaction and contour, and as it appears in not more than one-half of the sperma- tids, a condition which the odd chromosome necessarily fulfills from the fact of its not dividing in the first spermatocyte division, we seem to be justified in concluding that the body x of the sperma- tids is a derivative of the odd chromosome of the spermatocyte. There is nothing unusual about the formation of the spermatozoon. The “ Nebenkern”’ forms the sheath of the axial filament (Fig. 41), the acrosome differentiates from the cytoplasm at the apex of the 486 Alice M. Boring head, the head forms by condensation of the chromatin (Figs. 44 to 47), passing through one rather diffuse stage (Fig. 46). Vanduzea arcuata Vanduzea arcuata was found in abundance on the locust trees near Cold Spring Harbor in June. The spermatogonial plates show 17 chromosomes, varying 1n size (Fig. 48). It 1s not possible to arrange them all in pairs, but at least two large pairs are well marked (a, and a,, 6, and b,). In the growth stage, the odd chro- mosome appears as a long, darkly-staining body, without a smooth contour. It is at first bent upon itself in different forms (Fig. 49), and later lies at full length along the nuclear membrane (Fig. 50), resembling the same stage in Entilia sinuata (Fig. 15). In the equatorial plate of the first spermatocyte division, there are 9 chro- mosomes, two of which are larger than the others (Fig. 51, a and b), corresponding to the four large ones in the spermatogonial plate; ais slightly larger than db just as a, and a, were slightly larger than 6, and b,. This point certainly counts as evidence that each spermatocyte chromosome represents not an indefinite segment of the spireme, but two individual spermatogonial chromosomes. The odd chromosome can be recognized by 1 its eccentric position. Fig. 52 shows all the chromosomes but » in metakinesis, and in Fis. 53% is passing to one pole undivided. Figs. 54 and 55 show variations in the position of x in anaphase; it does not Re lag behind, but may even precede the other chromosomes to the pales The second spermatocyte equatorial plates, containing g and 8 chromosomes, respectively, are shown in Figs. 56 and57. Each has one large chromosome a, one not quite so aoe: b, and six small ones of about the same size. Fig. 56 has a ninth chromosome of intermediate size which must be the odd chromosome, as x in the first spermatocyte plate has a corresponding intermediate size (Fig. 51). All the chromosomes divide in this division, including the odd one, as is shown in all of the lateral views of the metaphase (Fig. 58) and of the anaphase (Fig. 59). Half of the spermatids contain the odd chromosome, and half do not (Figs. 60 and 61). S permatogenests 487 Ceresa taurina. Three species of Ceresa were found near Cold Spring Harbor on the morning-glory vines and tall weeds, during the last three weeks of July. Unfortunately the chromosomes of the spermato- gonial plates in all three forms are too close together to make it Spee to count them. They all have the same reduced number of chromosomes and a peculiar deposition of chromatin on the nuclear membrane in the growth period. As this phenomenon is most pronounced in Ceresa taurina, | shall give the details for this form. In the contraction stage, the chromatin is massed at one side of the nucleus in a number of darkly-staining loops with their bases united in a dense flat chromatic plate, which stains more deeply than the loops (Fig. 62). As the loops spread through the nucleus, they stain less, making the contrast with the black plate more intense (Fig. 63). In the rest stage (Figs. 64 to 67), the reticulum does not take basic stains at all; the chromatin plate appears in various forms, sometimes continuous and sometimes broken up into two, three, or four parts. By the time a split spi- reme is formed, it has been almost entirely dissolved (Fig. 68), and in the prophases, no trace of it is left (Fig. 69). When these masses dissolve, the odd chromosome becomes visible as a round, smooth body (Figs. 67 and 68), which probably was concealed in the midst of the chromatic plate as far back as the contraction stage, but its presence was obscured by the similarity of its staining reac- tion to that of the other chromatin. As to the meaning of this deposition of chromatin on the nuclear membrane, it seems possi- ble that it is basichromatin thrown out from the chromosome loops in the contraction stage, and that it takes no part in the fur- ther formation of the chromosomes, since it disappears before the next division. The only case at all similar which I can find in the literature is that of Gryllus campestris described by Voinov (’04). He claims that all the chromatin is gathered into the “corps nucle- inien double,” leaving the non-stainable achromatic substance spread through the nucleus, and that when the spireme forms, the chromatin is added to it again from this structure. He neglects the distinction between oxy and basichromatin, and thinks that when all 488 Alice M. Boring the stainable chromatin 1s aggregated into one body, there is no chro- matin left elsewhere. ‘The situation is much clearer if looked at from Conklin’s point of view (’02): although the nucleus in the rest stage does not take basic stains, it still contains chromatin in the form of oxychromatin; this has the power of changing into basichromatin to form the chromosomes for division. ‘The basi- chromatin masses of the rest stage, with the exception of the odd chromosome, which here again shows its individuality by a differ- ence in behavior, are apparently rejected substances, which dis- appear without playing any further role in karyokinesis. In the prophase, the odd chromosome lies close to the nuclear membrane as in the forms previously studied, and in the metaphase it has a somewhat eccentric position (Fig. 70). The chromosomes here are so nearly of the same size that it is impossible to trace any individuals from cell to cell; but the odd chromosome, by virtue of its position and its univalence, can be followed until the second spermatocytes are formed. Figs. 71 to 73 show its varying behav- ior in metaphase; it may either follow or precede the other chromo- somes to the pole. ‘This fact is shown also by the two anaphase figures, 74 and 75. The second spermatocyte equatorial plates show the two numbers of chromosomes I1 and Io (Figs. 76 and 77), but the odd chromosome can no longer be distinguished from the others, either in metaphase (Fig. 78) or anaphase (Fig. 79). In all the spermatids (Fig. 80), there appears one large body (7) taking the basic stains, probably analogous to the body in the beetle sper- matids called a chromatin nucleolus by Stevens (’o6b). It is impossible to decide whether the odd chromosome in half the sper- matids keeps its individuality as was observed in Entilia and Van- duzea, for all the chromatin stains deeply and in some stages is broken up into many separate masses (Fig. 80). Ceresa bubalus The only external difference between this species and the fore- going one Is its greater size and the different angle of the prothor- acic protuberances. The only difference in the spermatogenesis as can be seen by Figs. 81 to g2, is that the mass of rejected chro- matin 1s not so conspicuous. In the bouquet stage (Fig. 81), the S permatogenesis 489 plate is not nearly so large as in the same stage of Ceresa taurina (Fig. 63). Fig. 82 represents one of the most extreme cases of the growth stage. Ceresa diceros The shape and size of this species is about the same as in Ceresa bubalus, but the coloring is different, being brown and white, instead of uniform green. ‘The spermatogenesis is practically the same, as Figs. g3 to 101 show, but a preparation from the testis of one could be distinguished from a preparation of the other, because the cells, chromosomes, and spindles of C. diceros are always smaller than those in C. bubalus. Atymna castanea This species was found on the chestnut trees exclusively, and was very abundant at the end of June and beginning of July. No spermatogonial plates in which the number of chromosomes could be counted were found. ‘The odd chromosome appears in the rest stage as a large round body with a smooth contour and an afhnity for basic stains (Fig. 102). In lateral view of the metaphase of the first spermatocyte division, it is apparent that it does not divide (Figs. 104 and 105), and in the anaphase it has the position usually characteristic of this order, between the plates of chromosomes, but nearer one pole than the other (Fig. 106). “The number of chro- mosomes in the first spermatocyte is again 11 (Fig. 103), two of them constantly larger than the others (a and). ‘These two large chromosomes appear in all the second spermatocyte plates, whether they have 11 or 10 chromosomes (Figs. 107 and 108). All the spermatids contain a chromatin nucleolus (Fig. 111), as in the genus Ceresa. There being apparently no other basic-staining body in any of the spermatids, the odd chromosome in half of them must take part in the formation of the general reticulum like the other chromosomes. Campylenchia curvata Campylenchia curvata was found in sweepings from various weeds throughout July. The material showed all desirable stages. 490 Alice M. Boring Many spermatogonial plates were found, some of which it was possible to count. It seems that there must be one short period in the arrangement of the chromosomes into the plate, when they are spread further apart than at any other tme. Judging from the behavior of the chromosomes of the first spermatocyte in coming into the equatorial plate, this more open stage must occur when the chromosomes are first drawn into a flat plate from their scat- tered position in prophase. Later as metakinesis begins and the mantle fibers pull from the two poles, the chromosomes are drawn closer together and the diameter of the plate becomes smaller. Fig. 112 shows a very clear spermatogonial plate, with 19 chromo- somes. It is possible here to group the chromosomes into 9g pairs with one left over; only the two most distinct pairs are lettered, a, and a,, long and slender, 5, and 5,, a little shorter and thicker. The two chromosomes formed by the fusion of these pairs are designated by a and d in Fig. 114, the equatorial plate of the first spermatocyte, and in Figs. 117 and 118, the equatorial plates of the second spermatocytes. “he number of chromosomes in the equatorial plates are what would be expected after finding 19 in the spermatogonia; 10 in the first spermatocytes, and 10 andq, respectively, in the second. In the rest stages (Fig. 113), x ap- pears as usual, but there are also present two other smaller bodies with the same staining reaction, m, and m.. I have called them m-chromosomes, as they have all the characteristics of Wilson’s m-chromosomes in the rest stage of the Hemiptera Heteroptera (o5c); they are of equal size and they take the basic stains like the odd chromosome. As unfortunately they are not enough smaller than some of the other chromosomes to be readily distinguished in the spermatogonial plate, or to be traced through the prophase of the first spermatocyte to the spindle, it is impossible to see whether they really represent one pair whose fusion has been delayed. The odd chromosome appears as usual in metaphase (Fig. 115) and anaphase (Fig. 116) of the first spermatocyte division, and as usual is not distinguishable in the metaphase (Fig. 119) or ana- phase (Fig. 120) of the second division. In the spermatids, a basic-staining body appears in half the nuclei (Figs. 121 and 122), and so must here (as in Entilia and Vanduzea) represent the odd S permatogenesis 491 chromosome, rather than the chromatin nucleolus of the other Membracidz studied. Enchenopa binotata Enchenopa binotata was found throughout July at Cold Spring Harbor on the locust and wild cherry trees, on blackberry bushes and sometimes in general sweepings of weeds. Its spermato- genesis has been the most puzzling of any form studied and the following account is given tentatively, with the intention of going over the work as soon as more material can be obtained. The first facts to be noticed are that all the chromosomes appear as dumb-bells in the metakinesis of the first spermatocyte (Fig. 128), there is no lagging chromosome in the anaphase (Fig. 130), and all the second spermatocytes have 10 chromosomes (Fig. 131), the same number as the first spermatocytes. In iron haematoxylin preparations extracted to the same degree as in other material, no darkly-staining body appears in the rest stage, but in those extracted for a shorter time, a long twisted body appears against the pale spireme (Fig. 124). This can occasionally be traced into a stage where the spireme has segmented (Fig. 125), but never any further, as it does not assume a compact rounded shape until the other chromosomes become condensed. ‘The question arises as to whether this body in the growth stage represents two sper- matogonial chromosomes and consequently divides in both sper- matocyte divisions as all bivalent chromosomes do; or whether it is univalent, analogous to most odd chromosomes in insects, but divides in the first spermatocyte division and not in the second, thus differing from all the other Hemiptera Homoptera studied and resembling most of the Heteroptera. ‘There were a few sper- matogonial plates in such a stage that it was possible to count the chromosomes, but these did not have the chromosomes as clearly spread apart as in most the other species studied. In five plates, 19 chromosomes were counted (Fig. 123) and in two, 20. One of those with 20 may, however, be deceptive; two of the chro- mosomes are much smaller than any in the other plates, the plate is at the surface of the section, and as x in Fig. 1231s V-shaped, it is possible that the bend of the V was cut off and the two small chro- 492 Alice M. Boring mosomes may really be but one. Other evidence for the univa- lence of one chromosome is its occasional appearance in early metaphase of the first spermatocytes when it has not yet assumed the dumb-bell shape (Fig. 129), and a few second spermatocyte metaphases where it apparently does not divide (Fig. 133). If it does not divide in the second spermatocyte division, the second spermatocyte spindle should always appear as it does in Fig. 133 rather than as in Fig. 132, unless the odd chromosome is usually in the center surrounded by the other chromosomes. ‘That this probably is true is indicated by several cases like Fig. 135, the two anaphase groups of one second spermatocyte spindle, a having 9 chromosomes and } 10. ‘There is a space in a corresponding to the chromosome marked x in 6. This evidence is anything but satisfactory, but the possibility of such an exception to the general rule that the odd chromosome divides in the first spermatocyte division, is too interesting a fact to leave unmentioned. Here again one large chromosome in the first spermatocyte (Fig. 126) is represented by two in the spermatogonia (Fig. 123, a, and a,), and by one in the second spermatocyte (Fig.131,a). Fig. 127 shows an occasional first spermatocyte with 11 chromosomes, implying a delay in the fusion of one pair. Here we find the chromatin nucleolus in all the spermatids (Fig. 136). Fasside The testes of the Jasside are pale yellow in color, and there- fore very easy to dissect out. ‘The follicles are about three times as long as broad; this makes it easier to trace the development from stage to stage than in the Membracide. My material includes six species, four of them caught at Cold Spring Harbor in July, and the other two, Agallia sanguinolenta and Phlepsius irrotatus, at Bryn Mawr in October. Chlorotettrix unicolor and C. vividus This material was fixed and preserved as belonging to one species, but study of the sections showed two different reduced numbers of chromosomes, 11 and 9g. ‘This led to a careful com- S permatogenesis 493 parison of my specimens with those in the collection at the Acad- emy of Natural Sciences, Philadelphia. “There proved to be two species, C. unicolor and C. vividus, in which the only marked difference is the width of head and thorax. Some of my specimens are slightly narrower than others, so I have probably mixed the two species, and cannot state whether the g chromosomes belong to C. unicolor or to C. vividus. The resting spermatogonium has a reticulum of oxychromatin and linin and a plasmosome, which stains black in iron hematox- ylin, but shows its achromatic nature in thionin (Fig. 137). There were no good spermatogonial plates in the material with the smaller number of chromosomes, but a lateral view of the spindle is shown in Fig. 138, and the anaphase in Fig. 139. ‘The chromatin then passes into a contraction stage which is very dense, but contains several clear vacuoles (Fig. 140). This has a very different appearance from the contraction stage of the Membracide. A spireme stage follows where the chromatin again fills the nucleus and still stains deeply (Fig. 141). The odd chromosome is first visible in the rest stage (Fig. 142) where the chromatin stains least and is most scattered. It is closely applied to the nuclear membrane as was usually the case among the Membracidz. ‘The spireme splits longitudinally (Fig. 143), and then becomes seg- mented (Fig. 144). In all stages the odd chromosome can be distinguished by its small size. In the prophase of the first sper- matocyte division, it can be recognized by its rounded contour; in the equatorial plate, by its eccentric position (Fig. 145); in the lateral view of the metaphase (Fig. 146), by its undivided condi- tion; and in anaphase, by its lagging behind at one pole of the spindle (Fig. 147). In the equatorial plates of the second sper- matocytes with g chromosomes, it can still be recognized by its small size (Fig. 149). As it divides in the second spermatocyte division, there is no indication of it in a lateral view of the meta- phase (Fig. 150), or anaphase (Fig. 151). “Two of the g chromo- somes are larger than the others (a and /in Fig. 145), and they keep their individuality in the second spermatocyte (a and 6 in Figs. 148 and 149). In all the spermatids, there is one condensed body, which resembles the body called a chromatin nucleolus in 404 Alice M. Boring five species of the Membracida. In the early spermatid, this is the only condensed body distinguishable (Fig. 152), but later when the chromatin becomes more diffuse, it appears that half the sper- matids have another smaller condensed body (Figs. 153 and 154), which is lacking in the other half. “This must be the odd chromo- some, observed 1n the same stages of three species of Membracide. In a still later stage, when the reticulum is arranged around a series of clear vacuoles, this difference is still to be observed; all the cells have the one large body, but only half have the small chromosome (Figs. 155 and 156). After this, both bodies disappear, the chro- matin reticulum becomes slightly more condensed at first (Fig. 157), the nucleus then elongates but keeps the vacuoles (Fig. 158), and finally condenses into the head of the spermatozoon (Fig. 159). The acrosome is differentiated from cytoplasm at the apex of the head. Fig. 160 is the spermatogonial plate of the species with the larger number of chromosomes. It contains 21 chromosomes, four larger than the others, not differing conspicuously in size among themselves (a,, a, 6, b,). The first spermatocyte equa- torial plate has 11 chromosomes, and they show the same size relation as those of the other species, two large ones and one small odd chromosome in an eccentric position (Fig. 161). This plate simply has two more chromosomes of intermediate size than the other. [he second spermatocyte plates again show the two large chromosomes (Figs. 162 and 163), the total numbers being 11 and 10, instead of g and 8. Diedrocephala coccinea A few scattered individuals were found in July in general sweep- ings, but in August an abundance of material was obtained from the blackberry vines. The spermatogonial plates show 23 chro- mosomes, two larger than the others (a, and a, in Fig. 164). In the postsynapsis stage, the odd chromosome is not surrounded by the spireme, as has been the case in the forms described above, but it stands out distinctly by itself in the clear part of the nucleus (Fig. 165). In the rest stage, it is still of the same size and in the same position, although the nucleus grows much larger and the S permatogenests 495 chromatin becomes scattered and diffuse (Fig. 166). The first spermatocyte shows the odd chromosome as a medium-sized body, eccentric in the plate of 12 chromosomes (Fig. 167), and not divid- ing in metakinesis (Fig. 168). In anaphase, it lags behind the others (Fig. 169). The two large chromosomes of the spermato- gonia have fused into a single large one in the first spermatocyte (a in Fig. 167), and this keeps its individuality in the second sper- matocytes (a in Figs. 170 and 171). Half the second spermato- cytes have 12 chromosomes, and half 11. The spermatids all have the chromatin nucleolus, and half of them the odd chromo- some (Figs. 174 and 175), as in Chlorotettrix. Diedrocephala mollipes This species resembles Diedrocephala coccinea in shape, but not in color, being bright green instead of red and green striped. Its spermatogenesis is also similar (Figs. 176 to 185), but the cells and chromosomes are smaller (cf. Fig. 177 and 167). They both have the same number of chromosomes, 12, but Diedrocephala mollipes has no one chromosome markedly larger than the others. The spermatids have both a chromatin nucleolus and an odd chromosome. Phlepsius irrotatus The spermatogonial plate contains 15 chromosomes, two larger than the others (a, and a,, Fig. 186). ‘These are represented by a in the first spermatocyte (Fig. 188a) and also in the second sperma- tocytes (Figs. 191 and 192, a). The growth period shows the odd chromosome (x) as a round body with even contour (Fig. 187). The univalent chromosome x has the peculiarity here that it never comes to lie in a flat plate with the other chromosomes in the first spermatocyte division, as is indicated in Fig. 189. ‘To get all 8 chromosomes, the equatorial plate must be drawn at two different foci (Figs. 188aand 188b). ‘The odd chromosome always precedes the others to the pole (Fig. 190), never taking the lagging position characteristic of the species previously described. We have noted that this sometimes takes place in other forms (Vanduzea arcuata, and the three species of Ceresa), but Phlepsius is the first form 496 Alice M. Boring where this position is invariable. ‘The second spermatocytes con- tain 8 and 7 chromosomes (Figs. 1g1 and 192). ‘The spermatids all contain the chromatin nucleolus (Figs. 195, 7, and 196, m) and half of them, an odd chromosome (Figs. 195 and 196, «). Agallia sanguinolenta No spermatogonial plates were found in this form. ‘The odd chromosome appears as usual in the growth period (Fig. 197). There are 11 chromosomes in the first spermatocyte (Fig. 198), and 11 and 1o in the second (Figs. 200 and 201). ‘The odd chro- mosome does not divide in the first spermatocyte metakinesis (Fig. 199), but passes to one pole after the other chromosomes in anaphase (Fig. 206). The spermatids all contain a chromatin nucleolus, and half of them, the odd chromosome (Figs. 203 and 204). Figs. 205 to 207 are drawn from aceto-carmine preparations at the same magnification as Figs. 197 to 204. Cercopide The testes of the Cercopidz are situated near the posterior end of the abdomen. ‘They are white in color, and each follicle is round, with a comparatively long duct joining it to the vas deferens. The material comprises four species, and the spermatogenesis of none of them resembles very closely that of the species studied by Stevens (’o6b). Clastoptera obtusa Thrs species was found on the alder at Cold Spring Harbor. ‘The resting spermatogonium stains very lightly and has a plasmo- some (Fig. 208). In preparing for division, the chromatin forms a spireme, which becomes more dense, and then segments (Fig. 209). There are 15 chromosomes in the spermatogonial equa- torial platesall of about the same size (Fig. 210). “The division is longitudinal as usual (Figs. 211 and 212). After the telophase, the chromosomes soon become joined by linin connections (Fig. 213), form a compact spireme in early synapsis (Fig. 214), a dense mass in the contraction stage (Fig. 215) and a spireme loosely wound on itself in postsynapsis (Fig. 216). “The odd chromosome S permatogenests 497 appears inthe contraction stage distinct from the dense chromatin mass, and remains so in postsynapsis and the early growth stage (Fig. 217). It is from the first, a small, ovoid, smooth-contoured body, and still shows clearly when the spireme has segmented and the tetrads are forming (Fig. 218), and when the dumb-bells are formed (Fig. 219). It takes an eccentric position in the equatorial plate of the first spermatocyte (Fig. 220). It does not divide in the first spermatocyte division (Fig. 221), and is the last chromo- some to reach the pole in the anaphase (Fig. 222). As there are 7 chromosomes, plus the odd one, in the first spermatocyte, so there are 8 in half the second spermatocytes (Fig. 223), and 7 in the others (Fig. 224). [he odd chromosome behaves like the others in the second division (Figs. 225 and 226), and is not distinguish- able in the spermatids, all of which have a chromatin nucleolus (Fig. 227). In the development of the spermatid, the chromatin reticulum first becomes massed on the side of the nucleus toward the axial filament (Fig. 228), and then forms a dense U, leaving the rest of the nucleus clear (Fig. 229). The nucleus then elon- gates, still leaving a clear space toward the apex (Fig. 230). The mature spermatozoon has a solid dense chromatic head (Fig. 231). Aphrophora quadrangularis This species was found on the grass and low bushes in July near Cold Spring Harbor. Originally a small quantity of material was collected and tried in aceto-carmine, as it was supposed to be the same species that Stevens (’06b) had found in Maine and described. But the reduced number of chromosomes proved to be 11 instead of 12, so material was fixed in Gilson and kept to be studied at a convenient time. [he material was obtained from two distinct localities, but not kept separate. ‘The sections showed follicles with 11 chromosomes and a few with 12. Whether this difference corresponds with the difference in locality it is unfortunately not possible to say. Another peculiarity is that the form with 12 chro- mosomes does not resemble, in some of its stages, the form with 12 chromosomes described by Stevens. ‘The most important stages of the form with 11 chromosomes are shown in Figs. 232 to 242. There are 21 spermatogonial chromosomes (Fig. 232) and 498 : Alice M. Boring 11 and 10 second spermatocyte chromosomes (Figs. 238 and 239). The odd chromosome can be traced as an individual as far back as the contraction stage (Figs. 233, x,and234,x). Aplasmosome (p) also appears in the growth period, the thionin clearly bringing out the difference between the two. One of the 11 chromosomes is larger than the others, as is shown in Figs. 235, 238, 239. [he odd chromosome does not divide in the first spermatocyte division (Figs.236 and 237). The spermatids all contain a chromatin nucle- olus (Fig. 242). A few stages of an individual with 12 chromo- somes are shown in Figs. 243 to 248. ‘This series much more nearly resembles that of the other form from Cold Spring Harbor with 11 chromosomes, than that of the form found in Maine with 12 chro- mosomes. [he Maine form has no contraction stage (Stevens ~ ’o6b, Figs. 240 to 249), while this form has a distinct one with the odd chromosome and the plasmosome outside of the spireme in the clear part of the nucleus (Fig. 243). The only possible con- clusion seems to be that three species (so determined by the differ- ences in spermatogenesis) have been up to this time grouped as one, and all called Aphrophora quadrangularis. Aphrophora 4-notata Aphrophora 4-notata is interesting especially in connection with Aphrophora quadrangularis, as being another case of differ- ence of chromosome number within the same genus. Aphrophora 4-notata has 14 chromosomes for the reduced number (Fig. 250) and consequently 14 in half of the second spermatocytes (Fig. 253) and 13 in the other half (Fig. 254). The odd chromosome is pres- ent in the spireme stage (Fig. 249), and does not divide in the first spermatocyte division (Figs. 251 and 252). Fulgoride The testes of the Fulgoride are orange-colored and show through the thin white walls of the abdomen. ‘The separate folli- cles are oblong. Of the four species in my material, three belong to the genus Peeciloptera, and one to the Amphiscepa, but according to the spermatogenesis, P. bivittata is much more like the Amphis- S permatogenesis 499 cepa than like the other two species of Peeciloptera. P. septen- trionalis and P. pruinosa were found on the nettle and the other two species came from sweeping low grasses. In this material, the cells and chromosomes are large and the achromatic struc- tures especially well preserved. The material fixed in Flemming, and stained in thionin makes some of the clearest preparations included in this study. Poeciloptera septentrionalis The resting spermatogonia of this form are small and stain lightly (Fig. 256). In preparation for division, a spireme is formed, each granule of which splits longitudinally (Fig. 257). The chromatic part of the spireme segments, retaining the linin connections and also an indication of the longitudinal split (Fig. 258). [here are 27 chromosomes in the spermatogonial plate, two longer than the others (a, and a, of Fig. 259). Fig. 260 shows distinctly that this division follows the preliminary longitudinal split. After the telophase, the chromosomes become more diffuse and join into a spireme (Fig. 262). “This spireme contracts into a small dense ball at one side of the nucleus (Fig. 263), and then the cell goes through a long growth period in which the diameter is at least doubled. ‘The odd chromosome appears as soon as the spireme becomes pale enough to conceal it no longer (Fig. 264). Then a pair of m-chromosomes appears and a small plasmosome (Fig. 265). [he plasmosome and odd chromosome both increase in size, the latter having a vacuole in the center (Fig. 266). The odd chromosome has now attained its full size, but while the cell and nucleus continue to increase, the plasmosome keeps on grow- ing (Fig. 267). Even though it is now larger than the odd chromo- some, it stains scarcely at all, while the odd chromosome and the m-chromosomes stain a deep blue, thus demonstrating the valu- able differentiating powers of thionin. In the next stage (Fig. 268) the odd chromosome and the plasmosome are unchanged, but the spireme stains more deeply and shows a longitudinal split. The m- chromosomes no longer appear, they have probably become indis- tinguishable from the other spireme segments. “The plasmosome and odd chromosome still keep the same relative size in the pro- 500 Alice M. Boring phase, while the tetrads are forming (Fig. 269), the plasmosome sometimes not being dissolved until after the spindle is formed (Fig. 271). There are 14 chromosomes in the equatorial plate of the first spermatocyte (Fig. 270), one of them being marked by its eccentric position, another by its large size. This large chromo- some keeps its individuality in all the second spermatocytes, those with 14 chromosomes (Fig. 273), and those with 13 (Fig. 274). The odd chromosome does not divide in the first spermatocyte division (Figs. 271 and 272), but does in the second (Figs. 275 and 276). ‘The development of the spermatid in this family is very peculiar. ‘The nucleus stains quite deeply, so that nothing more can be made out than that there seems to be one condensed body in each spermatid (Fig. 27ga). The “‘Nebenkern” goes through a complicated development somewhat similar at first to that de- scribed by Baumgartner (02). First delicate fibers are formed in it (Fig. 277), then it appears as a long coiled fiber in a clear space, surrounded by a definite membrane (Fig. 278). ‘This space becomes separated by a partition into two tubes, each containing several shorter fibers (Figs.27ga and b). ‘These tubes and fibers both become elongated (Fig. 280). ‘The tubes grow still longer and smaller in diameter, and at the same time twist around each other in an irregular spiral (Fig. 281a). Cross sections through different portions of these twisted tubes indicate that they must also be constricted in places (Fig. 281b). They finally become flattened, presenting some such an appearance as in Fig. 282a, and in cross section as in Fig. 282b. In this species, the chromo- somes in the female somatic cells could be counted, and proved to be 28 in number (Fig. 283), there being the same two long ones that appeared in the spermatogonial plate. ‘The significance of the even number in the female, and the odd number in the male will be pointed out in the theoretical considerations. Peeciloptera pruinosa Peeciloptera pruinosa resembles the last described form exter- nally in every character but color, being a grayish purple instead ofa pale green. ‘The principal stages are shown in Figs. 284 to 293, the only difference being that there are two large chromosomes S permatogenesis 501 instead of one, in the first spermatocyte equatorial plate (Fig. 285) and also in the second spermatocyte plates (Figs. 288 and 289). The chromatin in the spermatid nucleus does not stain so deeply, and here it can be demonstrated that there is achromatin nucleus in all of the spermatids (Figs. 292 and 293), and the odd chromosome besides in half of them (Fig. 292). Here also the female somatic chromosome number is 28. Fig. 294 shows some of the chromo- somes overlapping each other, but they are really entirely separate from one another, lying at slightly different levels; it isa late pro- phase stage of an egg follicle cell before the chromosomes are drawn completely into one plane. Amphiscepa bivittata All this material came from larve. The different stages are shown in Figs. 295 to 304. The spermatogonial plates contain 25 chromosomes, two pairs of long ones, one pair longer than the other (Fig. 295). In the rest stage, there are no m-chromosomes, but two plasmosomes are present (Fig. 296). “The first spermato- cyte plate shows two large chromosomes, one larger than the other (Fig. 297), corresponding to the two large pairs of the spermato- gonium. ‘he plasmosomes here persist into the metaphase (Fig. 298). ‘The odd chromosome is quite small (Figs. 297, x, 298, x; 299, x) and does not divide in the first division. Chromosomes a and b of the first spermatocyte retain their relative sizes in the second spermatocytes, both those containing 13 chromosomes (Fig. 300), and those with 12 (Fig. 301). Peeciloptera bivittata Peeciloptera bivittata very closely resembles the last described species, even to the number and relative sizes of its chromosomes (Figs. 305-313). It has two plasmosomes in the growth period, and one or both of these persist in a most remarkable fashion even to the anaphase of the second spermatocyte division (Fig. 312). The size of the chromosomes and cells is greater than in Amphis- cepa bivittata. 502 Alice M. Boring THEORETICAL CONSIDERATIONS Individuality of the Chromosomes The theory of the individuality of the chromosomes was first proposed by Boveri (’88) as a result of his work on Ascaris. He found a constant number of chromosomes in each species, always half this number in the two maturation divisions, and the original number restored by fertilization. Every year adds to the number of species found conforming to these rules, and consequently making Boveri’s theory more plausible. Beginning with Sutton’s work in 1g00, many species have been shown to give evidence of a more direct nature, and among these, the Hemiptera Homop- tera can be classed. In the first place, it is a sign of individuality, when we are able to pick out one chromosome in every equatorial plate by some characteristic size, shape or position. ‘This can be done for 14 out of the 22 species of Hemiptera Homoptera studied, the characteristic usually being the large size of the chromosome (see Peeciloptera septentrionalis, Figs. 270, 273, 274). Secondly, all evidence that supports Montgomery’s hy- pothesis of the union of paternal and maternal chromosomes in synapsis necessarily supports the theory of the individuality of the chromosomes. In Peeciloptera septentrionalis, the large chromo- some in the spermatocytes (Fig. 270) 1s represented in the sperma- togonia (Fig. 259) by two large chromosomes. Half of the chro- mosomes in each spermatogonial plate must have come originally from the spermatozo6n, and half from the egg. Only one large chromosome could be received from the spermatozoon, according to Fig. 270, therefore the other large one must have come from the egg. As these two large chromosomes, one paternal and one maternal, are represented by a single chromatic element in the spermatocyte, this must be formed by the union of a paternal with a maternal chromosome of the spermatogonium. Thus we see that the Hemiptera Homoptera are in accord with Montgomery’s hypothesis of synapsis and reduction. In the third place, the behavior of the odd chromosome supports Boveri’s theory. In the Hemiptera Homoptera, the odd chromosome can seldom be identi- fied in the spermatogonia, but from the contraction stage to the S permatogenesis 503 anaphase of the first spermatocyte, and sometimes to the meta- phase of the second spermatocyte (Figs. 56 and 149) its individ- uality is marked. It takes the basic stains when the rest of the chromatin takes acid stains; it frequently has a smooth round con- tour in the early prophase, when the other chromosomes are irreg- ular rods or tetrads; it usually is closely applied to the nuclear mem- brane until that is dissolved, and then keeps an eccentric position in the first spermatocyte equatorial plate; it does not divide in this divi- sion, and either precedes or follows the other chromosomes to the pole. In Vanduzea arcuata (Fig. 56), where it is intermediate in size, and in Chlorotettrix (Fig. 149), where it is the smallest chro- mosome, its individuality is still marked in the second spermato- cyte. Finally the facts that have brought about the dropping of the old discussion about prereduction and postreduction, speak for the individuality of the chromosomes, in that they show the essential point of reduction to be the separation of each maternal chromosome from its paternal mate, and their distribution to differ- ent spermatozoa. ‘The uselessness of insisting on prereduction or postreduction is shown within the order Hemiptera, where the odd chromosome may divide in either division; in the Heteroptera, it usually divides in the first, while in the Homoptera, the usual place of division is the second spermatocyte, but Archimerus and Banassa are exceptions in the former and Enchenopa in the latter. Value of the Number of Chromosomes in Taxonomy and Evolution McClung (’05) states that for Orthoptera, a certain number of chromosomes is characteristic for each family, the chromosome grouping marking the genus, and the relative size of the chromo- somes indicating the species. Unfortunately this is not true for the Hemiptera Homoptera as the number varies within the family and even within the genus, being constant for the species only. The case of Aphrophora quadrangularis may make this doubtful, although it seems more probable that two or three species have previously been included under one name, than that in the same species, the reduced number should be sometimes 12 and some- times 11, which would not accord with the simplest laws of heredity. 504 Alice M. Boring Montgomery has for many years endeavored to determine the stage of evolution by the number of chromosomes that a species possesses, those having few being considered higher in the scale than those with many. ‘The chromatin nucleoli were supposed to be degenerating chromosomes as a species evolves to a higher form. But he has recently collected data from all the scattered literature, tabulated the number of chromosomes and the species, and finds that there is no such correlation (’06b). Inthe Hemiptera Homop- tera there is no reason for considering Vanduzea arcuata, with 9 chromosomes, more highly evolved than Entilia sinuata, with 11, or Phlepsius irrotatus, with 8, more so than Pceciloptera septen- trionalis, with 14. Sex Determination We have seen in the historical review of the work on tracheate spermatogenesis, that the most recent and reliable work all points to a dimorphism of the spermatozoa in the forms with an odd chro- mosome or an unequal pair of chromosomes. McClung was the first to suggest that the one characteristic that most generally divides the animal kingdom into two equal classes is sex, and that therefore, the dimorphism of sex and of spermatozoa may be causally connected. ‘There is need of careful statistical work on the proportion of males and females among different species of insects. In general collecting, however, one gets an impression of equality in numbers. McClung’s theory was a brilliant guess, which the work of Stevens and Wilson has substantiated. The Hemiptera Homoptera furnish additional evidence for this theory. Females of many of the species were sectioned for o6go- nial and somatic equatorial plates. Only two furnished the desired stages, Poeciloptera septentrionalis and Pececiloptera pruinosa. In both the spermatogonial number is 27, the spermatozoa pos- sessing 13 and 14 chromosomes, and the female somatic number 1s 28. Stevens and Wilson have shown that there is no difference between the somatic number and the unreduced number in the germ cells; in the female, both numbers are even, in the male, both are odd (or even, when a small chromosome is included). As the female somatic number in Peeciloptera is even, the odgonial S permatogenesis 505 number must also be even, and all the maturated eggs necessarily possess the same number of chromosomes, 14._ Applying Wilson’s Co6b) formula for sex determination to the Peeciloptera, we have the following: I Egg (14 chromosomes) + Spermatozoon (14 chromosomes) = Female (28 chromosomes). Il Egg (14 chromosomes) + Spermatozoon (13 chromosomes) = Male (27 chromosomes). Here again it is possible to apply Castle’s (’00) theory of sex as a Mendelian character, which has been so fully elaborated and applied to the case of the odd chromosome by Wilson. _ It involves the assumption of two kinds of eggs, male and female, as well as the two kinds of spermatozoa which are actually to be observed. It also involves the assumption of selective fertilization: an egg bearing the female determinant must be fertilized by a spermato- zoon with the male determinant, while an egg bearing the male determinant must be fertilized by a spermatozoon with the female determinant. In case II of the above formula when the egg is fertilized by the spermatozoon without the odd chromosome, the sex determinant must be introduced by the egg; and as in this case, a male is produced, the eggs fertilized by a spermatozoon without an odd chromosome must bear the male determinant, and the chromosome which has disappeared in the males must be the one with the female character. So in case I, where the egg 1s fertil- ized by the spermatozoon with the odd chromosome, the sperma- - tozoon must bear the male character and the egg the female; as this combination always results in a female, it 1s necessary to assume that the male character is recessive and the female domi- nant. he above formule can be extended to show these assump- tions and will read thus: I 9 Egg (14 chromosomes) + (%) Spermatozoon (14 chro- mosomes) = @ (@) Female (28 chromosomes). Il (#) Egg (14 chromosomes) + (0) Spermatozoon (13 chro- mosomes) = (3) (0) Male (27 chromosomes). This is the part of Wilson’s theory that deals with the case presented by Peeciloptera and presumably the other Hemiptera Homoptera. The facts as far as they go are not at variance with the theory. 506 Alice M. Boring SUMMARY 1 An odd chromosome is present in the spermatogenesis of 22 species of the Hemiptera Homoptera, as shown in each case by some or all of the following facts: a ‘The spermatogonia have an uneven number of chromosomes. b A dense body takes basic stains in the growth period. c One chromosome stands in an eccentric position in the first spermatocyte equatorial plate. d In the metaphase of the first spermatocyte division, one chromosome does not divide, and has half the valence of the others, as shown by its spherical shape when the others are like dumb- bells. e In anaphase of the first spermatocyte division, one chromo- some at one pole behaves differently from the others, either pre- ceding or lagging behind. 7 Half of the equatorial plates of the second spermatocytes contain the same number of chromosomes as those of the first spermatocytes, but half contain one less. g Half of the spermatids contain a condensed body, taking basic stains, which is the odd chromosome. 2 [The odd chromosome shows certain variations in behavior, either individual or specific. a In the anaphase of the division where it does not divide, in some cells it may precede the other chromosomes to the poles, while in others it lags behind them. This individual variation is a characteristic of certain species, the three species of Ceresa and Vanduzea arcuata, while most of the species studied have the odd chromosome always lagging be- hind, and Phlepsius irrotatus has it always preceding the others. b In Enchenopa binotata, it divides in the first division, and in the second division, where it does not divide, it neither precedes nor lags behind the others. c The shape of the odd chromosome in the growth -period varies. It may be always spherical or ovoid with a smooth con- tour, as inthe Fulgoride, Cercopida, Jassidz, and some of the Membracide. It may be long and uneven in contour as in Van- duzea arcuata and Enchenopa binotata. ; S permatogenesis 507 It may pass through both forms in different stages, as in Entilia sinuata. 3 Inthe spermatids of 19 species; that is, all except three of the Membracidz, there is a chromatin nucleolus in all of the sperma- tids entirely independent of the odd chromosome. In seven of these species, the odd chromosome is present also in half of the spermatids, in others there is no indication of it. In the three Membracidz without the chromatin nucleolus, Entilia sinuata, Vanduzea arcuata, and Campylenchia curvata, the odd chromo- some is present in half of the spermatids. 4. In the genus Ceresa, in the contraction stage some of the basichromatin is thrown out from the chromatin loops and per- sists through the growth period as a chromatin deposition on the nuclear membrane and finally dissolves without apparently taking part in the formation of the chromosomes for the first spermato- cyte division. 5 In three species, Campylenchia curvata, Poeciloptera septen- trionalis, and Poeciloptera pruinosa, a pair of m-chromosomes remain condensed in the growth period. 6 The number of chromosomes has no significance for group- ing species into families. In reduced number, in the Membracidz, 5 species have 11 chromosomes 2 species have 10 chromosomes I species has — g chromosomes in the Jassidz, 2 species have 12 chromosomes 2 species have 11 chromosomes I species has —_g chromosomes I species has 8 chromosomes in the Cercopide, species has 14 chromosomes species has 11 chromosomes I I species has 12 chromosomes I I species has 8 chromosomes in the Fulgorida, 2 species have 14 chromosomes 2 species have 13 chromosomes 7 [he number of chromosomes has no significance for group- ing species into genera. 508 Alice M. Boring Chlorotettrix unicolor, I1 chromosomes Chlorotettrix vividus, g chromosomes Aphrophora quadrangularis, 11 or 12 chromosomes Aphrophora 4-notata, 14 chromosomes Peeciloptera septentrionalis, 14 chromosomes Peeciloptera bivittata, 13 chromosomes 8 The number of chromosomes is constant for each species. In the case of Aphrophora quadrangularis, where there have been found both 11 and 12 chromosomes, probably two species are pres- ent, which have not been separated in classification. g The only points in the spermatogenesis in which all of the species of one family resemble each other more closely than they do the species of the other families are the appearance of some of the growth stages and the transformation of the spermatid into the spermatozoon. 10 In fourteen of the species studied, the individuality of cer- tain chromosomes can be traced from the spermatogonium to the second spermatocyte, a pair of similar chromosomes 1n the sperma- togonium bearing the same size relation to the other chromosomes of the equatorial plate as a single chromosome bears to the others in the first and second spermatocyte plates. In all the species, the odd chromosome can be traced as keeping its individuality from the growth period to the anaphase of the first spermatocyte division, in Chlorotettrix and Vanduzea arcuata to the metaphase of the second spermatocyte division, and in Enchenopa binotata, from the spermatogonial plate to the telophase of the second sper- matocyte division. 11 In all 22 species, there is a dimorphism of the spermatozoa, which probably corresponds to the natural dimorphism of sex. 12 ‘[wo species of Fulgoridz in which the female somatic num- ber of chromosomes is 28, while the spermatogonial number is 27, furnish further proof for the theory of sex determination advanced by McClung, Wilson and Stevens. Bryn Mawr College May 4, 1907 S permatogenesis 509 BIBLIOGRAPHY BaumGarTn_ER, W. J., ’02—Spermatid Transformation. Kans. Univ. Sci. Bull., ty Dara 7 °o4--Some New Evidences for the Individuality of the Chromosomes. Biol. Bull., vi, p. 1. VAN BENEDEN, E.,’83—Maturation de |’Oeuf, la Fecondation et la Division Cellu- laire. Paris. Berry, E. H., ’06—The “Accessory Chromosome”’ in Epeira. Biol. Bull., xi, P2293: Biacxman, M. W., ’o1—Spermatogenesis of Myriapods. I Notes on Spermato- cytes and Spermatids of Scolopendra. Kans. Univ. Quart., x, p. 61. ’03—Spermatogenesis of Myriapods. II On the Chromatin in the Spermatocytes of Scolopendra heros. Biol. Bull., v, p. 187. ’o5—Spermatogenesis of Myriapods. III Spermatogenesis of Scolo- pendra heros. Bull. Mus. Comp. Zool., Harvard, xvii, p. 1. ’o5b—Spermatogenesis of Myriapods. IV On the Karyosphere and Nucleolus in the Spermatocytes of Scolopendra subspinipes. Proc. Am. Acad. Arts and Sci., xli, p. 331. Bovert, T., '87—Zellen-Studien I. Jena. °88—Zellen-Studien II. Jena. Burscuu1, O., ’71—Vorlaufige Mittheilung iber Bau und Entwicklung der Samen- faden der Insekten und Crustaceen. Z. wiss. Zool., xxi, p. 402. CastLe, W. E., ’03—The Heredity of Sex. Bull. Mus. Comp. Zool., Harvard, xl, p. 189. Conkitn, E. G., ’02—Karyokinesis and Cytokinesis. Jour. Acad. Nat. Sci., Philat, xin, p- 5; Foot, K. anp E. C. SrroBeti, ’07—The “Accessory Chromosome” of Anasa tristis. Biol) Bull; xn, p: 119; Gross, J., ’04—Die Spermatogenese von Syromastes marginatus. Zool. Jahrb. f. Anat. u. Ont., xx, p. 439. ’06—Die Spermatogenese von Pyrrhocoris apterus. Zool. Jahrb., xxi, p- 269. Henkin, H., ’90a—Untersuchungen tiber die ersten Entwicklungsvorgange in den Eiern der Insekten. II Ueber Spermatogenese und deren Beziehung zur Eientwicklung bei Pyrrhocoris apterus. Z. wiss. Zool., li, p. 685. *gob—Ueber Reduktionstheilung der Chromosomen in den Samenzellen von Insekten. Internat. Monatsschr. f. Anat. u. Phys., vii, p. 243. Hertwic, O., ’90—Vergleich der Ei- und Samenbildung bei Nematoden. Arch. f. Mikr. Anat., xxxvi, p. I. 510 Alice M. Boring Mark, E. L., *81—Maturation, Fecundation and Segmentation of Limax cam- pestris. Bull. Mus. Comp. Zool., Harvard, vi, p. 173. McCune, C. E., ’98—A Peculiar Nuclear Element in Male Reproductive Cells of Insects. Zool. Bull., 11, p. 187. *oo—Spermatocyte Divisions of the Acrididz. Kans. Univ. Quart., ix, P: 73: °o1—Notes on the Accessory Chromosome. Anat. Anz., xx, p. 220. o2a—Spermatocyte Divisions of the Locustida. Kans. Univ. Quart., x1, p. 185. o2b—Accessory Chromosome—Sex Determinant? Biol. Bull., ii, p. 43. °o5—The Chromosome Complex of Orthopteran Spermatocytes. Biol. Bull., 1x, p. 304. McGr11, C., ’04—Spermatogenesis of Anax junius. Uniy. Missoun Stud., II, 5. Mepes, G., ’°05—Spermatogenesis of Scutigera forceps. Biol. Bull., ix, p. 156. Montcomery, |. H., ’97—A Preliminary Note on the Chromatin Reduction in Spermatogenesis of Pentatoma. Zool. Anz., xx, p. 457. *98—Spermatogenesis of Pentatoma. Zool. Jahrb., xii, p. 1. *99—Chromatin Reduction in Hemiptera. Zool. Anz., xxii, p. 76. ‘oo—Spermatogenesis of Peripatus. Zool. Jahrb., xiv, p. 277. ora—A Study of the Chromosomes of the Germ Cells of Metazoa. Trans. Am: Philosac.5 30% ps is 3 *orb—Further Studies on the Chromosomes of the Hemiptera Heterop- tera. “Proc: Acad. Nat: Se.,'Phila.; lini, p. 261- °04—Some Observations and Considerations on the Maturation Phenom- ena of Germ Cells. Biol. Bull., vi, p. 137. *o5—Spermatogenesis of Syrbula and Lycosa. Proc. Acad. Nat. Sci, Phila., lvii, p. 162. °o6a—Terminology of Aberrant Chromosomes and their Behavior in certain Hemiptera. Science, n. s., xxiii, p. 36. *0o6b—Chromosomes in the Spermatogenesis of the Hemiptera Heterop- tera. Trans. Am. Phil. Soc., n. s., xxi, p. 97. Moore AND Rosinson, '07—On the Behavior of the Nucleolus in the Spermato- genesis of Periplaneta americana. Quart. Jour. Mic. Sci., xlvin, P- 571- Nowtin, W. N., ’06—A Study of the Spermatogenesis of Coptocycla aurichalcea and Coptocycla guttata, with especial reference to the Problem of Sex Determination. Jour. Exp. Zodl., 111, p. 583. Orrte, H., ’06—Samenreifung und Samenbildung von Locusta viridissima. Zool. Ainz., XXX, p- 520- PauLMIER, F. Cc *98—Chromatin Reduction in Hemiptera. Anat. Anz., xiv, p- 514. : S permatogenesis 511 Paurmier, F. C., ’99—Spermatogenesis of Anasa tristis. Jour. Morph., xv, sup- plement, p. 223. PratnEr, G., ’86—Die Karyokinese bei den Lepidoptera als Grundlage fiir eine Theorie der Zelltheilung. Internat. Monatsschr. f. Anat. u. Histol., ill, p. 341. vom Ratu, ’91—Ueber die Reduktion der Chromatischen Elemente in der Samen- bildung von Gryllotalpa. Ber. d. Naturf. Ges. Freiburg, vi, p. 62. ’92—Zur Kenntniss der Spermatogenese von Gryllotalpa vulgaris. Arch. f. Mikr. Anat., xl, p. 102. *95—Neue Beitrage zur Frage der Chromatinreduktion in der Samen- und Eireife. Arch. f. mikr. Anat., xlvi, p. 168. *g4—Ueber die Konstanz der Chromosomenzahl bei Thieren. Biol. Cen- tralbl., xiv, p. 449. SapaTIER, A., ’90—De la Spermatogénése chez les Locustides. Compt. Rend. Acad., Paris, cxi, p. 797. DE SinETY, R. ’01—Recherches sur la Biologie et |’Anatomie des Phasmes. La Cellule, sax; p. 127. STEVENS, N. M., ’o5a—A Study of the Germ Cells of Aphis rose and Aphis ceno- ° theres Jounséxps Zool, 1,)\p. 312: o5b—Studies in Spermatogenesis with especial reference to the ‘‘ Acces- sory Chromosome.” Carnegie Ins., Wash., pub. 36. °o6a—Studies on the Germ Cells of Aphids. Carnegie Ins., Wash., pub. 51. *o6b—Studies in Spermatogenesis II. A Comparative Study of the Het- erochromosomes in Certain Species of Coleoptera, Hemiptera, and Lepidoptera, with especial reference to Sex Determination. Car- negie Ins., Wash., pub. 36, 11. Sutton, W. S., ’oo—Spermatogonial Divisions in Brachystola magna. Kans. Unive Oudre six p= 135: *02—Morphology of the Chromosome Group in Brachystola magna. Biol. Bull., iv, p. 24. °03—Chromosomes in Heredity. Biol. Bull., iv, p. 231. vy. La VaALeTTE St. GeorGE, ’85—Spermatologische Beitrage I. Arch. f. mikr. Anat., xxv, p. 581. Verson, E., ’8g—Zur Spermatogenese. Zool. Anz., xi, p. 100. *94_Zur Spermatogenese bei der Seidenraupe. Z. wiss. Zool., lviii, p03: Vornoy, D. N., ’03—Spermatogénése d’été chez le Cybister roeselei. Arch. d. Zool. Exp. et Genl., 4 series, 1, p. 173. 512 Alice M. Boring Voinoy, D. N.,’04—Sur une Disposition special a la Chromatin dans la Spermato- génese du Gryllus campestris Reproduisant des Structures Observés seulement dans l’Ovogénése. Arch. d. Zool. Exp. et Genl., ¥ series, 11, Notes et Revues, p. 63. Wattace, L. B., ’00—The Accessory Chromosome in the Spider. Anat. Anz., XVIll, Pp. 327. 05—Spermatogenesis of the Spider. Biol. Bull., vili, p. 169. Weismann, A., °87—On the Number of Polar Bodies and their Significance in Heredity. Essays on Heredity. Oxford. Wricox, E. V., ’°94—Spermatogenesis of Caloptenus femur-rubrum. Preliminary Notice. Anat. Anz., x, p. 303. *95—Spermatogenesis of Caloptenus femur-rubrum and Cicada tibicens. Bull. Mus. Comp. Zool., Harvard, xxvii, p. 3. *96—Further Studies on the Spermatogenesis of Caloptenus femur- rubrum. Bull. Mus. Comp. Zool., Harvard, xxix, p. 193. °97—Chromatic Tetrads. Anat. Anz., xiv, p. 194. Wirson, E. B., ’05a—Chromosomes in Relation to the Determination of Sex in Insects. Science, n. s., Xx, p. 500. o5b—Studieson Chromosomes. I ‘The Behavior of the Idiochrosomes inthe Hemiptera. Jour. Exp. Zodl., 11, p. 371. ’o5c—Studies on Chromosomes. II The Paired Microchromosomes, Idiochromosomes, and Heterotropic Chromosomes in the Hemiptera. Jour. Exp. Zool., 11, p. 507. *o6a—Studies on Chromosomes. III Sexual Differences of the Chro- mosome Groups in Hemiptera, with some Considerations on Deter- mination and Inheritance of Sex. Jour. Exp. Zodl., iii, p. 1. °0o6b—A New Theory of Sex Production. Science, n. s., xxii, p. 189. °07—The case of Anasatristis. Science, n.s., xxv, p. IQI. ZWEIGER, H., ’06—Die Spermatogenese von Forficula auricularia. Zool. Anz., XX, p= 220. DESCRIPTION OF PLATES. The figures were drawn with the aid of the Zeiss-Abbe drawing camera, No. 111. Figs. 1-46 were drawn with a Leitz oil immersion obj. 1'z and a Zeiss compensating oc. 12, Figs. 47-313 with a Zeiss apo- chromatic oil immersion obj. 2 mm., oc. 12. They have been reduced one-third, giving a magnification of about rooo diameters. Abbreviations Used on Plates a; and a2= one pair of spermatogonial chromosomes. a = a bivalent primary spermatocyte chromosome representing a; and ap. b; and = one pair of spermatogonial chromosomes. b = a bivalent primary spermatocyte chromosome representing };and bp. my and mz = a pair of m-chromosomes. n = chromatin nucleolus. p, pi, p2 = plasmosomes. x = odd chromosome. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. 6 CO) Qin FF Wo et [Sy ta (o) 14 15 16 Prate I Entilia sinuata (Family Membracide) Spermatogonial rest stage. Spermatogonial spireme. Spermatogonium, segmentation of the spireme, each segment longitudinally split. Spermatogonium, condensation of the segments of the spireme. Spermatogonial equatorial plate, 21 chromosomes. Spermatogonial metaphase. Spermatogonial anaphase. Spermatogonial telophase, formation of nuclear membrane. Spermatogonial telophase, polar view. First spermatocyte, contraction stage. First spermatocyte, early postsynapsis stage. First spermatocyte, late postsynapsis, fine spireme. First spermatocyte, coarse spireme. First spermatocyte, rest stage. First spermatocyte, split spireme. First spermatocyte, spireme divided into 11 split segments. Figs. 17-19 First spermatocyte, early prophase, tetrad formation. Fig. 20 First spermatocyte, late prophase, dumb-bell formation. Fig. 21 First spermatocyte, equatorial plate, 1r chromosomes. Fig. 22 First spermatocyte, metaphase, chromosomes still tetrads. Figs. 23,24 First spermatocyte, metaphase. Figs. 25,26 First spermatocyte, anaphase, centrosomes divided for the second division. Fig. Fig. Fig. Fig. 27 28 29 30° First spermatocyte, telophase. First spermatocyte, telophase, polar view. Rearrangement of chromosomes for the second spermatocyte division. Second spermatocyte, equatorial plate, 11 chromosomes. SPERMATOGENESIS PLATE I Auice M. Borinc A. M. B. del. MEMBRACID Tue JourNnaL or ExpertIMENTAL ZOOLOGY, VOL. Iv, NO. 4 Pirate I] Entilia sinuata (continued) Fig. 31 Second spermatocyte, equatorial plate, 10 chromosomes. Fig. 32 Second spermatocyte, metaphase. Fig. 33 Second spermatocyte, anaphase. Fig. 34 Second spermatocyte, anaphase, polar view. Fig. 35 Second spermatocyte, telophase. Fig. 36 Spermatid, first stage. Figs. 37, 38 Spermatids, second stage, half with «, half without. Figs. 39, 40 Spermatids, third stage, half contain x, half do not. Figs. 41,42 Spermatids, formation of axial filament, half contain x, haif do not. Figs. 43,44 Spermatids, condensation of the chromatin, half contain x, half do net. Figs. 45, 46 Spermatids, later stages. Fig. 47. Mature spermatozoon. Vanduzea arcuata (Family Membracide) Fig. 48 Spermatogonial equatorial plate, 17 chromosomes. Figs. 49, 50 First spermatocyte, growth period. Fig. 51 First spermatocyte, equatorial plate, 9 chromosomes. Fig. 52 First spermatocyte, metaphase. Figs. 53-55 First spermatocyte, anaphase. Figs. 56,57 Second spermatocytes, equatorial plates, containing 9 and 8 chromosomes, respectively. Fig. 58 Second spermatocyte, metaphase. Fig. 59 Second spermatocyte, anaphase. Figs. 60, 61 Spermatids, half contain x, half do not. PLATE II SPERMATOGENESIS Atice M. Borine A. M. B. del. MEMBRACID © THE JourNAL or ExprrIMENTAL ZOOLOGY, VoL. IV, NO. 4 Pirate III Ceresa taurina (Family Membracide) Figs. 62, 63 First spermatocyte contraction stage, a mass of rejected basichromatin at the base of the loops. Figs. 64-66 First spermatocyte, rest stage, showing rejected basichromatin. Fig. 67 First spermatocyte, rest stage, showing x in the midst of the rejected chromatin. Fig. 68 First spermatocyte, split spireme stage. Most of the rejected chromatin has dissolved, showing x plainly. Fig. 69 First spermatocyte, prophase. Fig. 70 First spermatocyte, equatorial plate, 11 chromosomes. Figs. 71-73 First spermatocyte, metaphase. Figs. 74,75 First spermatocyte, anaphase. Figs. 76, 77 Second spermatocyte, equatorial plates, containing 11 and 10 chromosomes, respectively. Fig. 78 Second spermatocyte, metaphase. Fig. 79 Second spermatocyte, anaphase. Fig. 80 Spermatid, with chromatin nucleolus. _ Ceresa bubalus (Family Membracide) Fig. 81 First spermatocyte, synapsis stage, showing rejected chromatin. Fig. 82 First spermatocyte, rest stage, showing rejected chromatin. Fig. 83. First spermatocyte, equatorial plate, 11 chromosomes. Figs. 84, 85 First spermatocytes, metaphase. Figs. 86, 87 First spermatocytes, anaphase. Figs. 88, 89 Second spermatocytes, equatorial plates, containing 11 and to chromosomes, respect- ively. Fig. go Second spermatocyte, metaphase. Fig. 91 Second spermatocyte, anaphase. Fig. 92 Spermatid, with chromatin nucleolus. Ceresa diceros (Family Membracide) Fig. 93 First spermatocyte, rest stage, showing rejected chromatin. Fig. 94 First spermatocyte, equatorial plate, 11 chromosomes. Fig. 95 First spermatocyte, metaphase. Fig. 96 First spermatocyte, anaphase. Figs. 97, 98 Second spermatocytes, equatorial plates, containing 11 and 10 chromosomes, respect- ively. Fig. 99 Second spermatocyte, metaphase. Fig. 100 Second spermatocyte, anaphase. Fig. tor Spermatid with chromatin nucleolus. Atymna castanea (Family Membracide) Fig. 102 First spermatocyte, rest stage. Fig. 103. First spermatocyte, equatorial plate, 11 chromosomes. | | | PLATE IIl SPERMATOGENESIS Atice M. Borinc ies 62 A. M. B. del. MEMBRACID-E THE JournaL or ExperRIMENTAL ZOOLOGY, VOL. IV, No. 4 Prate IV Atymna castanea (continued) Figs. 104, 105 First spermatocyte, metaphase. Fig. 106 First spermatocyte, anaphase. Figs. 107, 108 Second spermatocytes, equatorial plates, containing 11 and 10 chromosomes, respect- ively. Fig. 109 Second spermatocyte, metaphase. Fig. 110 Second spermatocyte, anaphase. Fig. 111 Spermatid, with chromatin nucleolus. Campylenchia curvata (Family Membracide) Fig. 112 Spermatogonial equatorial plate, 19 chromosomes. Fig. 113 First spermatocyte, rest stage. Fig. 114 First spermatocyte, equatorial plate, 10 chromosomes. Fig. 115 First spermatocyte, metaphase. Fig. 116 First spermatocyte, anaphase. Figs. 117, 118 Second spermatocytes, equatorial plates, containing 10 and 9 chromosomes, respect- ively. Fig. 119 Second spermatocyte, metaphase. Fig. 120 Second spermatocyte, anaphase. Figs. 121,122 Spermatids, half with x, half without. Enchenopa binotata (Family Membracide) Fig. 123. Spermatogonial equatorial plate, 19 chromosomes. Fig. 124 First spermatocyte, spireme stage. Fig. 125 First spermatocyte, early prophase. Fig. 126 First spermatocyte, equatorial plate, 10 chromosomes. Fig. 127. First spermatocyte, equatorial plate, 11 chromosomes, occasionally found. Figs. 128,129 First spermatocytes, metaphase. Fig. 130 First spermatocyte, anaphase. Fig. 131 Second spermatocyte, equatorial plate, 10 chromosomes. Figs. 132, 133 Second spermatocytes, metaphase. Fig. 133 x does not divide in this division. Fig. 134 Second spermatocyte, anaphase. Fig. 135a and b Second spermatocyte anaphase, two plates from the same spindle, 9 chromosomes in one, Io in the other. Fig. 136 Spermatid, with chromatin nucleolus. Chlorotettrix unicolor and Chlorotettrix vividus Family fasside) Fig. 137 Spermatogonial rest stage. Fig. 138 Spermatogonial metaphase. Fig. 139 Spermatogonial anaphase. Fig.140 First spermatocyte, contraction stage. Fig. 141 First spermatocyte, spireme stage. Fig. 142 First spermatocyte, rest stage. Fig. Fig. Fig. Fig. 143 First spermatocyte, split spireme stage. 144 First spermatocyte, prophase. 145 First spermatocyte, equatorial plate, 9 chromosomes. 146 First spermatocyte, metaphase. | ———— ee SPERMATOGENESIS PLATE IV Atice M. Borinc 107 108 109 A. M. B. del. MEMBRACID AND JASSIDAE Tue JourNAL or ExpERIMENTAL ZOOLOGY, VOL. IV, NO. 4 Pirate V Chlorotettrix unicolor and Chlorotettrix vividus (continued) Fig. 147 First spermatocyte, anaphase. Figs. 148, 149 Second spermatocytes, equatorial plates, containing 8 and 9 chromosomes, respect- ively. Fig. 150 Second spermatocyte, metaphase. Fig. 151 Second spermatocyte, anaphase. Fig. 152 Spermatid, first stage. Figs. 153,154 Spermatid, second stage, half with x, half without. Figs. 155,156 Spermatid, third stage, half with x, half without. Figs. 157,158 Late spermatid stages. Fig. 159 Head of mature spermatozoén. Fig. 160 Spermatogonial equatorial plate, 21 chromosomes. Fig. 161 First spermatocyte equatorial plate, 11 chromosomes. Figs. 162, 163 Second spermatocytes, equatorial plates, containing 11 and 10 chromosomes, respect- ively. Diedrocephala coccinea (Family Fasside) Fig. 164 Spermatogonial equatorial plate, 23 chromosomes. Fig. 165 First spermatocyte, postsynapsis stage. Fig. 166 First spermatocyte, rest stage. Fig. 167 First spermatocyte, equatorial plate, 12 chromosomes. Fig. 168 First spermatocyte, metaphase. Fig. 169 First spermatocyte, anaphase. Figs. 170,171 Second spermatocytes, equatorial plates, containing 12 and 11 chromosomes, respect- ively. Fig. 172 Second spermatocyte, metaphase. Fig. 173, Second spermatocyte, anaphase. Figs. 174,175 Spermatids, half without x, half with. Diedrocephala mollipes (Family fasside) Fig. 176 First spermatocyte, rest stage. Fig. 177 First spermatocyte, equatorial plate, 12 chromosomes. Fig. 178 First spermatocyte, metaphase. Fig. 179 First spermatocyte, anaphase. Figs. 180, 181 Second spermatocytes, equatorial plates, containing 12 and 11 chromosomes, respect- ively. Fig. 182 Second spermatocyte, metaphase. Fig. 183 Second spermatocyte, anaphase. Figs. 184,185 Spermatids, half without «, half with. SPERMATOGENESIS PLATE V Autce M. Borinc a b 148 on bat? He 2 “ill o> @ ® *€). Seats 0,003 155 156 157 158 160 154 ie 180 183 184 185 A. M. B. del. JASSID THe Journa or ExpERIMENTAL ZOOLOGY, VOL.IV, NO. 4 Pratre VI Phlepsius irrotatus (Family Fasside) Fig. 186 Spermatogonial equatorial plate, 15 chromosomes. Fig. 187 First spermatocyte, rest stage. Fig. 188a and b First spermatocyte, equatorial plate, and the odd chromosome x. Fig. 189 First spermatocyte, metaphase. Fig. 190 First spermatocyte anaphase. Figs. 191, 192 Second spermatocytes, equatorial plates, containing 8 and 7 chromosomes, respect- ively. Fig. 193 Second spermatocyte, metaphase. Fig. 194 Second spermatocyte, anaphase. | Figs. 195,196 Spermatids, half without x, half with. A gallia sanguinolenta (Family fasside) Fig. 197 First spermatocyte, spireme stage. Fig. 198 First spermatocyte, equatorial plate, 11 chromosomes. Fig. 199 First spermatocyte, metaphase. Figs. 200, 201 Second spermatocytes, equatorial plates, containing 11 and 10 chromosomes, respect- ively. Fig. 202 Second spermatocyte, early anaphase. Figs. 203, 204 Spermatids, half without x, half with. Fig. 205 First spermatocyte, equatorial plate, aceto-carmine preparation. . Fig. 206 First spermatocyte, anaphase, aceto-carmine preparation. Fig.207 Second spermatocyte, equatorial plate, aceto-carmine preparation. Clastoptera sbtusa (Family Cercopide) Fig. 208 Spermatogonial rest stage. Fig. 209 Spermatogonial prophase. Fig. 210 Spermatogonial equatorial plate, 15 chromosomes. Fig. 211 Spermatogonial metaphase. Fig. 212 Spermatogonial anaphase. Figs. 213, 214. First spermatocyte, early synapsis. Fig. 215 First spermatocyte, contraction stage. Fig. 216 First spermatocyte, postsynapsis stage. Fig. 217 First spermatocyte, spireme stage. , Fig. 218 First spermatocyte, early prophase, tetrad formation. Fig.219 First spermatocyte, prophase, dumb-bell formation. Fig. 220 First spermatocyte, equatorial plate, 8 chromosomes. a Fig. 221 First spermatocyte, metaphase. Fig. 222 First spermatocyte, anaphase. Figs. 223, 224 Second spermatocytes, equatorial plates containing 8 and 7 chromosomes, respect- ively. Fig. 225 Second spermatocyte, metaphase. ~~? eS a”) lh Lee SPERMATOGENESIS PLATE VI Auice M. BorinG a, *, 186 191 : Ga 197 198 35 200 201 202 @& ® Me afta\y an & ‘S hh A is @e eo @e.@ ity! @® = QPP mH rig 203 204 205 aoa te 209 210 au 212 213 A. M. B. del. JASSIDZZ AND CERCOPID: THE JourNAL of ExPeRIMENTAL ZOOLOGY, VOL. IV, NO. 4 Pirate VII Clastoptera obtusa (Continued) Fig. 226 Second spermatocyte, anaphase. Fig. 227. Early spermatid, with chromatin nucleolus. Fig. 228 Spermatid, formation of axial filament. Figs. 229, 230 Later spermatids. Fig. Fig. Fig. Fig. Fig. Fig. Fig. 231 232 235 eB 4, 235 236 237 Mature spermatozoon. A phrophora quadrangularis with 11 chromosomes (Family Cercopide) Spermatogonial equatorial plate, 21 chromosomes. First spermatocyte, contraction stage. First spermatocyte, spireme stage. First spermatocyte, equatorial plate, 11 chromosomes. First spermatocyte, metaphase. First spermatocyte, anaphase. Figs. 238, 239 Second spermatocytes, equatorial plates, containing 11 and 10 chromosomes, respect- ively. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. 240 241 242 243 244 245 246 247 248 2a) 250 251 252 Second spermatocyte, metaphase. Second spermatocyte, anaphase. Spermatid, with chromatin nucleolus. A phrophora quadrangularis with 12 chromosomes (Family Cercopide) First spermatocyte, contraction stage. First spermatocyte, equatorial plate, 12 chromosomes. First spermatocyte, metaphase. First spermatocyte, anaphase. Second spermatocyte, equatorial plate, 12 chromosomes. Second spermatocyte, anaphase. ; Aphrophora 4-notata (Family Cercopide) First spermatocyte, spireme stage. First spermatocyte, equatorial plate, 14 chromosomes. First spermatocyte, metaphase. First spermatocyte, anaphase. Figs.253,254 Second spermatocytes, equatorial plates, containing 14 and 13 chromosomes, respect- ively. Fig. 255 Second spermatocyte, anaphase. SPERMATOGENESIS PLATE VII Auice M. Borinc 253 254 255 A. M. B. del. CERCOPID£ Tue Journar or EXPERIMENTAL ZO6LOGY, VOL. IV, NO. 4 Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Pratre VIII Peciloptera septentrionalis (Family Fulgorida) 256 Spermatogonial rest stage. 257 Spermatogonial split spireme. 258 Spermatogonium, spireme segmented and condensed, segments split. 259 Spermatogonial equatorial plate, 27 chromosomes. 260 Spermatogonial metaphase. 261 Spermatogonial anaphase. 262 First spermatocyte, early synapsis stage. 263 First spermatocyte, contraction stage. 264 First spermatocyte, spireme stage. Figs. 265,267 First spermatocyte, rest stages, growth in size of nucleus and cell. Fig. Fig. Fig. Fig. Fig. 268 First spermatocyte, split spireme stage. 269 First spermatocyte, prophase, tetrad formation. 270 First spermatocyte, equatorial plate, 14 chromosomes. 271 First spermatocyte, metaphase. 272 First spermatocyte, anaphase. Figs. 273,274 Second spermatocytes, equatorial plates, containing 14 and 13 chromosomes, respect- ively. Fig. Fig. 275 Second spermatocyte, metaphase. 276 Second spermatocyte, anaphase. Figs. 277,278 Spermatids, formation of fibers in the ‘‘ Nebenkern.” Fig. Fig 279a Spermatid, ‘‘Nebenkern” separated by a partition into two tubes. -279b Cross section of ‘‘ Nebenkern” structure as in 2792. Fig. Fig. .281b Cross sections of tubes of 281a. 280 Spermatid, elongation of fibers and tubes. 281a Spermatid, irregular spiral of twisted tubes. . 282a Spermatid, further twisting and flattening. . 282b Cross section of 282a. . 283 Female somatic equatorial plate, 28 chromosomes. SPERMATOGENESIS PLATE VII Auice M. Borinc A. M. B. del FULGORID Tue JourNnaL or ExPERIMENTAL ZoGLOGY, VOL. IV, No. 4 Prate IX Peciloptera pruinosa (Family Fulgoride) Fig. 284 First spermatocyte, rest stage. Fig. 285 First spermatocyte, equatorial plate, 14 chromosomes. Fig. 286 First spermatocyte, metaphase. Fig. 287 First spermatocyte, anaphase. Figs. 288, 289 Second spermatocytes, equatorial plates, containing 14 and 13 chromosomes, respect- ively . Fig. 290 Second spermatocyte, metaphase. Fig. 291 Second spermatocyte, anaphase. Figs. 292,293 Spermatids, half with x, half without. Fig. 294 Female somatic equatorial plate, 28 chromosomes. Amphiscepa bivittata (Family Fulgoride) Fig. 295 Spermatogonial equatorial plate, 25 chromosomes. Fig. 296 First spermatocyte, rest stage. Fig. 297 First spermatocyte, equatorial plate, 13 chromosomes. Fig. 298 First spermatocyte, metaphase. Fig.299 First spermatocyte, anaphase. Figs. 300, 301 Second spermatocytes, equatorial plates, containing 13 and 12 chromosomes, respect ively. _ Fig. 302 Second spermatocyte, metaphase. Fig. 303 Second spermatocyte, anaphase. Fig. 304 Spermatid. Peciloptera bivittata (Family Fulgoride) Fig. 305 First spermatocyte, rest stage. Fig. 306 First spermatocyte, equatorial plate, 13 chromosomes. Fig. 307 First spermatocyte, metaphase. Fig. 308 First spermatocyte, anaphase. Figs, 309,310 Second spermatocytes, equatorial plates, containing 13 and 12 chromosomes, respectively. Fig. 311 Second spermatocyte, metaphase. Fig. 312 Second spermatocyte, anaphase. Fig. 313 Spermatid. SPERMATOGENESIS PLATE IX AutcesM. Borinc 289 A. M. B. del. FULGORID Tue Journat or ExprrIMENTAL ZOOLOGY, VOL. IV, NO. 4 ~~ CONTRIBUTIONS FROM THE ZOOLOGICAL LABORATORY OF THE MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD COLLEGE. E. L. Marx, Director. No. 191. PH REACTIONS OF THE POMACE FLY, DROSO- Pow wave RCOPHILA LOEW, TO, ODOROUS SUBSTANCES WILLIAM MORTON BARROWS Wirth Five Ficures Te trod Metre erccusrwyorr ser. Saveur ace tevocsrsi siayeveiel ate 2: s SkGRvara oe ely = otere Seno naveRevedyy sto eokenetscatetane ieee 515 Hil Pees DETAINEES ets eperta SUIINEL 67ers Sch ie eo crs bias cee tes ee ee eI Se eB Ee o6 cor 536 OCNISUS GT 210) 1) AEApS OR ate Atti Ree iar eae a eer ART ra SSA. oo SEA ry 537 I INTRODUCTION Drosophila ampelophila is a small fly about three millimeters in length belonging to the family Drosophilide. It lays tis eggs on fermenting fruit, which serves as food for both the larvae and the adults. The ease with which large numbers of these insects can be reared in the laboratory during the winter as well as the summer, and the definiteness with which they react to many forms of stimuli, make them favorable subjects for experimentation. Since they find their food with great certainty even in the dark, a habit that seemed to involve the sense of smell, I was led to take up an investigation of their reactions to odorous substances. Where the flies were abundant, it was noticed that they often entered bottles and other receptacles containing alcohol. ‘The fact that the fermenting fruit upon which they feed is continually generating alcohols and other related compounds, led me to sus- pect that it was these substances that served to attract the flies, and that they therefore probably presented a clear case of chemotro- pism among air-inhabiting animals. Tue JourNaL or ExPERIMENTAL ZOOLOGY, VOL. IV, NO. 4. 516 William Morton Barrows The experiments recorded in this paper were undertaken to determine, if possible, first, to what substances Drosophila is chemotropic, and secondly, in what way the fly finds its food. The work was carried on in the Zodlogical Laboratory of Har- vard University, under the direction of Prof. W. E. Castle and Prof. G. H. Parker, to whom I am indebted for much valuable advice and careful criticism. “II EXPERIMENTS 1 Preliminary Experiments For the preliminary experiments, which were planned to ascer- tain whether certain substances were stimulating or not for the flies, the following apparatus was devised. A two-dram vial closed at the end by a cork stopper was arranged asa trap. Pierc- ing the stopper and reaching nearly to the bottom of the vial was a glass tube with a caliber of about two millimeters (Fig. 1). The bore of the tube was large enough to allow a fly to creep through easily and yet small enough to make it difficult for the fly to turn around after having once started into the tube. ‘The tube pro- jected beyond the cork on the exterior about three millimeters. If a fly once got halfway down the tube leading into the vial, the chance of its backing out or finding its way out later was very small About 1 cc. of the substance to be tested was placed on a piece of filter paper in the vial. Five vials thus charged with substances to be tested usually formed the set of traps. At least one of these was always used as a control in that it contained filter paper wet with distilled water only or with some other material used as a check. The traps, with their open ends directed toward the light, were placed in a vertical glass cylinder 20.5 cm. high and 17.5 cm. in diameter. ‘The bottom edge of the cylinder rested on a sheet of clean filter paper and the top was closed by a glass plate. The atmosphere in the cylinder was kept moist by the evaporation of distilled water exposed in a small vessel Many hungry flies, usually one hundred, were liberated in the glass cylinder and left there for twenty-four hours. By hungry flies is meant those which Reactions of the Pomace Fly to Odorous Substances BUT had been supplied with distilled water but had been kept from food for twenty-four hours. If they are kept without food much longer than this, they begin to die and few survive sixty hours. After the flies had been allowed twenty-four hours in which to enter the traps, the experiment was discontinued and the indi- viduals in each trap were counted. In this way some idea of the influence of different substances on the movements of the flies could be ascertained. These experiments were preliminary, in that they aimed only to determine what substances called forth positive reactions. ‘The Fig. 1 Vial arranged as a trap. results, which are necessarily fragmentary, because of the difficulty of dealing with odors in a quantitative way, are given in Table I. From Table I it is apparent that of the ten substances which were tested singly in aqueous solutions, the flies gave definite positive reactions to four, namely amyl, and especially ethyl alcohol, acetic and lactic acid (Experiments 1, 2, 3, 5,6). ‘The remaining substances to which the flies did not react positively or did not react in numbers large enough to be significant, are propyl alcohol, butyric acid, valerianic acid, glucose, amyl valerianate and acetic ether. Experiments which were undertaken subse- 518 William Morton Barrows TABLE I Numbers of flies out of one hundred, which, in each of eight experiments, entered the several traps charged with odorous substances. The duration of each experiment was twenty-four hours No. of Experiment Substance in the trap Navel Hic trap HimpeAcctiomcitiha percent. sa sy.t ee nual ste s. ae DM eoholerouper centysarel a eerackea ooo nee O eet 4 ae Ru Wiateri(control)sreinncamerpticastitar ee oe eee I 4 Acetic acid 4 per cent with trace of aceticether............ 5 5 Glucose and waterscsstrversite shiner caine erelter eee roe ° | TpAceticacidl4 pencenthw. verses ose ee eee | I UieowAl coholaospencentaysee-ture career a eee tree Boe Ri Water (control) ie aerate Neich Civ ae aeR eee he ore | 4 Acetic acid 4 per cent with trace of aceticether............. 61 ib Glucosé:and waters ccna cise bere reer eee ° 1 Alcohol ro per cent containing acetic ether .og per cent..... 85 2rAlcoholito;per centescrr itm ter eae er eee ite 12 oar 3° Water (control)).5 2.28 ton atntaet tees ah serra ro) A Aceticiacidia pericentsn) waste eit tae ee eee ete rere 2 5 Acetic acid 4 per cent containing acetic ether .04 per cent... 19 1 Alcohol ro per cent with trace of butyric acid.............. 43 2 Alcohol ro per cent containing 2 per cent hydrochloric acid. ° Bienes patentee elses 3, Aleoholixo'percent (control) 524.5 siltasace nese te wn anotr 26 4 Alcohol ro per cent with trace of valerianic acid............ 14 5 Alcohol io per cent with trace of isoamyl acetate........... ° 1 Water with trace of valerianie acid’... <.05- <1 --ea eee fe) 2) Butyriciacid 2 per cen tier..,.02, se) 15 Set eee eae I Loenee sors aeheraetine a: Water (controll) pry ate. nat yraiteiiae eae ieee penioe neers 2 | A Eropylialconol)2) percentawcns: (ssi ina teek etcetera ° 5 Water with a trace of amyl valerianate................... fo) ( ToWaterwithiaitrace of aceticethen eesti eee eee | fo) \ lg Tate a tedeniperiennt.' Aoi t cu sory Bao eee aes 6 6.. q Water (controll)’ scocctae. sue accra iy renee er eee bee fo) 4 Amylialcoholes pen centerrasat tee rele lel vrei anit reer 3 1 Alcohol ro per cent with a trace of acetacetic ester ......... ° | 2 Alcohol ro per cent with a trace of isobutyl acetate ........ 9 Tittle rehire ne a @Alcohol re;pericent|(Control) yam vies ine eee | is 4 Alcohol ro per cent with a trace of methyl acetate.......... 29 5 Alcohol 10 per cent with a trace of isoamyl acetate......... | 8 ( 1 Alcohol ro per cent containing .o5 per centosmicacid...... | | 2 Alcohol ro per cent containing 2 per cent hydrochloric acid. ° Chae Santon Sacer | 3) Alcoholimo. per cent|(controll) aeeeerer ease ee I | 4 Alcohol 10 per cent containing 2 per cent nitric acid........ 5 Alcohol ro per cent containing 2 per cent acetic acid........ 8 Reactions of the Pomace Fly to Odorous Substances 5,19 quent to those under discussion, however, showed that the flies did enter or attempted to enter traps which contained acetic ether in aqueous solution, so that it 1s possible that under other conditions the flies might be positive to some of the substances to which in these preliminary experiments they seemed not to be. In Experiment 2 one trap was charged with 4 per cent acetic acid and another trap with 4 per cent acetic acid containing a small amount of aceticether. ‘lhe number of flies found in the trap which contained the mixture of acid and ether greacly exceeded that in the trap charged with acid only. A similar contrast was observed in Experiment 3. ‘This increase of flies in the trap containing the mixture is evidently due to the acetic ether. A similar phenomenon was seen when to Io per cent ethyl alcohol, a small amount of acetic ether, isobutyl acetate, or methyl acetate was added (Experi- ments 3 and 7). Isoamyl acetate may possibly also be classed with these substances, though the results of Experiments 4 and 7 do not show an entire agreement in this respect. An increase was also obtained when acetic or butyric acids were added in small amounts to the alcohol. ‘This property may be slightly shared by valerianic acid and possibly by osmic acid, but it is certainly not a characteristic of hydrochloric and nitric acids which seem to be strongly repellent (Experiments 4 and 8). All the organic substances tested in these preliminary experi- ments are found in fermenting fruits and the test conditions which gave the highest positive numerical results are probably those which simulated most closely the natural optimum conditions. 2 Experiments with Alcohol, Acetic Acid and Acetic Ether As a more complete analysis of the effect of some of the odorous substances used in the preliminary experiments was desirable, it was decided to test more fully acetic acid, ethyl alcohol and acetic ether. [hese were chosen because they are commonly found in fermenting fruit, the first two in quite large quantities and the third in traces. “lo make the tests more accurate, a new piece of apparatus was constructed in which two traps were so placed in the sides of a leaden trough that each fly as it passed 520 William Morton Barrows through the trough had a chance to enter the charged trap or the check trap or to react to the odor issuing from the former. The plan and elevation of the apparatus used are shown in Figs. 2 and 3, respectively. On a wooden base 4, some 30 cm. in length Fig. 2 Fig. 3 Figs. 2 and 3 Plan and elevation of apparatus used in testing the reactions of Drosophila to odors. A, wooden base; B; leaden trough; C, zinc slide for closing exit from the receiving chamber, E; D, glass } plate; E, inverted glass dish forming a receiving chamber; F, inclined way; G, inverted glass dish form- ing a collecting chamber; H, hole in glass plate serving as entrance to collecting chamber; J I’, entrance ; tubes to trans; F, suction tube. and 12 cm. in width, was mounted a leaden trough B, which was 14 cm. long and about 2.5 cm. broad except at the far end (right in the figure) where it expanded into a chamber 3 cm. square. The Reactions of the Pomace Fly to Odorous Substances §21 passage in the trough was 5 mm. deep, and rr mm. broad and extended from the near end, which was closed by a zinc slide C to the square chamber at the far end. ‘The trough was covered bya glass plate D, which fitted to the lead Slosdie enough to make it practically air-tight. Against the near end of the trough was an inverted cylindrical pies dish £, which served to hold the flies to be tested. ‘This dish was peed to the level of the glass plate by a block of wood. Its chamber communicated with the passage of the trough by a short inclined way F, which allowed the flies to pass into the trough when the slide C was open. At the opposite end of the trough and resting upon the glass plate was another inverted glass vessel G, communicating with the trough by means of a small hole H through the glass plate. This vessel served as a reservior to hold the flies that had been tested. A fly creeping from chamber £ along the trough to chamber G passes between the open ends of the two traps, which were inserted opposite each other through the walls of the trough at / and /’.. A small glass tube piercing the far wall of the chamber at F was con- nected by a rubber tube with a suction apparatus, by means of which a current of air could be drawn through the trough at any desired rate. he suction apparatus consisted of a large bottle filled with water, closed at the top by a stopper with two holes. Through one hole was inserted a bent glass tube, which served asasiphon. ‘The other hole was filled by a short glass tube which connected with the rubber tube from ‘f and served to admit the air under external pressure to the partial vacuum formed by the siphon. Reference to Figs. 2 and 3 will show that when the siphon was allowed to run at a given rate, controlled by a clamp on the rubber tube, a current of air flowed from EF through the trough, past the ends of the traps to the outlet . The aim was to have this air current carry all the escaping odorous particles away from the mouth of the trap. To test the apparatus, hydro- chloric acid was allowed to evaporate in chamber FE and this gas was drawn by the air current along the trough and past one of the traps which was charged with anne water. White fumes of ammonium chloride were formed at the mouth of the trap /, and deposited along the path of the current. The dotted line in Bie, 522 William Morton Barrows marks the edge of the current of ammonium chloride, which shows that none of the ammonia moved against the current toward £. In using the apparatus it was so placed that the flies, which are positively phototropic, would creep under the influence of light from chamber £ to chamber G. A number of hungry flies were placed in chamber £; the current of air was then started; and the two vials were placed in position, one containing the substance to TABLE II The numbers of flies which reacted positively to each of eight different strengths of ethyl alcohol. In each experiment the number of flies used was one hundred e+ | Number of flies that | ‘S & iS) ig 2 5 Number of flies that ga 8.8 See ces turned toward but Paes g gy Se wm OH entered the ae | oO omeey Number of the | © % | did not enter the | g 2 5 ge oo ° a =! Qu =} experiment | <4 .& =] ) eter ce: Begs hl mabe | palin Stecc San | a> Alcohol | Control | Centrol | Alcohol | 3B 3S BSS | ies! | | Oo 8 «a oH ees trap | trap | trap =| trap | = | I 100 ° ° ° ° ° ° Dricntscmeiaein tes 75 ° I ° I ° 2 ORO RE AO BAOEE 50 ° 3 Te) ° 10 3 4 25 2 ° | 15 I 4 a Gidooceunacogs- 25 8 2 | 3 fo) ops eae emionio od 20 ° ° 16 ° 7 9 HSAs cree wees I 2 I 6 ° 7 5 II I lo Spiced HOD Beao 15 2 ° 12 I eyavelerelope a samaiere [eee ° I I ° 2 | 7-5 I fe) 10 10 I 4 ° DU ts Grete tone. tse 5 ° 3 o | I | Ty, 2 ° 3 fr) 5 3 | 3 1.75 Aes epee 5 I I I | 2 PA actioud ster 5 ° ° 5 | ° | | | | MG sie tere yee ray= tote water | ° I 2 | 2 z 3 | be tested, the other containing water used asa control. ‘The slide was then opened far enough to allow a few flies at a time to pass down the trough toward the light. The flies, that reacted posi- tively by turning abruptly toward the traps, were counted as were those that entered either of the traps. After fifty flies had been admitted to the trough, the slide was closed and the traps were interchanged. Now fifty more flies were allowed to pass through Reactions of the Pomace Fly to Odorous Substances 523 the apparatus and their reactions recorded. ‘These records added to those of the preceding fifty constituted the records of one experi- ment. When the different strengths of the same substance were tested on different days, the last strength used on the previous day was first tested in order to be sure that the hungry flies in stock had remained uniform in their response to this stimulus, Having ascertained this, the experiments were carried forward as though they formed a continuous series. ‘The results of these experiments are given in Tables II to VI. TABLE III The numbers of flies which reacted positively to glacial acetic acid and to different strengths of this acid in water. In each experiment the number of fltes used was one hundred | . | ; ‘Sow 7 | Number of flies that | ‘o Weror ce ws os. Number of flies that | oS 2 oe 3 0 Sih turned toward but | & ¢ lanier ba} entered the : oS 2 Number of the | ‘SO &, did not enter the Berane g 2 3 5 a ov 5 experiment a AS. | o-8 g oP as Acid Control Control | Acid gaat lags os | alesse coi z trap | trap trap | trap eS | = Rin glacial ° ° 5 I Pare 1 Doss 50 fo) ° 6 ° 6 ° plot 2 fe) fo) 12 3 5 | 7 Il 4 Arse: 25 ° fo) | 10 I Sone 20 fo) ° 7 co) 7 fo) Tree 15 I | 12 fo) 13 I stats 10 I fo) 25 fo) 26 ° S228 5 18 | I 16 fo) 34 | I 9... 3 ° 9 ° 12 | ° 10 I ° 2 fe) 3 | fo) | From Table II it can be seen that the greatest number of positive reactions to alcohol were obtained at 20 per cent concentra- tion, while strengths above or below this grade show a decrease in the number of positive reactions. It will be seen from Fig. 2, that a fly in passing from the near end to the far end of the trough enters the area of stimulation obliquely. Consequently one side of the animal must be stimu- lated before the other. Many of the flies entering the odor in this way give a positive reaction by turning toward the stimulated side. 524 William Morton Barrows This reaction indicates that the flies respond to a difference in the intensity of the stimulus on the two sides of their bodies. The following peculiar response was very often observed. After flies had passed the traps they would often suddenly turn and run back with a characteristic zigzag motion to the mouth of the charged trap. While testing with 50 per cent alcohol, it was noted that about one-third of the flies reactedin this manner. Evidently they are able to follow the current of odor back to its source. These responses will be further discussed in a subsequent part of this paper. TABLE IV The numbers of flies which reacted positively to each of three different strengths of acetic ether in water. In each experiment the number of flies used was one hundred . 3 = ESE ate a Number of flies that) S & Ro) g = o umber 0 1es that = f= cs aie turned toward but Be oo ee Q entered the ; | a 45) | ane Number of the Oo a did not enter the Eg 2 5 5 2 S E a & fia Sele al ms = °o experiment = 5 Nl es 5 AEs a Ether Control Ether Control | = a ee = S ra un S Roa 5 4H Ss F trap trap trap trap = = Trace 8 ° 7 33 6 !\ 19 a | | / Boos 8 ° I 5 | ° If | i< OE 4 ° ° 9 fo) \ 8 : 7 ne 4 ° ° 7 2) { Gee 2 ° fo) 13 fo) \ 11.5 ? Dade freee 2 ° ° | if) 2 | f It is plain from an inspection of Table III that the largest number of positive reactions was produced by 5 per cent acetic acid. Not only is this true, but, when the trap was charged with this strength, about half of the flies which at first failed to respond to the stimulus returned through the current of odorous material from the far end of the trough back to the mouth of the trap. These responses are not included in the table. Acetic ether is soluble in water only to the extent of about 8 per cent, and when used in such high concentration it affects the flies in a singular manner. ‘They show intense excitement and struggle at the mouth of the trap for a chance to enter, yet when one has succeeded in entering it backs out almost immediately. Reactions of the Pomace Fly to Odorous Substances 525 It is probable that the dissolved ether evaporates rapidly forming an almost saturated atmosphere inside the trap, and this is known to kill the flies in less than three minutes. Hence the excessive stimulation probably causes them to back out of the entrance to the trap into which they had been enticed by the less concentrated vapor. Acetic ether is never so abundant in decaying fruit as in the weakest solutions (2 per cent) tested in these experiments. In order to make a mixture which should combine the optimum strengths of alcohol and acetic acid, equal volumes of 40 per cent alcohol and Io per cent acetic acid were mixed. ‘This mixture then contained 20 per cent of alcohol and 5 per cent of acetic TABLE V The numbers of flies which reacted positively to each of four different strengths of a mixture of equal parts of 40 per cent alcohol and 10 per cent acetic acid diluted with water. In each experiment the num- ber of flies used was one hundred eo Number of flies that} 6 x # | 6 & 2 Riek #30 : “=a A 5 aon Number of flies that ; go 38 MK oo turned toward but | y 5 2 Cucus en entered the ' tay SL wey Number of the > did not enter the hfs z oe are 2 a . ia lee gi Saomic experiment = 8 —— |) BS see ee i aa 1 SS | | eeeemn connec. Se Olea ore 5 Charged Control Charged | Control § © E s 2 g qi trap trap trap trap | = I 100 13 fo) 17 4 30 4 2 50 26 5 2 ° | 28 5 3 25 29 I 2 ° 31 I 4 T2015 27 5 12 ° 39 5 acid. “Table V shows the results obtained by testing flies with this mixture either pure or diluted with water. A solution of 124 per cent of this mixture, which is equal toa mixture of 24 per cent alcohol and 3 per cent acetic acid, gives a slightly higher number of positive reactions than is given either by 5 per cent acetic acid (Table III) or 20 per cent alcohol (Table II). The numbers of the positive reactions are not significantly large, yet it is probable that the mixture is uniformly more stimulating than alcohol or acetic acid alone. Table VI shows the results obtained by testing the mixture con- taining 20 per cent alcohol and 5 per cent acetic acid (Table V) to 526 William Morton Barrows which had been added 8 per cent of acetic ether. Of the dilutions used, 12} per cent of the mixture induced the largest number of flies to react positively. It is probable that the experiments were complicated by the presence of a higher per cent of acetic ether than is met with under natural conditions. In Table I, Experi- ment 3, about .o4 per cent of acetic ether was added, respectively, to 10 per cent alcohol and to 4 per cent acetic acid, and in both instances there was an immense increase in the number of re- sponses as compared with the responses to those reagents alone; this increase must have been due to the slight amount of acetic ether present. We may safely conclude that acetic ether probably plays some part in the reactions of Drosophila to normal food. TABLE VI The numbers of flies which reacted positively to each of three different strengths of the mixture of alcohol, acetic acid and acetic ether. In each experiment the number of flies used was one hundred Number of flies that} — ' ~ _— t Sie.) TiNambercok dies ch ~ jae |) ae ae ror umber o ies that Si aa turned toward but} § 5 Gh aha we ) a2 f= entered the ; 2) 4 1 Soa Number of the! ° i did not enter the cinta g 2 : aS =_—s —_es es : —_— =) =] experiment s, ae, Bacar rs 3 | | iS} & = € | Charged | Control | Charged | Control | 3 § § | = SS eet ort | So Hw OS 3S HY o trap trap trap trap an a | Hoarodnd apiwiKe sou ICO | ° I 4 ° 4 I Defic aeteis: Mecekecrell 12%. | 13 2 7 fo) 20 3 aSaoGCDOO TIS. Ge 6+ 4 ° 5 ° 9 The foregoing experiments show that Drosophila is positively chemotropic to alcohol, acetic acid and under certain conditions to acetic ether. The optimum strengths of alcohol and of acetic acid are 20 and 5 per cent, respectively, while that of acetic ether is uncertain, but must be only a fraction of I per cent. Table VII, made up from data given by Leach (’05), shows that alcohol and acetic acid are commonly found in cider vinegar, fermented cider, and California sherry in per cents that are close to those which call forth the largest numbers of reactions in Droso- phila. Acetic ether is found in these fluids in very slight traces. Reactions of the Pomace Fly to Odorous Substances ney 3 Experiments on the Directive Effects of Odorous Substances Having determined that these flies are chemotropic to fermenting fruit, | turn to the second question, In what way does the fly find its food? To ascertain the accuracy of flight toward the food, experi- ments were carried out in the following way. About one hundred hungry flies were liberated in a large laboratory room. A few minutes after their liberation a tumbler, freely exposed on the top of a table and containing fermenting banana was opened. As the hungry insects in flying through the room passed near the table they eventually discovered the banana. When they were six or more feet from the tumbler they showed a rather characteristic TABLE VII Amounts (in per cent) of acetic acid, and alcohol found in cider vinegar, fermented cider, and California Sherry (Leach ’05) Acetic acid in per eent Alcohol in per cent Number Substamees © | ———2@—-——-—___——__ | — —$$____—_——- ; | | ; of samples Max. | min. Av. | Max. | Mun. Av. Cider vinegar.... 7.61 | 3.24 4.65 | 44 Fermented cider.| 6.59 24 z518 |. (6285 Tat 4.72 16 California sherry | 79 25 21.85 8.22 66 vibratory flight. Short rapid excursions were made through the air, up and down, forward and backward, right and left. Sometimes the fly came nearer the tumbler and under such circumstances it often remained in its vibratory flight in this new situation. As it approached to within about three feet of the tumbler the excursions shortened and the fly oriented more accurately to the source of the odor, though the flight still showed considerable vibration to right and left while the head of the fly was directed generally toward the tumbler. When about two feet from the tumbler, the vibratory movements grew less and less extensive and the flight became more rapid and more accurately directed toward the tumbler. The last six or eight inches of the journey was made in nearly a straight line to the edge of the tumbler or to its base. 528 William Morton Barrows It 1s a source of continual surprise to see how accurately and quickly many flies will find food. Not only do the flies find food easily and certainly during flight, but they can also find it successfully when creeping. To test this the following experi- ments were tried. A small piece of fermenting banana was placed on a glass plate one inch square and the glass plate with the banana was put in the center of a square sheet of paper ruled into 25 squares each one an inch on a side (Fig. 4, ato b). ‘The glass cylinder used 1n the first experiments (p. 516) was then placed over this paper in such a position that the four corners of the paper just touched the lower edge of the cylinder. To prevent air cur- rents from driving the odorous particles away, the chamber was closed by a glass plate above. Hungry flies were admitted singly at the bottom of one side of the chamber. ‘They flew as usual toward the light side and upward, but in the course of a minute started to creep down the glass toward the bottom of the chamber. As they came on the paper their course was carefully plotted on a duplicate sheet of ruled paper. ‘The courses given in the diagrams (Fig. 4) show the characteristic paths traversed by twelve flies. ‘These paths showthat inmost cases the flies took the most direct route in order to reach the food, 7.¢., they did not merely run upon it by chance. ‘The paths are usually so direct that it appears as if the flies found the food by sight, but that this is not so, is shown by subsequent experiments, in which after the removal of the antennz the flies seldom found the food, though their eyes were intact. It is clear that both in flight and in creeping the movements toward food are at first irregular and afterward more accurately directed, so that the fly eventually takes an almost straight course to the food. ‘The beginnings of the courses in flight and in creep- ing are such as to suggest the trial and error method of response. ‘This view is supported by what is sometimes seen in the zigzag course taken in the return of flies against the current of odor in the trough in the experiment described on p. 524. “The conclusions of the courses however in both flight and creeping are so accurately directed toward the food that trial and error can play no part in explaining this condition; one is forced on the contrary to assume Reactions of the Pomace Fly to Odorous Substances 529 6 21 up Q s a e. & gh. ii 4h. Hi Fig. 4 The irregular lines represent the paths traversed by twelve flies in reaching a piece of fer- menting banana placed in the center of the area. The area was § inches square and the fly almost always started from some point on the outer edge of the area. 530 William Morton Barrows some such method of orientation as is implied in the theory of tropisms. ‘The more usual conception of the tropism theory, as advocated particularly by Verworn, is to the effect that when an animal is unsymmetrically stimulated it turns until it is symmetric- ally stimulated and either faces toward or away from the source of the stimulus and then moves in the appropriate direction. It is evident from this that symmetrical stimulation is an essential feature of this theory. Loeb has extended this view in the sense that he often implies that the stimulus acts directly on the loco- motor organs, but I do not regard this as an essential part of the tropism theory, and, as I shall show presently, this modification of the theory has no application in this case. The question, then, is, are the accurately directed responses of Drosophila dependent on symmetrical stimulation? If this is the case, one would natur- ally turn to the antennz of this animal as the symmetrical recep- tive organs of smell for such reactions. 4. Experiments to Determine the Position and Function of the Olfactory Sense Organs It is generally believed that in most insects the antennz ate the seat of the olfactory organs. In some species these organs are placed in pits, whilein others they are exposed on the surface of the antenna. Mayer (’79) described a pit in the third segment of the antenne of a species of Drosophila which he considered an olfactory organ. I have found that Drosophila ampelophila has a large sac-like pit situated in the end of the terminal! (third) segment of the antenna, which contains sense cones. Fig. 5 shows a front view of the head of Drosophila and the position of this pit in the left antenna. The following experiments on flies which had been deprived of the third segments of their antennze show that without doubt the olfactory organ in Drosophila is located in this segment. It was found by repeated trials that the antenne could not be satisfactorily covered with gum to keep out the stimulating odor, nor could they be burned off without considerable injury to the fly. ‘The method finally employed was to place a fly, already etherized,on its back on a glass slide and to hold it down with a small camel’s- Reactions of the Pomace Fly to Odorous Substances 531 hair brush, which in turn was held in place by a small rubber band. Secured in this way under the lens of a high-power dissecting microscope, the third joints of the antenne were cut off by a pair of fine embryological dissecting scissors. There is a deep division between the second and third segments of the antenna, and the third segment was usually removed without injury to the second. After the effect of the ether! had passed away, the fly operated upon seemed in most respects perfectly normal. Such flies were left for twenty-four hours without food, but were supplied with Fig. 5 Front view of the head of Drosophila showing in dotted lines the position of the olfactory pit in the terminal segment of the left antenna. water. At the end of this time they were liberated singly in the large cylinder used in the experiments described on p. 516, and were carefully watched for five minutes in order to be certain that their behavior was normal. Having ascertained this, their ability to find a piece of fermenting banana in the cylinder was tested. The time required to find the food was recorded, or if the food was not found in fifteen minutes, the experiment was discontinued. 1 Tt should be noted that the process of etherizing has no lasting effect on the ability of the flies to scent food, 7. e., normal flies after recovery from ether find food with as great certainty as they did before etherization. 532 William Morton Barrows Of fourteen flies which were thus tested all with one exception failed to find the food within fifteen minutes. In the exceptional case food was found in twelve minutes, but the insect’s course was such that it obviously came to the food by accident. During these experiments the flies were carefully watched, and although many of them came within 1 cm. of the food, they did not find it. To test this matter further, four of the flies with defective anten- nz were allowed to wander in a small glass tumbler until they found the food, which they apparently did by accident. When one foot of the fly touched the food or a small drop of water the tongue was immediately put down and the animal began feeding. It is not impossible that two transparent hairs which are found beneath the claws of each front foot, and have the appearance of sense hairs, may be instrumental in giving rise to this feeding reaction. It therefore seems certain that the sense of smell is absent, or at least greatly reduced, in flies which have lost the terminal joints of both antennz. In order to determine the rela- tive time taken to find food by flies with and without the terminal segments of the antenna, six normal flies were admitted to the chamber and the time recorded which elapse before they reached the food. ‘These flies were then operated upon; the distal seg- ments of both antennz were cut off and they were allowed to rest twenty-four hours, when they were again tested and the time similarly recorded. ‘The results of these experiments are given in Table VII!. From this table it will be seen that the normal hungry fly finds the food in about two and one-half minutes, while the same fly after having beer operated upon seemed unable to locate the food. We may conclude: first, that Drosophila does not find its food by sight, but by smell, and when this sense is lost it reaches the food only by accident; and, secondly, that the olfactory sense organs—at least those which are concerned with finding food—are localized in the third or distal segment of the antenna. The fact that in Drosophila the antenne are the principal organs concerned in the reception of olfactory stimuli and that they are symmetrically placed on the body of the animal, leads to the conclusion that these flies orient to odorous centers in the way Reactions of the Pomace Fly to Odorous Substances 533 assumed by the tropism theory. Forward locomotion would be called forth by an equal stimulation of the two antennz and lateral movements by an unequal stimulation of these organs. If this view is correct, circus movements ought to result after the removal of one antenna even though the stimulating atmosphere contains a uniformly distributed odorous substance. It was, therefore, thought desirable to experiment upon flies from which one antenna had been removed in order to produce excessive unilateral stimulation. Before the operation the flies were tested for five minutes in pure air and five minutes in an atmosphere with odor, to make certain that they were normal, 7. e., that they did not turn more TABLE VIII Records of the times which six flies took to find food before and after the terminal segments of their antenne had been removed | Time in minutes in which the Numbers of the flies | — = — - - ——- Normal flies found the food Injured flies found the food li Bema a See o RECO Zot iiss totes 3 SRS epee ee | 4.0 Bieter icteric ice ee mc7/5 Flies all failed to find the food at the [eo ort ORO OTS Oe * R15 end of twenty minutes. Beeler ayer oneeaus We chase eh 2.0 Dost die Soe eee Ione 1.0 PNET ACER le rare avers [ais craven oie sistsis 2.45 frequently to the right than they did to the left or vice versa. ‘The terminal segment of the right or the left antenna was then cut off from each fly. After the removal of this segment, the flies were fed and allowed to remain twenty-four hours, when they were again hungry. They were then admitted singly to the cylinder containing only pure air and watched for five minutes. Without exception they reacted as they did under similar circumstances previous to the operation. A little odor was now blown into the cylinder from a wash bottle partially filled with fermenting banana and the flies were again admitted singly and their movements carefully watched. If the fly moved in a circular path, as was 534 William Morton Barrows usually the case, a record of one circle was put down for the fly when it had turned continuously through an arc of 360 degrees in one direction. A circle was called positive if the fly moved with its normal antenna on the side toward the center and negative if it moved in the opposite direction; 7. e., a fly having its left antenna cut and moving always with its right side next the center would be said to be describing a positive circle, etc. “Table LX shows the records of twelve flies tested in the manner described. TABLE Ix The numbers of positive and negative circles made by twelve flies which had only one functional antenna. Each fly was tested singly for a period of five minutes in a uniformly odorous atmosphere The number of circles made ina Numbers of the flies Normal antenna positive negative | direction direction eaaoeavaceoueno pp Geaaa dive Kodo monmon soa right 6 ° eos aqnoascEMdoUsooponsodeiaos sdnanbos tae left 4 ° [IROOM AOSD DOE e ae coo Haar ccna cc asa hatemIeI right 4 ° ligGbo 5.00.61. 00 DOD EAen tse Oe scace on an aw ORO Ok left 3 ° ROBO PECGOs AoC or Ue add duc a Foun aD ono Oe left 5 I Ped oN oD Avian oo aPE ROTOR SOUL QO Una O OGeIAR OD I9 OC right 5 I Ac each a ahege era ciate cc et NERRYS Meio eda eg redey eee ios tors right 5 ° Gt ah MU AA Jee Beek ime SERRE Bap Me, Sar left 2 ) LE Geen Bin RIGGS AO ORS Oto ONO ETE ee oa eo | left 2 ° BRP ee eu Ae eo ate tsa ay Mors Nees SLO retiss left 5 ° (Tere ait ch Dh Ooo eck aeciie acto waee ec left 2 ° TG Pe aoe ote eee eae TORO ate ele eas left 3 2 AD Obal Si sre ct-p PNR VONIAS yoeoe iether Velnin: easel ave eae a PES oye 46 4 From this set of experiments it will be seen that forty-six out of fifty of the circles, or 92 per cent, were made in a positive direction, 1.e., toward the normal antenna. As this antenna is obviously the one stimulated, it is clear that the flies must orient to unequal, unilateral stimulation.’ 2 Since this paper was written Kellogg (’07, p. 153) has recorded circus movements in the males of the silkworm moth after the removal of one antenna and on exposure to odors. Reactions of the Pomace Fly to Odorous Substances 535 III THEORETIC DISCUSSION The experimental results recorded in this paper show very con- clusively that the reactions of Drosophila to odorous matessats are by no means uniform, but vary in method under different circumstances. When the stimulus is very weak little more than random movements are excited, but when the stimulus is some- what stronger trial and error movements gradually prevail, where- by the fly becomes approximately oriented toward the odorous material, much as has been emphasized for many lower animals by Jennings (04). Finally the orientation to the odorous material becomes very accurate and the fly may be said to take an almost direct course to it. It is clear that the latter part of the course is accomplished by methods in the main free from anything that can be described as trial and error. Since under a like degree of stimulation flies, after the loss of an antenna, carry out circus movements with great regularity, it seems impossible to explain the movements under these conditions in any other way than on the basis of the tropism theory. This theory has been stated in several ways. As applied to chemical stimulation Verworm (’99, p. 429) declares: “The word chemotaxis is applied to that prop- erty of organisms that are endowed with the capacity of active movement by which when under the influence of chemical stimuli acting unilaterally they move toward or away from the source of the stimulus. Where there is an approach to the source of the stimu- lus, there is postive chemotaxis, where there is a removal from the source negative chemotaxis. Unilateral stimulation with chemical stimuli is only realized when the concentration of the substance in question gradually increases from the living object in one direc- tion.” The method by which Drosophila finds its food is directly com- parable to that observed by Harper (05, p. 33) in the reactions of Perichzta to weak and strong light. ‘This earthworm orients away from the source of a weak light stimulus by frequent random movements, z.¢., by the trial and error method. But when the light stimulus is greatly increased the orientation is direct, random movements toward the light are suppressed altogether and the 536 William Morton Barrows worm appears to move directly away from the light without notice- able trial movements. It seems to me probable from experiments described by Pearl (03, pp. 623-670) that planarians may follow some such method in finding food. However, as the animal is not highly special- ized, the distance through which it can orient accurately is small and the result is not striking. IV SUMMARY 1 Drosophila ampelophila is a small fly peculiar for its fond- ness for fermenting fruit. . 2 These flies are positively chemotropic to amyl and especially ethyl alcohol, acetic and lactic acid and acetic ether. 3 Acetic ether, isobutyl acetate and methyl acetate, when added in small amounts to 10 per cent ethyl alcohol, greatly increase its attractiveness. A similar increase is noted where acetic or butyric acids are added to the alcohol. All these organic sub- stances are found in fermenting fruits. 4 The optimum strengths of ethyl alcohol and acetic acid as determined by the number of positive reactions given to different strengths is 20 and 5 per cent, respectively, while a mixture con- taining 24 per cent alcohol and 3 per cent acetic acid gives a slightly higher number of positive reactions than is given by either 5 per cent acetic acid or 20 per cent ethyl alcohol. Alcohol and acetic acid are commonly found in cider vinegar, fermented cider, and California sherry in per cents that are close to those which call forth the largest number of reactions in Drosophila. 5. Drosophila does not find its food by sight, but by smell, and when this sense is lost it reaches its food only by accident. ‘The olfactory sense organs—at least those which are concerned with finding food—are located in the third or terminal segment of the antenna. 6 When one antenna is lost and the other antenna is stimu- lated by food odor, circus movements are carried out in such a way as to prove that the fly orients normally by an unequal stimulation of the antenne. Reactions of the Pomace Fly to Odorous Substances ely) 7 Drosophila, when stimulated by weak food odor, first shows random movements, 7. ¢., it attempts to find the food by the method of trial and error, but as the fly passes into an area of greater stimu- lation, these movements give way to a direct orientation. ‘This orientation is a well defined “tropism” response. ‘These reactions of Drosophila are paralleled by those of Perichzta to strong and weak light and possibly also by the food reactions of planarians. V BIBLIOGRAPHY Harper, E. H., ’05—Reactions to Light and Mechanical Stimuli in the Earthworm Pericheta bermudensis (Beddard). Biol. Bull., vol. x, no. 1, pp. 17-34. Jennincs, H. S., ’04—Contributions to the Study of the Behavior of Lower Organ- isms.. Carnegie Institution of Washington, Publication no. 16, 8vo, 256 pp. Kettoce, V. L., °07—Some Silkworm Moth Reflexes. Biol. Bull., vol. xii, no. 3, Pp- 152-154. Leacu, A. E., ’05—Food Inspection and Analysis. New York, 8vo, xiv+ 787 pp., 40 pls. Logs, J., ’97—Zur Theorie der physiologischen Licht- und Schwerkraftwirkungen. Arch. f. ges. Physiol., Bd. 66, Heft 9-10, pp. 439-466. _ Mayer, P., ’79—Zur Lehre von den Sinnesorganen bei den Insecten. Zool. Anz., Jahrg. 2, No. 25, pp. 182-183. Peart, R., ’03—The Movements and Reactions of Fresh-water Planarians: A Study in Animal Behavior. Quart. Jour. Micr. Sci., vol. xlvi, pt. 4, PP 5°95 /14- Verworn, M., ’99—General Physiology. Translated by F.S. Lee. London, 8vo, xvi + 615 pp. ws f ‘ 4 2 a , , . s ; - Cree : age - fd A a Z ae Uy, ~ ‘ j ow + i, a ™ ‘ q hoo . +, a 7 5 Veer ey } a Lae . : fi " ae i ’ ai * be a) Pr é [ Ks 7 ‘ 5; ae Sree et sa aes a ry oe ie \ ¢ Ra) he a ia os ¥ Pa ‘—on wm the why © - A / ef Jas THF ty VROTA MORE UT AtP Ee eek eee 27 . ~ - ‘ in ( seat va ter Sees Cotes ICA MPS iar J 7 ' : bg rr pow 7. 7, > i > TAM ae at —~* Ld ) ‘oh ; i «i? i Joe , - : P rT 7 - . + . . = ¥ . me v » = ite. : . ts Z 7 | . . 7 ae ° . >. * / CONTRIBUTIONS FROM THE ZOOLOGICAL LABORATORY OF THE MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD COLLEGE. E.L. Mark, Director. No. 192. ii hE Ch OF TEMPERATURE ON THE: MIGRA- TION OP Tite RETINAL PIGMENT IN DECAPOD CRUSTACEANS BY EDGAR DAVIDSON CONGDON With SEvEN Ficures I INTRODUCTION The last few decades have witnessed a gradual accumulation of knowledge concerning photomechanical changes in the retinal pigment of vertebrates, cephalopods and arthropods. It has also been shown that light influences the melanophores of the reptile skin much as it does the retinal pigment cells. “The effects of temperature upon the pigment migration of the melanophores, especially in Anolis and Phrynosoma, have been discussed recently by Parker (06). Only two reports of the effects of temperature on retinal pigment have appeared and these both refer to the frog. Kihne (79, p. 334) stated that in frogs which were subjected to low temperature in darkness, the retinal pigment extended farther toward the light between the cones than it did in those which had been subjected to high temperature in the dark. Herzog (’95) subsequently investigated this subject at greater length. He agreed with Kihne that below 18°C. any decrease of temperature causes a distal migration of pigment and any increase, a proximal one. Above 18° C. he believed the result was the reverse. It was suggested to me by Prof. G. H. Parker that the decapod crustaceans would be favorable objects for the study of the influ- ence of temperature upon pigment migration because of the marked photomechanical changes often found in their eyes. ‘The prawn, Palamonetes vulgaris Stimp., and the crayfish, Cambarus bartonu Gir., were chosen as being easily obtainable species whose photomechanical reactions were well known. The conditions found in Palazmonetes will be considered first. Tue JourNnat or ExpERIMENTAL ZOOLOGY, VOL. IV, NO. 4. 540 Edgar Davidson Congdon II PALAEMONETES VULGARIS STIMP. The eyes of the prawn consist of a stalk and terminal bulb. The former contains a series of four optic ganglia, which are enlargements of the optic nerve. ‘The bulb is made of numerous rod-like ommatidia, which extend somewhat radially from a base- ment membrane (Fig. 1, mb. ba.) near the end of the stalk to square facets (cta.) in the cuticula covering the bulb. An ommatidium may be roughly divided into a distal two-thirds, consisting chiefly of cone cells, and a proximal third containing the rhabdom. ‘The four cone cells (cl. con.) lie parallel to the axis of the ommatidium, and in their distal portions are closely associated to form the glassy cone. They taper proximally from the cone to the rhabdom whence they possibly extend as processes to the basement mem- brane. Between the cone and the facet are two small corneal hypodermal cells (cl. crn.) Six distal retinular cells full of black pigment together form a sheath around the cone cells (Fig. 3). They do not belong exclusively to one ommatidium, but each one serves as a partial sheath for three cones. ‘The lower third of the ommatidium contains seven proximal retinular cells (Fig. 1, cl. px.), lying close together and parallel to the ommatidial axis; they extend as long processes into the region proximal to the basement membrane. Distal to this membrane they unite to form the spindle shaped rhabdom (rhb.) ‘They contain black pig- ment. Associated with them are one or two whitish accessory pigment cells. Fibrillz from the distal optic ganglion extend up through the proximal retinular cells to end in the rhabdom. Distally, the central parts of the ommatidium are transparent and convey the light to the rhabdom, which is thus open to stimulation. The photomechanical changes of Palamonetes have been de- scribed by Parker (’97). In increasing light the distal retinular cells migrate as wholes in a proximal direction, thus restricting, as Exner (’g1) has pointed out, the amount of light that enters the deeper parts of the eye. ‘The pigment of the proximal retinular cells is at the same time carried distally along the sides of the thabdom, probably by protoplasmic streaming within these cells. This process also reduces the amount of light that can reach the Migration of the Retinal Pigment 541 thabdom. In decreasing light the distal pigment cells mvoe distally till they surround the cone and the proximal pigment is In consequence of the carried below the basement membrane. reget? one (PRAY wiht ee ‘OD Rr SEX SEK0 ig dens gece EX, Hooy Se be) Pa 009 So obs 30° light 10°. light 30%. dark.10%. dark. Fig. 1 Fig. 2 Fig. 4 Figs. 1 to 4 longitudinal sections of the ommatidia of Palemonetes, magnified 200 diameters, show- Fig. 3 ing the distribution of the retinal pigment under the following conditions: cl. con., cone cell; cl. ern., corneal hypodermal cell; c/. dst., distal Fig. 1, in light and at 10° C. retinular cell; c/. px., proximal retinular cell; c#., cone; mb. ba., basement membrane; cta., cuticula; rhb., rhabdom. i Fig. 2, in light and at 30° C. Fig. 3, in dark and at 10°C. Fig. 4, in dark and at 30°C. movement of the proximal pigment, the light-colored accessory pigment about the rhabdom is exposed and may serve, as Exner 542 Edgar Davidson Congdon observed, for areflecting apparatus. [hese pigment migrations plainly tend to protect the eye from over-stimulation by strong light and to increase its chances of stimulation in weak light. It is evident from the foregoing that light conditions must be taken into account in testing the effects of temperature. To do this, a series of experiments in the dark and another in the light were planned, each of which included three temperatures: 10°, 20° and 30°C. One extreme of temperature in each series would increase and the other decrease the effect of light, if indeed tem- perature is a factor in the migration of pigment. As the photo- mechanical changes in Palzmonetes are ordinarily completed in about two hours, I extended my experiments to only two and a half hours. The three experiments of each series were performed at the same time. Care was taken in the light series to have the three aquaria for the three experiments placed close together so as to receive equal illumination. Six to twelve individuals were taken for each experiment. All of the animals came from a com- mon supply and were similar in size and sex. One set of experi- ments was conducted twice, once with each sex. No difference in responsiveness was found between males and females. The animals were put into quart glass jars filled with water and these jars were placed each in a two-gallon cylindrical aquarium. The latter was filled with water at the desired temperature; thus the water in the inner jar containing the animals was brought in half an hour to the required temperature. Preliminary trials showed that 10° C. and 33° to 35° C. were not harmful to the prawns when thus gradually produced, though the higher tem- perature would cause at times the death of the animals if suddenly applied. In the dark series there was no easy means at hand for maintaining the desired temperature in the dark-proof box with- out admitting light. Consequently the box was kept closed dur- ing the whole of the experiment, the water thus being allowed to change gradually toward room temperature. ‘This resulted in a variation of about 3° C. during the experiment, an amount not sufficient to predjudice the result. At the end of treatment the animals were plunged into water at 80° C. for a fraction of a min- ute, until fixation was accomplished. “The eyes were then prepared Migration of the Retinal Pigment 543 for examination by being hardened, cut into sections and stained with borax carmine. Eyes of the different experiments showed a very perfect series of migration stages both for the proximal and the distal pigment. In the dark series, the proximal DieseD was always proximal to the basement membrane. At 10° C. (Fig. 3) it was close against the basement membrane; at 20° C. the solid of its distal mar- gin below this membrane was equal to about one-fifth the length of the rhabdom; at 30° C. (Fig. 4) the distance was two-fifths of the length of the rhabdom. In the light series at 10° C. (Fig. 1) there was a strong concentration of the proximal pigment into the ends of the cells just distal to the rhabdom. Only a little pigment could be seen proximal to this. At 20° C. more of the pigment was proximal in position and surrounded the rhabdom. At 30° C. (Fig. 2) the pigment was rather dense around the rhabdom yet not so abundant as at the distal ends of the cells. Both series show that with increasing temperature the proximal pigment moves proximally and with decreasing temperature it moves distally. The distal retinular cells can not be said to show as pronounced a response as the proximal ones did, yet the series was convincing. In all preparations from the light series the distal retinular cells were proximal to the cone; in the dark series they were at least partly surrounding it. At 20° C. in the light the distal pigment was in large part distributed evenly along the cone cells; a small part was collected at the top of the proximal retinular cells. Low temperature increased the effect of the light by massing all of the pigment just above the proximal retinular calle. The Wen tempera- ture produced an even distribution of the pigment fogs the cone cells with no proximal accumulation. In the dark condition the distal pigment cells completely covered the cone at high tempera- ture (Fig. 4). Low temperature (Fig. 3) resulted in a proximal migration equal to one-third the length of the cone. In general, increased temperature causes distal, and decreased temperature, proximal migration in distal retinular cells. It may be said that in both types of pigment cells in Palamonetes the effect of increased temperature is opposite to that of incre sed light. 544 Edgar Davidson Congdon Parker found a migration of the accessory pigment, but as it is not easy to determine this even in its relation to the light, I did not occupy myself with its relation to heat. III CAMBARUS BARTONII GIR. In the crayfish Cambarus bartonu the reaction of the retinal pigment to changes in light and in temperature were by no means so clear as in the prawn. Exner (’g1, pp. 108-109) long ago made a similar statement about Astacus, so far as light reactions were con- cerned. In experimenting on Cambarus the factor of light was elimin- ated by confining the temperature treatment to animals in the dark. The series of temperatures tested was 2°, 10°, 14°, 22°, 29°, 34°; 39° and 41°C. The high temperatures 39° and 41° C. resulted in the death of half the animals and produced variable conditions in the eyes of the survivors. ‘The experiments were discarded as probably dependent upon irregular moribund conditions. ‘There was no reason to think that 2° C. was harmful. As a precaution against shock the more extreme temperatures were only gradually applied. Animals that were subjected to 2° C. were kept at this temperature for twelve hours, so that, if the cell processes were somewhat retarded by the cold, sufficient time would be given for the reaction to become complete. ‘The animals were kept at the other temperatures for at least two and a half hours, as in the case of Palamonetes. The low temperatures were obtained by putting the animals in a shallow dish supplied by water cooled by ice. For the high temperatures, a tank was arranged in a light-proof box so that water of a desired temperature could be replenished without admitting light. After the experi- ments the eyes of the animals were fixed in water at 80°C. The tough cuticula was then removed and the retina sectioned and stained in borax carmine. | The distal retinular cells of Cambarus showed in the different experiments a considerable diversity of positions and could not be reduced to a simple series conformable to the differences of tem- perature. I am inclined to ascribe this condition to some fault in ila Migration of the Retinal Pigment 545 my methods. ‘The average positions of the proximal pigment of different animals formed a natural series, although the average i MY, Fig. 5 Fig. 6 Fig. 7 Figs. 5 to 7 longitudinal sections of retinular andsub-retinal portion of the eye of Cambarus, magnified 350 diameters, showing the distribution of the retinal pigment under the following conditions: Fig. 5, in dark and at 2°C. Fig. 6, in dark and at 22°C. Fig. 7, in dark and at 34° C. interval between the temperatures employed was only about 6° C. At 2° C. (Fig. 5) a considerable amount of the proximal pigment 546 Edgar Davidson Congdon was found clustered around the rhabdoms or between them. Some- times a little was scattered in the region between the rhabdoms and the basement membrane. Eyes of animals that had been kept at 34° C. (Fig. 7) frequently showed no pigment at all above the basement membrane. In some cases a small amount of pigment was scattered near the rhabdoms. Although there were frequent individual differences, the remaining experiments yielded a natural series of results between the two extremes mentioned. The proximal pigment of Cambarus moves therefore, like that in Palamonetes, distally with decreasing temperature and proximally with increasing temperature. Light and heat have opposite effects. As in Palzmonetes, the movement in response to temperature 1s much less than in response to light. IV. DISCUSSION Parker (’06, p. 410) 1n a recent paper summarized the effects of light on the melanophores of the reptile skin and on the proximal retinular pigment in the arthropods in the statement that all migration due to increased light is distal and so toward the source of light, and all migration due to decreased light is proximal. He also gives experimental evidence that increased temperature produces a proximal migration and decreased temperature a distal one in the melanophore pigment of reptiles. “The proximal pig- ment of the decapod crustaceans Palamonetes and Cambarus, as shown in this paper, falls under the same rule. Herzog’s obser- vations on the influence of temperature on the migration of the pigment in the frog’s retina agrees with this statement for tempera- tures below 18° C., but not for those above this point, where the reverse 1s said to be true. Aside from this observation of Herzog’s, which needs confirmation, all evidence points toward a general law for temperature the reverse of that for light; namely, increase of temperature causes proximal migration, decrease of temperature, distal migration. In most instances of the migration of retinal pigment, the proc- ess has an adaptive value in controlling the amount of light that reaches the receptive organs. Possible adaptations may also be Migration of the Retinal Pigment 547 easily pointed out in the migration of the melanophore pigment of reptiles. On the other hand, the migration of the retinal pigment in crustaceans as caused by change of temperature seems to me not to be adaptive, for it is always small in amount and it occurs at temperatures higher and lower than those commonly experienced by the animals. It seems reasonable that the migration is closely associated with the accelerating effect of heat on the chemical changes in the melanophore cell, and even if the migration were more marked, it is difficult to see what advantage it would give to its possessor. A probably related instance of lack of adaptation has been described by Hess (’05, p. 423) in the cephalopod eye, which often required one to two days in which to complete che pigment migration. ‘This length of time renders the migration at best a very imperfect means of adaptation for sur- roundings in which light and darkness follow each other at half-day intervals. V SUMMARY In both Palemonetes and Cambarus the proximal retinal pig- ment migrates distally when the temperature is lowered and proxi- mally when it is raised. In Palamonetes the distal pigment migrates proximally when the temperature is lowered, and distally when it is raised. In all cases increased temperatures cause a pigment movement the reverse 1n direction to that produced by increased light. The effect of temperature is much weaker than that of light. In the eyes of crustaceans retinal pigment migration due to temperature changes is probably not adaptive. BIBLIOGRAPHY EXNeER, S., ’91—Die Physiologie der facettirten Augen von Krebsen und Insecten. Deuticke, Leipzig und Wien, vii +206 pp., 7 Taf. Herzoa, H., ’05—Experimentelle Untersuchungen zur Physiologie der Bewegungs- vorgange in der Netzhaut. Arch f. Anat. u. Physiol., Physiol. Abt., Jahrg. 1905, Heft 5-6, pp. 413-464, Taf. 5. Hess, C., ’05—Beitrage zur Physiologie und Anatomie des Cephalopodenauges. Arch. f. ges. Physiol., Bd., 109, Heft 9-10, pp. 393-439, Taf. 5-8. 548 Edgar Davidson Congdon Kutune, W., ’79—Chemische Vorgange in der Netzhaut. M.L. Hermann, Hand- buch der Physiol., Bd. iii, Theil 7, pp. 235-342. Parker, G. H., ’97—Photomechanical Changes in the Retinal Pigment Cells of Palzmonetes, and their Relation to the Central Nervous System. Bull. Mus. Comp. Zool. Harvard Coll., vol. xxx, no. 6, pp. 273-300, I pl. *99—The Photomechanical Changes in the Retinal Pigment of Gam- marus. Bull. Mus. Comp. Zool. Harvard Coll., vol. xxxv, no. 6, pp. 141-148, I pl. °06—The Influence of Light and Heat on the Movement of the Melano- phore Pigment, especially in Lizards. Jour. Exper. Zodl., vol. ii, no. 3, pp. 401-414. From the Zoological Laboratory, Columbia University. OBSERVATIONS AND EXPERIMENTS ON REGENER- ATION IN LUMBRICULUS! S. MORGULIS MACHO AGE Oa ereyte ere res A ororee ete ee tes cto tS 5 Sys wlohe, Sia “ool a cg oncrahe Ves oveoEh eine ence Ne eeere heen lnsatoaatentie 549 ANNLOLOMYOLME PLOGUCHOMN aris lamas clasts escherehe ols ele eet ene er Nearer toe 557 GpeWereneration onrereneratedi pants, mraiseria-ioci/e fri talyo alse rl ictal te siete eters astral 566 1D). UXCRCOGH WE iedorano} (9) OS So gacseaqne occa pan OO Sao Ononm noo onaucd eS oonnanooaddds 570 Figs SOme COMMENLS OD ANLehOULereMeLa tlOlMeyar teeta ater yl eie ter atelier tet ienel eee erento 571 IRargeravereahentonnl ahatel Zea WOAIM ONG co poaw SMMOD AD oa CAME DOO DBOODAEnEDRAB EEG Locmsopoonobenaoeoe 572 SURI ANY topes ate, Sette 2 Ao) eve eTeTe IS eVSlors lel e vie) A Sis Soe sleyD apm als VEE 6 Fee +s hGO El aides BEE Siel nelelhoerertes 573 LAST OT MTGE, con sons os obec boo Pad oO ono apeo DIDO Dono OUHO DAD AnADOSODDsOpECUBuOOGAAGCeE 574 This paper is the outcome of some experiments carried out mainly during the months of last autumn. ‘The problem was sug- gested to me by Prof. T. H. Morgan, under whose direction this work has been done, and it gives me great pleasure to avail my- self of this opportunity to express my sincerest gratitude to-Prof. T. H. Morgan for having awakened in me an active interest in the subject, and also for his kindness in revising the manuscript. INTRODUCTION The worms, 4-5 cm. long, are made up of 100—-180 or even more segments, each bearing four pairs of sete. ‘The first seven to eight segments are readily distinguished from the rest of the body, and to them is applied the name “head.” A median dorsal blood vessel gives off in each segment a pair of lateral branches, and also a pair of blunt finger-like diverticula, called the “ blood glands,” which are absent in the first nine or ten segments. 1 This is the species originally described by Leidy as Lumbriculus limosus, and kindly identified for me by Dr. Percy Moore, as Thinodrilus limosus. Tue JourNAL or ExpERIMENTAL ZOOLOGY, VOL. IV, NO. 4. 550 S. Morgulis The worms were found along theedges of ponds between decaying leaves. In aquaria they can be seen on the bottom with their tails deeply embedded in the mud, and with the anterior parts of the body extending out like poles, always at an acute angle with the level of the bottom. When the water is disturbed, or when they are approached by a pipette, they instantaneously disappear under the mud. To mechanical stimuli, such as irritation by means of a sharply or bluntly pointed object, Lumbriculus responds more or less readily: the more anteriorly the stimulus is applied the more rapidly and vigorously the worm reacts. No reaction, or only a very slight one, is apparent when the tail is touched. If the ante- rior end is stimulated by a current of water or by repeated tickling with a needle, the worm remains quiet and motionless for a while. This quietness may be easily misinterpreted as a failure to per- ceive the stimulus, but such a conclusion would be erroneous. Prof. C. O. Whitman gives a very interesting account of a similar, but very much more striking kind of reaction in Clepsine.? He points out that if the surface of the water is touched with the point of a needle just above the animal’s back, it ceases to respire and becomes quiet, aptly called by Prof. C. O. Whitman—deceptive quiet. That the animal, although motionless, is in a state of active resistance, is shown by the rigidness of the musculature and some other symptoms. In the case of Lumbriculus I could not actually see the strained condition of the musculature, but the stretching out of the sete is probably an indication of such a con- dition. : Clepsine in the condition of “active resistance” adheres to the dish so firmly that it is very difficult to force its hold. ‘With one end detached, the other will often hold against a pull strong enough to snap the body in two.” I have observed similar cases in Lum- briculus. When stimulated, the worm sometimes becomes so firmly attached to the surface of the leaf, that it is impossible to draw it into the pipette, although a part of its body may be broken off by the forcible current of water in and out of the pipette. ‘This, 2Prof. C. O. Whitman: Animal Behavior. Biol. Lectures. Regeneration in Lumbriculus 551 however, happens very rarely. Lumbriculus also shows a thigmo- tactic reaction. In dishes of water the worms show a tendency to crowd together, rolling up in a ball, at times of considerable size, containing as many as twenty to thirty worms.’ Separate pieces taken from different levels also show the same reaction to mechanical stimuli as do the corresponding regions of the normal worm. Only the anterior pieces however come together to form balls. AUTOTOMY OR REPRODUCTION It is well know that of all the Oligochata Lumbriculus possesses the greatest capacity of breaking off pieces of their body when subjected to stimuli. ‘*Wie schon gesagt, sind die Thiere (Lum- briculus) ausserst empfindlich gegen aussere Reize und zerreissen sich haufig schon bei ganz leiser Berthrung an der getroffenen Stelle. . . . . Deshalb findet man so haufig im freien Wasser, noch mehr unter den in Glasgefassen geztichteten Uhieren, verstum- melte Individuen, da sie beim Anstossen an die Wande oder son- stige harte Gegenstande leicht zerbrechen.””* O. F. Miller’ also speaks of this behavior. If, he says, Lum- briculi be put into dishes—‘‘so wird man bald an ihnen den Schwanz vermissen; selbst in ihrem naturlichen Aufenthalt trifft man wenige unbeschadigt an; die meisten sind in Begriff, einen neuen Schwanz, andere einen neuen Kopf, noch andere beides zu ent- wickeln . | . . . Demnachscheint dieses Zertheilen ihnen naturlich zu sein, und vielleicht das Mittel der Erhaltung threr Art.”’ In Table I, I have summarized Bilow’s figures regarding the condition in which the worms are found in nature. ‘These data were obtained from a very large number of individuals. From this table we see at a glance how large 1s the proportion of individuals regenerating, especially those regenerating their tails, 3C. M. Child gives a similar description of the behavior of a fresh water nemertean Stichostemma “¢Several specimens in a jar of clear water will often aggregate in a single mass, crawling over and between each other and finally becoming nearly quiet.’ The Habits and Natural History of Sticho- stemma, Amer. Natur., xxxv, 1901. 40. Diffenbach: Anatomische und systematische Studien au Oligochaéte limicole. 24 Ber. Ober- hass. Ges., Giessen, 1886. 5O.F. Miiller: Von Wiirmern des siissen und salzigen Wassers, Kopenhagen, 1771. 552 S. Morgulis which justifies Bilow’s question—“‘ob denn allen Vhieren mit irgend welchen regenerirten Enden die einst verlorenen Stucke von Feinden abgerissen seien, um als deren Nahrung zu dienen, oder ob nicht etwa der Lumbriculus bei seiner eminent weitge- henden Regenerationsfahigkeit sich selbst verstitimmelte, d. h. in Sticke reisse oder zerfalle um aus diesen Stiicken ganze Thiere entstehen zu lassen und auf diese Weise durch einfache Querthei- lung, also ohne vorher angelegte Knospungszone, sein Geschlecht fortzupflauzen.” Bulow decides the question in favor of the latter possibility. Miller, as we have seen, is inclined to give a similar interpretation. On the other hand this general view is disputed by some observers. Thus Diffenbach accounts for the breaking up as the result of the attacks of enemies. In order to throw some light upon this ques- TABLE I | Individuals with | Individuals with | Individuals with | | Non-regenerat- Month | regenerating Repener zune | zepeuergouE | ing individuals heads tails heads and tails | ily PEGCED Umer terre 15.6 42.5 | 227.7 5.8 September, per cent...... | 22 | AST ay oe 25 tion, z.e., to test whether the breaking up is merely a case of “ reflex throwing off of parts of the body, or autotomy,’”® or whether it is to be looked upon as a normal process of reproduction, the following observations were made. I found that in the American form of Lumbriculus there is lack- ing the high degree of sensibility described by almost all students of the European form. Worms kept in any kind of a dish, in the presence or absence of mud, never broke up into pieces. Draw- ing them into a pipette, and then spurting them out in a strong current of water, repeated many times in succession, never caused them to break apart in any way; nor did they react when put out on a paper or in the palm of my hand. ‘To be sure, care must be taken not to injure the extremely delicate body-wall of the worms, 6 Prof. T. H. Morgan: Regeneration, p. 110, 1901. Regeneration in Lumbriculus 553 as this may result in a splitting off at the point of injury. But even in such a case the splitting is by no means a direct result to the irritation since the pinching in two is very slow, and requires sometimes hours for its completion. I tried to imitate the attack of an enemy by grasping either end of the worm, and letting the other end hang down. ‘The suspension of the worms 1n such a position for several minutes gave absolutely no results. Neither could I verify Miller’s statement that a breaking off of portions of the worm’s body 1s effected by the addition of alcohol to the water. According to my observations alcohol as well as some other substance effects a breaking through of the body-wall at various points, and a consequent loss of blood through the broken blood vessels, but these ruptures of the body are never deep enough at the start to bring about the separation of portions of the worm, and if the worms be brought back into pure water, do not proceed any further. As a matter of fact some worms may recu- perate perfectly from this treatment. [he same thing will happen if the worms be subjected to drying. ‘The body-wall will actually burst open in many places, but if put back into water in due time, the worms may remain intact and continue to live. These results, I think, show pretty conclusively that the break- ing off of parts of the worms is in some way or another connected with an injury to or distortion of the body-wall and is by no means the expression of a reflex action. This difference in behavior of the European and American Lumbriculus may appear to be due to a difference in species, were it not that v. Wagner, a careful student of Lumbriculus, throws some doubt upon the remarkable behavior ascribed to the Euro- pean species. “Ich habe schon 1893 mitgetheilt, dass es mir, trotzdem ich seit Jahren Lumbrikeln halte, nicht gelang, den spontanen Zerfall derselben beobachten zu konnen; ich médchte die Thiere noch so unsanft behandeln, niemals reagirten dieselben durch plotzliches Zerbrechen . . . . . Seither (’00) habe ich mich, wenn auch mit wiederholten Unterbrechungen, doch weitere Jahre mit der Haltung und Beobachtung von Lumbrikeln befasst, 554 S. Morgulis und ich kann wieder neuerlich sagen, das mir die Erscheinung der Selbstzerstiicklung niemals vorgekommen 1st.” To the same effect we read in Mrazek’s tecent paper — “Wenn es nach den Ausfiihrungen v. Wagner’s noch eines weitern Beweises bediirfte, dass die Lumbrikeln keineswegs allzu empfindlich sind gegen mechanische Reize, so wurden meine Erfahrungen einen solchen in hinreichendem Masse liefern.”’ If, then, the pinching apart is not a case of autotomy, let us con- sider whether the pinching off of pieces of the worms is a mode of reproduction. ‘The strongest evidence in favor of this view is Bulow’s record of the number of regenerating individuals found in | nature, referred to above, and his own experiments, which may be summed up thus: out of twenty-five worms, handled with utmost care, in course of five to six weeks, fifteen worms have divided into fifty-one pieces, which gives an average of 3.5 piece to each worm. Other observers, as v. Wagner, for instance, favor this view simply because at a certain period they found the number of worms in the aquaria increased, the worms being at the same time smaller. It is to be regretted that Builow does not state definitely how much each of the worms classified as “individuals regenerating their tails’ was actually regenerated. To fill up this gap I have recorded the amount of regenerated tissue in worms found in their natural environment. ‘The number of worms was unfortunately much smaller than that used by Bulow. The measurement given in Table II do not pretend to be exact, and, indeed, were made rather roughly, nevertheless they will serve sufficiently for my purpose to show the size relations between the old and new tissue. This table shows that most of the worms that might be referred to as “individuals with regenerating tails” regenerate in reality only the tips of the tails. “The ease with which the worms lose this portion of their body is due to the exceedingly delicate body-wall of this region. ‘“‘Der Korper ist ungemein murbe, so dass na- mentlich das Hinterende leicht abbricht, wird aber in kurzer Zeit regenerirt. 7 7 Fr. Vejdovsky. Regeneration in Lumbriculus Bey) If we should disregard this slight amount of regenerated tissue, and group all these individuals as “nonregenerating,” the whole aspect of conditions will be utterly changed. ‘There would be us = Oo) AR Res Ch) Oo + > + pW kN F Ww WW mWwW Ww A nnn nN Mn Part missing TABLE II Part regenerating General remarks | | —— Tip of tail. Very tip of the cail. Very tip of the tail. Very tip of the tail. Very tip of the tail. Very tip of the tail. Tip of the tail. Very tip of the tail. | Tip of tail, 7mm. | Tip of tail, 7mm. | Tip of tail, 6 mm. Tip of tail, 3 mm. | Tip of tail, 3 mm. Tip of tail, 4mm. Tip of tail, 5 mm. Tip of tail, 4 mm. Head Tail, 1.5 cm. Tip of tail, 6 mm. Tip of tail, 5 mm. Tip of tail, 5 mm. Tip of tail, 6 mm. Tail, 1.5 cm. Tip of tail, 8 mm. | Tip of tail, 5 mm. Tail, 1.5 cm. Head and Tail, 1.5 cm. | | Was injured, pinched off two hours later, Was injured, pinched off (noticed only next morning). Could not determine with surety that the tail was really regenerated. The tip of the tail consists of two to three zones of regenerated tissue. The regenerated tail had its tip regenerating anew. 3-7 per cent regenerating heads, 11.1 per cent regenerating tails (this number may be even larger); 3.7 per cent regenerating tails and heads, and 81.4 per cent of “nonregenerating.” 550 Si: Morgulis In the two cases, where the heads had regenerated, the size of the worms did not differ sufficiently from other intact worms, and there is no reason to suppose that these worms have necessarily resulted from a spontaneous division somewhere about the middle of the body. More likely the foremost few segments were lost by a mere accident. Could any of the small pieces give rise to new worms and in this way propagate the species ? | will show later that such pieces, unless large, will soon die. Moreover if the worms always divided in the middle of their body, we should expect to have as many worms with regenerating heads as those with regenerating tails, but this does not occur. A regenerated head has certain peculiar- ities that distinguish it from a normal head, that can be detected long after its completion. If it be true that the posterior pieces generally perish what good would such a process of division be for the continuation of the race. As to the evidence brought forward by v. Wagner, I can only say that his statement 1s too uncertain, for, the presence of a large number of smaller worms might be as much an indication of larval development as of an asexual multiplication. EXPERIMENTS A Smallest Part Capable of Regeneration Bonnét found that when he had divided a worm into sixteen pieces, each piece formed a perfect worm again. If we assume 120 segments as the average number for Lumbriculus, these pieces would have contained seven to eight segments each. In my own case a number of very small pieces one to five segments each were first tried. In order. to keep the smallest pieces alive, I first cut the worms, into large pieces and allowed the cut surfaces to heal over for eight- een to twenty-four hours, and then cut off smaller parts near the closed ends. By using this method I was enabled to keep alive about one-third of the entire number of pieces for a time long enough to give results. f Regeneration in Lumbriculus aS Experiment I (October 14). A number of pieces were cut off and allowed to heal for about eighteen hours. ‘Then still smaller pieces were cut off near the closed ends. a From the anterior half of the worms | obtained: two pieces of five segments each; five pieces of four; eight pieces of three; S1X pieces of two; and two pieces of one segment each. Most of these pieces were found dead about the second day after the operation. On October 19, the survived pieces, containing 53 45 45 45 35 2, 2 old segments, had all regenerated a head and atail. b From the posterior halves of the same worms, I obtained: five pieces of five segments each; eight pieces of four; nine pieces of three; two pieces of two; and two pieces of one segment each. October 19, the surviving pieces of 5, 5, 5) 45 45 45 45 3» 35 39 3> 2, 2 had all regenerated a head and a tail. This experiment shows that pieces containing only two segments are capable of regenerating. Experiment II. a_ By following the same method, pieces from the anterior half of the worms, consisting of 3, 3, 3, 3, 2, 2, 25 I, I, 1 old segments also produced a new head and tail. This experiment showed very distinctly that a single segment is capable of maintaining its existence, and of regenerating a perfect head, consisting, as is usually the case, of six segments, and also a tail. b From the posterior halves of the same worms, I got thirteen pieces of three segments each; sixteen of two segments and twelve of one segment. Of these there regenerated eight pieces of three segments, seven pieces of two segments each, and none consisting of one segment only. B Rate of Posterior Regeneration The object of the following experiments was to determine whether the length of the piece or its relative position in the worm’s body is directly. responsible for the rate of its posterior regenera- tion. Experiment III (October 25). About thirty-five worms were divided into seven parts each. The seventh piece (the tip of the 8 Archiv. f. Entwickelungsm., xiv, p. 586, 1902. 558 S. Morgulis tail) was not utilized in these experiments. Although the average length of these terminal pieces was somewhat over forty to forty- five segments they all died. Pieces of corresponding levels were kept in the same dish, so that they were all practically under like conditions. Accord- ing to the level to which the pieces belong, they will be named A, A» As Ay A, A, At the end of two weeks all the tails that had been regenerated by these pieces, were cut off; the number of their segments as well as the number of segments in pieces, by which they were produced, was recorded, and the results are given in lable III. Inthistableis also given the average num- ber of regenerated segments per one old segment (in the last line). By this method of calculation all the individual variations were obliterated, and at the same time this, so to speak, “ideal old seg- ment” with the corresponding “ideal’’ number of new segments served to indicate the regenerative power at a given level, and served also as a basis for the comparison of the rates of regenera- tion. In accordance with our method of calculation, we find that to each of the original segments at the first level, there were on the average 3.2 new segments; at the second level 3.3, at the third level 3.1, at the fourth level 2.6, at the fifth level 1.8, and at the sixth level only 0.9 new segments. There is no evidence of a correspondence between the number of old segments and the number of new, regenerated segments, provided the old parts are not very different in length. Glancing over the Table III we can see at once that in pieces at the first level, those having twelve segments produced thirty-two or fifty new ones; those having thirteen old segments thirteen or forty-two. In pieces at the second level also, those of eleven old segments regenerated twenty-nine or fifty-six, and those of sixteen may regenerate eighteen or fifty-four new segments, etc. ‘The same lack of correspondence will be found in all parts of this table as well as in subsequent ones. On the other hand if we compare the smallest pieces of a cer- tain level with the largest pieces of the same level we may some- times find very considerable differences in the amount of regener- Regeneration in Lumbriculus 559 ation. ‘Thus in pieces of the third level those of seven or nine segments regenerate thirty-five or thirty-eight segments respect- ively, whereas pieces of twenty-eight, twenty-nine and thirty-one TABLE III October 25, 1906/ November 8, 1906 At Ao As Ag As As = fe Say | oe 2 a On Ra ee 2 ie wo | 2 Po| ee | fo | fe | fe | fe | fe | SE | fo | fe | Pe | Pe Pease | ame | a2 sare! goer ell cane Neal hee rel ce leas ERG morsel a geen i a ceill Saseo hy eee) Sees eee e eee lees SE 42 |4 Zz Z Z a a | a Z Z Z Z Pies | , | 10 22 8 25 7 af tt 42 S540 II ie) II 22 LO | 45 Qj) ahs 12 31 II 28° 16 17 12 38 10 46 10 41 |} 12 49 14 Pe) |) | aes iw) 7H Il 29 10 AN 1 62 16 38 | 18 | 24 12 44 ut 35 LOR Poesy lS, 45 17 42 | 20 | 8 12 50 II 36 iti) 8 15 46 17 5G ||| 20 16 13 13 Il 47 DD pele sz, 15 58 19 fo) 22 22 13 32 Phy 48 13 ° 18 34 19 44 | 21 15 13 36 Il 56 13 37 | 18 | 42 20 34 21 23 19) ls oe 12 50 13 44 LOE ASO 20 42 my | 1 14 44 12 65 14 50 19 50 21 38 Zech tao 14 53 13 ° 14 56 19 80 22 21 | 25 48 14 55 13 42 14 66 20 45 22 35 28) \\)) 08 wh i Ge 14 42 15 50 20 49 22 40 30 9 15 48 15 66 16 58 20 55 22 48 30 | 40 Teel eS 15 70 16 60 20 58 23 43 32 30 16 40 16 18 17 fig |) 2X9) 60 25 21 42 32 20 58 16 49 17 48 21 fo) 25 42 AS Vie 27, 21 68 16 Es its) 55 22 280 a Zo 42 23 64 17 61 | 19 32 22 52 | 28 28 25 go 18 52 Apt Si Is) 22 53 28 42 19 66 | 28 68 22 54 28 58 24 65 29 60 23 | «49 30 45 26 65 Say th 70) yy 30 50 | 26 CONn en sai Ape || 30 55 ase 76 | Ba2. | 21088) | yal 2.6 1.8 0.9 segments regenerate as many as sixty-eight, sixty or ninety new segments. It is not in the least surprising, I think, that pieces so consider- 560 | S. Morgulis ably different in size should have different regenerative capacities; still these differences are of little account when we take into con- sideration the regenerative capacity of pieces intermediate in size. Thus a piece of only fourteen segments belonging to the same level of the worms regenerated sixty-six segments, or six segments more than the piece of twenty-nine segments. In view of this evidence we may safely assume that the length of a piece has no direct relation to the rate of its regeneration, and that the rate is dependent upon the position of a piece in relation to the worm’s head. ‘The pieces nearer to the head have the highest regenerative power, which gradually decreases as we pass from the front backward. I wish also to point out in this connection the great range of variability in the regenerative power of various worms, that can be easily seen on looking over every column of the table. nk Sera a ie Az cared aa As | Aw. | Diagram showing the rate of regeneration at different levels. In the accompanying diagram an attempt is made to demon- strate the rate of regeneration at different levels by means of rec- tangles. ‘The horizontal lines represent the “ideal” old segments of different levels, whereas the vertical ones represent the “ideal’’ number of new segments, regenerated at these levels. In either case a segment is expressed by a line 1 cm. long. The worms were kept in clear, filtered water, so that they did not have any food for two weeks, but since the pieces had no heads they would not have been able in any case to feed for a greater part of the time. After fourteen days (November 21) the pieces A, to A, were examined again. It will be remembered that on November 8, Regeneration in Lumbriculus 561 the regenerated tails of these pieces had been cut off, so that since then the worms had been regenerating anew. Only pieces of three levels (A,, A,, A;) were alive. The data are given in Table RV: TABLE IV November 8/ November 21 As eerie | As No. of segments No. of segments No. of segments|No. of segments No. of segments, No. of segments in old part | in new tail in old part in new tail | in old part | in new tail 9 | 16 9 16 6 12 9 | 33 1 | 27 i 14 10 15 | 12 6 12 4 Il | 24 | 13 34 15 22 12 10 14 24 15 | 23 14 | 22 15 46 15 | 25 14 | 27 16 22 18 | ° 14 | 32 17 17 | 18 20 14 | 38 17 22 | 18 21 15 25 17 | 28 | 18 22 15 32 l 18 | 20 19 17 16 26 18 | 22 | 20 6 16 34 19 a7 20 9 17 28 20 20 20 12 19 26 20 26 21 26 26 41 20 27 23 28 37) 43 20 29 24 13 29 45 20 33 24 22 21 24 25 40 24 44 26 26 30 36 | 27 27 | 28 17 | 36 49 1.8 | in | 1.0 This table shows that in course of the second period of two weeks, the pieces regenerated about one-half as many segments as were regenerated for the first period of two weeks. ‘The regener- ated tails were cut off once more, November 21. (It should be remarked in this connection that not only the regenerated tails, but also the regenerated heads as well were cut off every time.) 562 S. Morgulis After a third period of two weeks (December 5) the pieces A;, A, and A, were examined. ‘The result is recorded in the Table V. Again we see that the pieces have regenerated only about one- half as many segments regenerated for the second period of two weeks, and about one-fourth as many for the first two weeks. TABLE V November 21/ December 5 As Ag As No. of segments \No. of segments) No. of segments No. of segments No. of segments No. of segments in old part | in new tail inold part | innewtail | inold part in new tail - | 9 16 10 2 12 2 10 | 6 10 9 13 12 II 7 14 3 15 10 12 ° 14 10 17 Fi 13 ° 16 iy | 9 13 9 16 9 18 4 14 12 18 22 18 5 15 10 19 13 19 6 15 18 21 3 19 13 16 II 20 17 16 15 22 10 17 15 23 14 24 21 24 13 26 31 27 5 29 27 o.8 0.6 0.5 It is worth mentioning in this connection, that the ratio between the amounts of regeneration at various levels remains almost con- stant during the three successive periods of two weeks each. ‘Thus the ratio between the amounts of regeneration of A, and A,, A, and A,, A, and A,, for the first period will be: eae ete PASO | DW Aietaek hee. home ike. The same for the second period will be: ae Ia hs RG Dero ae ees 1 o5 rs ee) i340 Regeneration in Lumbriculus 563 And for the third period: 0.8 0.8 On = — | SS SS R — PO 0.6 cay ow Ors TABLE VI November 8 | November 21 Cees oe = | A atten ee ee Pa WANG rec etege ea ne arene eis ey?) te) i ol UNV UALS re of rel eh akc fesseatons 7 1.8 1.6 J IGT Mea eine oerarat tea! | Tes 1.2 Experiment IV. Another experiment was started on November 29. A number of worms were cut into seven pieces each, as in the previous experiment. Pieces of corresponding levels were kept in the same dish. ‘The dishes contained nothing but filtered water. No food was present. ‘he pieces will be named accord- ing to the level of the body B,, B,, B,, B,, B,, Bg. The worms used in this experiment had been in the laboratory for some weeks, so that their regenerative capacity was consider- ably lessened. In general worms just brought in from the pond are the best as far as their regenerative power is concerned, and lose it when kept in the laboratory. I regret that I did not take the record of the amount of regeneration at the end of the first two weeks, as it would be very interesting in connectign with the results obtained from this experi- ment. | At the end of four weeks (December 27) the pieces were exam- ined, the segments counted, and their numbers are ‘given in Table Vile This table shows practically the same result in regard to the relation of the regenerated to the old tissue, and that of the former to the level of the worm’s body, as the Tables II, IV and V. In this case, however, the regenerated tissue was not removed, and the pieces were left regenerating tll January 10, 1907, or fora total of six weeks. Table VIII gives the result of counting the segments at the end of six weeks. The number of new segments is the same as at the end of four 564 S. Morgulis weeks, 7.c., no new segments had regenerated posteriorly. The regenerated segments, however, grew larger, and all the microscop- ical ones at the tip of the tail became conspicuous, and their sete of considerable size. Slight differences in both these tables (VII and VIII) are probably due to some miscounts, which are almost inevitable. TABLE VII November 29, 1906/ December 29, 1906 Bi Be Bz Ba B; Be = ea ae E i : =a Po| fe | fo | fe | Bo | P| Pe | fe | Pe | Pe | fe | Pe SSS eS Pe Be eae ee | eel Re rong SRal Gee n= BoE Gea See escd e& eee A esshel eoris| ee Z Zz Z 4 Z a a va \@ |G Z Z | | II 25 10 21 10 7] 13 13 Sih |e) 16 7 12 20 II 2D, II 19 13 20 16 ° 16 15 12 40 12 24 12 30 13 Z00 1 16 21 17 5 13 35 13 26 12 30 13 g28 16 me || zo 12 14 35 13 32 14 12 16 18 yp || aa 23 28 15 26 13 36 14 26 16 34 1725) e2 gLGia! ea! i 15 40 14 Ay) ois 39 16 39 17 15 24 13 16 31 14 40 16 22 17 il We tty] 18 29 18 16 40 15 26 16 43 17 28 | 18 17 17 30 ig ele LG RL ats) PAS |j oe its 21 17 32 15 44 17 vie |) its 32 LW | 17 45 18 2Oin ie eee, 39 20 30 | 19 | 42 18 30 18 43. | 24 2A ZO) 46° | 260°") “05 18 38 18 45. | 24 40 22) 25 2D |), Eee 18 Ab i) oe 32 | 22 42 21 | 21 18 45 21 5°o (| | 23 35 ay |) eye) | 8 15 | fe 29 44 21 35. | ie 22 2 Gia | | 23 26 | 2a 36%) | a Np XS 2.26 2.26 | 1.80 | esi7fi6 1.09 0.62 Let us now consider this last experiment in connection with the previous one, and see what we can infer from them. In either case the pieces of the worms were regenerating without food for a period of six weeks. In the latter case, where the process was Regeneration in Lumbriculus 565 going on undisturbed, the parts formed a certain number of seg- ments, then the formation of new segments stopped, and only growth, or increase in size took place. It seems as if these pieces of worm, brought out of equilibrium by the operation, completed themselves, attained again a state of equilibrium, and formed TABLE VIII November 29, 1906/‘Fanuar y 10, 1907 B, By | B, | B, B; Bg See eo ee ee tee (fe) se oe eel eevee Pee salpaceen orn, | |ncee Nimes lero es = tn Ce) elt raven ome ack a4 ao Adsl ol) Sei S& 38 iS a ee ae B42) le Se Ze ie Z Pan. a Zz Z Ae a a Ane | | | II Dis | | kite) 21 | 10 ee 13 13 15 13 16 7 12 2001 | \ at Z2i|\| Enea acti sile wag 18 ie. i" 16 16 15 I2 | 40 12 22 12 30 13 30 16 | ° 20 12 mo a5 \- 13 26 14 12 13 32 16 21 23 28 14 35 13 ay ee. Ag Wy Pai | 17 II 24 | 7 15 26 | 13 36%). 15 37 16 34 tll) ag 24 14 15 40 14 ag | ae 22 16 40 17 15 29 18 16 40 14 36 | 16 43 17 27 17 LO 17 31 14 AO | 197 21 17 28 18 | 17 17 22. 15 2Heites Li 39 7 30 8 9\\ Ger 17 47 16 2300) 24 27 18 28 19 12 18 30 18 36 24 40 18 32 19 45 18 34 18 47 | 20 46 20 15 18 38 18 49° | lt ez 27 21 17 18 44 20 32 | 22 Az, 23 22 19 42 21 52 | 23 36 | ar 30 | 27 44 | 21 35 | 22 | 25 | 2300 areal | 24 pena 27 26 2.24 2.24 i715 1.74 1.09 | 0.66 dwarf worms. These were formed by pieces one-seventh the original length of the worms. If in accordance with Bulow’s data each worm divides spontaneously on an average into four pieces (3-5 pieces being the actual average), worms only twice as big as 566 S. Morgulis those obtained from my experiment, would result, and would also very likely remain dwarf worms. If my view in regard to the formation of dwarf worms from pieces of Lumbriculus is true, the worms if divided would become continually smaller and smaller, till they would be reduced to single segments. As a result of the continual cutting off of the regenerated tissue more new material was produced than when only one cut was made through the same level. Thus in the first experiment for the third period of two weeks there were to each old segment 0.5— o.8 of a new segment, while the pieces cut only once did not produce even a single segment for the same length of time. C Regeneration of Regenerated Parts ‘Tails that had been regenerating from parts of Lumbriculi for some time, were cut off so that none of the old tissue remained. Some of the tails had been regenerating for two weeks. After removal from the old part a new tail was allowed to regenerate in course of the following two weeks. ‘These were also cut off. a October 31. ‘Thirty such regenerated tails were obtained from pieces belonging to the third level, and thirty regenerated tails from pieces of the fifth level, of worms that were divided into eight parts each. ‘These tails had been regenerating since October 17. The tails regenerated by pieces of the third level we will call for convenience A, and those regenerated by pieces of the fifth level B. ‘These sixty tails were kept for two weeks, and when examined on November 14, eighteen A tails and thirteen B tails, or half the original number were alive. Those that survived regenerated new heads of five to seven segments, with one exception, of which I shall speak presently. The little worms contained on an average, probably, about forty-five to fifty-five small segments. In the course of two following weeks they did not produce any new posterior segments. That this is really so can be ascertained with certainty because in the regenerating tails the terminal five to six segments, which are the youngest, are microscopical in size. Dur- Regeneration in Lumbriculus 507 ing the time intervening between October 31 and November 14, all these microscopical segments increase in size very much, their setae became large, and the blood vascular system quite conspicu- ous, as in older segments; but no more “ anlage’ of segments have made their appearance. ‘The last large segment was contiguous to the anal segment. This result shows that when a new regenerated tail regenerates a head, and thus forms a little worm, although the segments grow larger, no new segments are laid down at ie posterior end. ahedicr the same fields true also for non-regenerated posterior ends of Lumbriculus, I do not know, since in all cases when pos- terior ends were removed they died. b November 14. Again twenty-eight regenerated A, tails, and thirty B, tails were obtained from the same pieces as the fore- going at the third and fifth levels, which had been regenerating from October 31 till November 14, or for two weeks. November 28, I found alive out of fifty-eight tails (A,, B,) eleven A, and seven B,. ‘They all had regenerated new heads, although the number of segments in these heads was in some cases only four. Thus about one-third of all the regenerated tails produced new heads, and eighteen very small worms of about twenty-five to thirty segments each were obtained. In this case also no new segments were laid down at the posterior end, but the segments became larger. c November 14, I cut in two all the thirty little worms formed by the regeneration of A and B tails, in order to determine the rate of their posterior regeneration. ‘The cut was made not quite in the middle, so that the posterior parts were somewhat longer than the anterior. After a lapse of two weeks, November 28, twelve anterior pieces which came from A, and eight anterior pieces from B, were found alive and regenerating new tails. None of the posterior pieces, although bigger than the anterior ones, had survived. (In other experiments a small percentage of the posterior pieces regenerated heads.) On the whole a very fair percentage (about 65 per cent) of the anterior parts survived the operation and continued to regenerate the missing tails. 568 S. Morgulis In the tables given below are recorded the number of segments present in the anterior pieces left after the operation of November 14 (c) and the number of new segments they had produced at the end of two weeks, November 28. In accordance with our method of calculating the average num- ber of new segments per one old segment, to define the regenerative dower, we will have 0.11 segments standing for this power in pieces from A and 0.06 segments cs that in pieces from B. TABLE 1X November 14/ November 28 Anterior parts from worms formed Anterior parts from worms formed by regeneration of A by regeneration of B No. of old seg- No. of new seg- No. of old seg- No. of new seg- ments ments ments ments 16 bud* 16 bud 16 bud 17 bud 19 3 19 bud 19 5 15 6 20 bud 20 bud 20 bud 23 4 20 4 26 bud 21 3 26 bud ae 5 20.5 1.3 23 4 Average No. of new segments oy bud per one old segment, 0.06 25 3 20.4 2.3 Average No. of new segments per one old segment, 0.11 * Whenever the term ‘‘bud’’ is used, it indicates that an unsegmented portion in which the anus lies is produced only, but no segments are deposited between this regenerated organ and the old tissue. Of course these data are so meager that it would hardly justify much speculation, but it seems to me nevertheless very suggestive. If we compare the rate of posterior regeneration in the little worms formed from the anterior parts of the regenerated tail pieces (A and B), we find that they stand to each other in a ratio very much like that of the rate of posterior regeneration in the old parts, from which the A and B tails originate. Regeneration in Lumbriculus 569 The parts of the third level regenerated from October 17 to October 31 the tails A at an average of 4.4 new segments for each old segment; and those of the fifth level regenerated in the same time the B tails at an average of 2.6 new segments. During the next two weeks, from October 31 until November 14 they had regenerated at an average 2.4 segments and 1.7 segments, respect- ively. ‘The ratio between these rates of posterior regeneration is: 7 a ae ea 4.4 2) A 26 ley) The ratio between the rates of posterior regeneration of the little worms formed from A and Bis: OnE 1.8 @106° The eighteen A, and B, tails, which survived and had regen- erated heads were also cut intwo. ‘This operation was performed on November 28, or two weeks after they had been separated from the old tissue. At the end of two weeks again (December 12), only seven ante- rior pieces from A, were alive. ‘The fate of these seven pieces is shown 1n the following ‘Table. TABLE X November 28 December 12 No. of old segments. No. of new segments. 14 bud 14 bud 15 bud 16 bud 17 bud 18 bud 27 bud This table shows that only the anus was formed, but no new segments were produced in these seven worms. Other experiments gave similar results and need not be recorded here. It is evident that a regenerated tail is not only capable of regenerating a head, from its anterior cut surface, but also of 570 S. Morgults replacing its posterior part when cut in two, and that the property of regeneration passes over to the new tissue together with the protoplasmic material it 1s built of. D A Case of Heteromor phosts One of the thirty-one A tails of the previous experiment regener- ated a tailin place of ahead. ‘This is the only indubitable case of heteromorphosis in Lumbriculus of which we have record. ‘That this was a genuine tail and not merely a misformed head can be easily proved (1 (1) by the number of regenerated segments, (2) by the position of the anal aperture; (3) by its “functional activities.” When an abnormal head develops it does not contain more than six to seven new segments. Here eighteen segments were regen- erated. ‘The segments gradually decreased in size and were micro- scopical near the distalend. ‘This sequence of segments 1s charac- teristic of the tail. In the regenerating head on the contrary all the material is laid down first, and then its segmentation appears. In this case the terminal aperture is not a mouth (which may some- times assume such an abnormal position) because it is round and not triangular in form, and lies in a knob of indifferentiated mate- rial, as in the case of the anus. ‘The best proof of its being a tail is the direction of the contractions of the blood vessels, dis- sepiments and entire musculature. Whereas in a head contrac- tion takes place from before backward, here it was inthe reverse direction, viz: in that characteristic for a tail. In the old tail the contractions were running in a direction opposite to that of the heteromorphic tail. ‘The waves of contractions in both tails started at their distal end, ran toward each other gradually slowing down and vanishing altogether in the vicinity of the point of their union. ‘lhe contractions in the old tail were more vigorous, but they never passed beyond the old part. This heteromorphic tail developed from one of a number of pieces that had been kept in the same dish. It could not therefore be due to an external influence. In order to see whether the old tail would again produce a heteromorphic tail or a head, and also to see what the heteromorphic tail would regenerate, I severed Regeneration in Lumbriculus 571 the heteromorphic tail from the old one, exactly at the line of their union. Unfortunately, both pieces soon died after the operation. E Some Comments on Anterior Regeneration Although this subject was pretty thoroughly studied by Bilow and v. Wagner, there are some points that have not been considered at all In the formation of the new head, abnormalities are not infre- quent. Double heads, arising immediately from the cut surface; or from a common stalk a little distance from the cut surface occur in about 5 to 10 per cent of the pieces. At the posterior cut sur- face, on the contrary, double malformations are of very great rareness. ‘The only instance I find recorded in the literature is that spoken of by Bulow, of a worm developing a double tail, and another similar instance which I found last summer. From the posterior cut surface there grew out two tails of somewhat differ- ent lengths, and the whole worm had a Y-shaped form. By observing the process of regeneration of an abnormal head one will be impressed by the constant movements that are going on inside the body-walls in a forward direction. ‘These exert a great pressure upon the delicate regenerated epidermis, causing it to protrude in many points. ‘This action may be largely responsible for these malformations, for if, on the other hand, the malforma- tions are supposed to be due to the operation, why should we not find abnormalities in the regenerating tail also? I have never observed any malformation of the head of Lumbriculi freshly caught in ponds. Another point that I wish to call attention to is the dissimilarity between a regenerated and a normal head. “Das Vorderende ist immer etwas grinlich oder griinschwarz, was von dem Pigmente herrthrt, welches namentlich die den Darm- kanal bedeckenden Driisen erfiillt.”» This green pigment is arranged segmentally in seven to eight very deeply colored bands, which give to the head a striped appearance. If from one to seven of the anterior segments are removed, the number removed will be restored, but the substituted segments lack the pigment, and are 9 Fr. Vejdovsky. 5/2 Se Morgulis perfectly transparent. If six to seven segments are cut off, so that one or two more stripes of the greenish pigment are left in the old tissue, the head will be perfectly repaired but not a single gran- ule of the old green pigment will be seen in the new segments, though I watched it for four weeks. Thus the two adjacent head segments, one from the old worm, filled with pigment, the other produced by the worm anew, pale and pigmentless, lie side by side. Heads regenerating from any other level of the worm’s body, far away from the pigmented region, are also without pigment. If worms under natural conditions do reproduce themselves by dividing into several pieces (four being the approximate average, calculated from Bulow’s experiments), should we not expect from this, that the majority of the worms in nature would have heads without pigment? In fact, such worms with pigmentless heads are not very frequent. It is true that after a lapse of a considerable time the pigment is formed anew in the regenerated head, and begins to develop from the distal end of the regenerated head. ‘This delayed development of the pigment in the regenerated head needs however a more - complete study. REGENERATION AND ADAPTATION The tip of the tails in Lumbriculi is almost always missing, or regenerating, which fact indicates that this portion of the worm’s body is easily injured and that these injuries are of a very frequent occurrence. Onthe other hand the power of regeneration is very feeble in this particular region as compared with that in the more anterior and less frequently injured regions. If we attempt to find a connection between these facts, we shall, contrary to Weismann’s claims, reach the conclusion, already expressed in 1898 by Prof. ‘T. H. Morgan in regard to the frequency of accidental injuries and the power of regeneration in the hermit crab, that “no such relation is found to exist.” ‘This low capacity of regeneration in a region where regeneration is always going on, seems to contradict the view that the capacity Regeneration in” Lumbri culus Sis to regenerate is an “adaptation of the organism to definite demands made upon it by conditions of life,’ and that it is “not the out- come of primary qualities of the living substance,” ‘‘not an inherent quality of the organism,” as it also contradicts the view that it is due toan “adaptation produced by natural selection.’’!” SUMMARY 1 Pieces of Lumbriculus containing only a single segment are capable of regenerating both a new head and tail. 2 Regeneration from a posterior end takes place more rapidly in pieces from the anterior region of the body, and gradually decreases as the pieces are taken from the more posterior region of the worm. 3. A piece ofa worm, when subjected tothe operation of cutting a few times will produce more new tissue for the same length of time than when subjected to cutting only once. 4. Norelation whatsoever between the number of old segments in a piece and its rate of regeneration can be found. 5 There is no relation between the available food and the rate of posterior regeneration at different levels of the worm. 6 In regard to its regenerative capacity each worm shows variations of its own. 7 Regenerated tails, when detached from the old part, are capable of regenerating new heads, but do not produce any new posterior segments. 8 Pieces of such regenerated tails are also capable of posterior as well as anterior regeneration, from the posterior and anterior cut surfaces. g The pigment of the regenerated head probably does not arise in connection with the old pigment, but develops anew. 10 In the case of the anterior regeneration, where only six to seven (eight) segments come back, the eighth (or ninth) to the tenth (or eleventh) segments of the old worm are dropped out. 11 ‘The experimental evidence, likewise that from observations, 10 A, Weismann: The Germ-Plasm, 1893. 574 S. Morgulis is opposed to the view that the breaking off of pieces of the worms with their subsequent regeneration, is a regular mode of reproduc- tion in Lumbriculus. LIST OF REFERENCES Bonnet, C., 1745—Traité d’insectologie. Seconde partie. Observations sur quelqueses pices de vers d’eau douce, qui coupés par mercoeux, devien- nent autant d’animaux complets. Paris, Butow, C., ’°83—Die Keimschichten des wachsenden Schwanzendes von Lumbri- culus variegatus, etc. Zeit. Wiss. Zool., xxxix. °83—Ueber Theilungs- und Regenerations-vorgange bet Wurmern. Arch. Naturg., xlix. CuiLp, C., ’06—The Relation Between Functional Regulation and Form Regula- tion. Jour. Exp. Zool., 11. DriescH, Hans, ’o6—Regenerierende Regenerate. Arch. Entwicklungsm., xxi. GruBe, E., ’44—Ueber Lumbricus variegatus Muller’s und him verwandte Anne- liden. Arch. Naturg. Hesse, R., ’94—Die Geschlechtsorgane von Lumbriculus variegatus Grube. Zeit. Wiss. Zool., v, 58. Iwanow, P., 03. Die Regeneration von Rumof und Kopfsegmenten bei Lum- bricus variegatus Gr. Zeit. Wiss. Zool., 75. Leipy, J., 50—Descriptions of some American Annelida abranchia. Jour. Acad. Nat. Sc. (Philadelphia), 2 series, 11. Morean, T. H., ’01—Regeneration, New York. °o2—Experimental Studies of the Internal Factors of Regeneration Arch. f. Entwicklungsm. xiv. ’°06—The Physiology of Regeneration. Jour. Exp. Zool., 11. Mrazexk, AL., ’06—Die Geschlechtsverhaltnisse und die Geschlechtsorgane von Lumbriculus variegatus Gr. Zool. Jahrb., xxi. Ranpo.pH, Harriet, '02—The Regeneration of the Tail in Lumbriculus. Jour. Morph., vu. Wacner, F. von., ’0o—Beitrage zur Kentniss der Reparationsprozesse bei Lum- briculus variegatus I. Zool. Jahrb., xii. °o5—Beitrage zur Kentniss der Reparationsprozesse be: Lumbriculus variegatus II. Zool. Jahrb., xxu. WeIsMann, A., ’93—The Germ-Plasm, New York. Vrypovsky, FR., ’84—System und Morphologie der Oligochzten. ZELENY, Cu., ’03—A Study of the Rate of Regeneration of the Arms in the Brittle- star Ophioglypha lacertosa. Biol. Bul., vi. CORRELATION AND VARIATION IN INTERNAL AND EXTERNAL CHARACTERS IN THE COMMON TOAD (BUFO LENTIGINOSUS AMERICANUS, Le C.) BY WM. Ee KELEVCOd i, Px. D: W.tru Six Figures AND TWENTY-TWO TABLES 1S ThitirorhtVea bodes s Abo oct cs CU cMmOte acid cole SORE Cm D RMAC AR aon O61nO,60-o0 oon cle anc 575 IML GQUimMEIATS o ood. cans oobsbsdoboesedsoonharopecnaponiauk bo to DulodgoeMogo soo IO dod0oRE 576 Nia RL eh cate ein ies iy eae Mette hag aust circa nen enstaP ae, aut», e-apallena eure (ele ne Mine: eee tone bates ener oe 577 it WET aor aon Reanim a erica acl e cho oe a ao aan eeinn tt aes Samm SMa Come tr ants 577 Ty: NV [Gece VOY 0 LSU pea tera Rte ee GHEE Did DE cari EL cree Ma PRD ei cabs reid ae tele hs on 578 a Numencaliratio between! the sexsi (ese. eit 1922 42 ene i=in etre) etee ste) eee 580 A NAGEL Gc oncaasmoe sbabo0cedeooc ode pusodogeadoorbonIgedporoddosDoosAaSOSOeus 580 G Coatdlaitessosdegocanossocsach suagcugoncomencoreaqweccnbaGdphonodginbosenas 587 Nive Generalediscussiomonthe:datalses eet actrees rere rye Sie cityecs io ire ot ne eee 597 1 Numericallratio between the Sexesijass acticcrn fant hva asa ware cyaveynets wast asere weuapetehs -eea New 597 2. NEEL bsp ocagkcadisaretoudsebnnaasoone saad doonanmoginocougonocods shop od mn coe 600 Gee Couiparativenatia bili iyfOn taeiseXcs/emaran ratte tclels tiie) eee renee 600 i Wnerhaaancy polbypomscccodéokan coos Yoosrooboowucascoouoroeonssoboc doco lS 601 c Comparative variability of external and internal characters...................... 601 Ri (Cicn@omecsngopaddod scour docnpecuonapamacgouddagueDcidoccy omatighbousscega< 607 fe Comparative cOULelaplOnOn tie SCXG Samer leer celal ate etter ee OO b Comparative correlation of external and internal characters............-.-++...--. 608 Vinee i Cerca te Cite Oeeh thy Sern Sealy ct ney io eayANS ee Oe et ASP es, Sieve ue eres he BoccdMhorsob oe co: 612 I INTRODUCTION The present study was undertaken with a view toward getting information first as to the degrees of correlation and variability of internal characters, such as the viscera and muscles, regarding which our knowledge is remarkably deficient, and second as to the general condition of correlation among a considerable number of characters in a single group of individuals. It has been pointed out frequently that selection acts not upon single characters or variations alone, nor as such, but upon the entire organism as the unit: that there results from this action a “balance of fitness”’ Tue JouRNAL or EXPERIMENTAL ZOOLOGY, VOL. IV, NO. 4. 576 Wm. E. Kellicott among a large number of characters—in fact, throughout the entire organization. ‘This condition of balance is to a certain extent measurable by coeficients of correlation which are merely indices of the corresponding change ina given character accompanying a certain change in another given character—‘‘relative” and “sub- ject.” The ideal method of approaching this matter is of course through multiple correlations involving the relation of at least three characters but this method is not practicable because of the time involved in carrying through such calculations. Consequently the method adopted here is merely to calculate the coeflicients of cor- relation between all of the measured characters 1n pairs. The material from which the data were drawn was unusually favorable for such a study. ‘Toads were readily secured, were not affected by capture and were easily kept in nearly normal condi- tions during the brief time elapsing between their capture and measurement. ‘The group studied was a perfectly homogeneous one, all in excellent nutritional condition, with precisely similar environmental conditions, and quite isolated geographically. The number measured and weighed was not large absolutely (425) but the fact that practically an entire colony of the particular variety under observation was collected and measured is almost unique. ‘The results are therefore at least free from errors due to the sampling of a larger population, errors which easily may be so great as to vitiate results even though a most careful preliminary study of the organism and its habits may have been made and particular care taken to procure a “‘random” sample. The fact that the subjects were of various ages, the growth phe- nomena not being taken into account, renders the data given here of no value for certain special purposes, but for the relations which are being sought here this is not an objection. We are dealing here with the conditions of correlation and variation in a total natural society of normal individuals. II SUMMARY ge ; : The material under consideration consisted of practically an entire colony of the common toad. Accurate measurements were Correlation and Variation in the Toad 577 taken of thirteen characters, including both external dimensions and internal organs. ‘The numerical ratio between the sexes was found to be 658 males to 1000 females. ‘The sexes are perfectly distinct with respect to size, variability and correlation. ‘The females are on the whole about 24 per cent larger, 23 per cent more variable and 10 per cent better correlated than the males. ‘The internal characters are about four times as variable as the external. The ratios between the average values of pairs of characters remain the same in the two sexes. ‘The distributions are all skew and nearly all negatively, apparently the result of including individ- uals of all ages over one year. The correlation coefficients are all relatively high. The exter- nal characters although less variable than the internal, are in the males 51 per cent and in the females 30 per cent better correlated than the internal characters. “Those individuals above the aver- age in any pair of characters show much less “‘scatter’”’ about the regression line than those below the average. In the general discussion these results are compared with those of other observers. Some of the points mentioned are the relation between efficiency and mass or dimensions in external and internal characters; the extremely high variability of internal characters; the relation between variability and correlation. The conclusion is reached that from the side of fitness or sur- vival conditions of correlation here seem to be more fundamental than conditions of variability, and that the general subject of cor- relation is of increasing importance. III THE DATA I Material The subjects from which these data! were secured consisted of a society of 441 individuals of the Common American Toad (Bufo lentiginosus americanus, Le. C.) |The society was one found on Cedar Point, Ohio, a low sandy point 200 to 300 yards wide extending for six or seven miles obliquely into Lake Erie and 1 The data were secured during the summer of 1905 at the Lake Laboratory of the Ohio State University. 578 Wm. E. Kellicott partly enclosing Sandusky Bay. “Toward the extremity of this point a colony of toads has become established the individuals of which differ in several minor respects from those of the main-land. The colony is fairly isolated geographically and inhabits the lake beach for a distance of about a mile and a half. During the day- time the toads lie buried about two inches below the hot surface of the sand and only a few feet or yards back from the water’s edge. Shortly after sunset they uncover and hop down to the water to pick up food carried in by the usually light wash. Food is more than abundant throughout the season and with no particular exer- tion all are able to maintain themselves in an extremely well fed condition. As it becomes fully dark they assemble in this fashion, sometimes just within reach of the water and at the beginning of the season were picked up easily in considerable numbers. Dur- ing the summer almost daily collections were made and toward the end subjects became so rare that ultimately only a dozen or two could be found during an entire week. The data therefore were drawn from practically an entire popu- lation occupying a uniform stretch of sandy beach about twenty feet wide and a mile and a half in extent where all were subject to conditions which were remarkably uniform though somewhat unusual for creatures of their kind. And while the total number of individuals measured was not large absolutely, yet it represents nearly 100 per cent of this homogeneous group and consequently the observations are free from errors such as result from the sampling of a larger population. 2 Methods All the toads one or more years old were collected: there is no way of determining their exact age. Since such characters as total weight or weights of viscera are liable to modification by the unusual conditions of confinement, care was taken not only to keep these conditions as nearly normal as possible but also to measure animals only recently removed from their natural sur- roundings. Immediately after collection in the evening the toads were placed in a large sand-box and no more were taken than could Correlation and Variation in the Toad 579 be measured the following day. In no case was any measurement taken from an individual which had been more than twenty-seven hours in captivity. Fig. 1 Ventral view of female toad in position for measurement. One-half natural size of toad weighing 41.0 grams. The characters measured and methods of measurement are briefly as follows: (Compare with Fig. 1.) 1 Total weight. The toad was brushed clean and weighed alive. “There was a gradual loss in weight during captivity chiefly due to defecation and evaporation of water. The rate per cent at which this loss occurred was determined in four series of different sized toads and the proper correction made for each individual according to hours elapsed from time of collection. Usually upon opening the abdom- inal cavity the bladder and alimentary canal were found empty; in the few individuals where this was not the case, proper correction was again made. The toad was then pithed (brain and cord) a broad incision being made to permit free loss of blood. While taking the succeeding measurements the toad was placed on its back and the blood drained off completely. For the next measurements the legs were placed in the position shown in Fig. 1. 2 Length of body. From tip of nose to end of body between thighs. ‘This of course includes the head. 3 Length of thigh. One-half the distance between the middles of the knee-joints when in the position shown in the figure. 4 Lengthof shank. Distance from middle of knee-joint to middle of ankle-joint. 5 Length of foot. Distance from middle of ankle-joint to tip of longest (second) toe when fully extended. 6 Length of leg. Sum of thigh, shank and foot. 7 Totallength. Sum of body and leg. 8 Width of mouth. Transverse extent between angles of mouth when closed. 9 Length of head. Distance from tip of nose to postero-dorsal margin of cranium exposed by pithing incision. 580 Wm. E. Kellicott 10 Weight of gastrocnemii. The ankle-joint was completely flexed and the tendo Achilles sectioned _ in a radius of the joint. The attachment of the muscle to the femur was then cut along the bone and the muscle removed. Both muscles were weighed on a balance sensitive to one milligram. They were handled only with forceps, by the tendo Achilles and were not allowed to dry. 11 Weight of liver. The abdominal cavity was then opened and the liver removed by cutting through its attachments and blood-vessels closely along its surface. It was then gently rolled in a towel until it ceased to stain and weighed as above. 12 Weight of ovaries. These were removed in same manner as liver except that being practically bloodless, they were not rolled. 13 Length of alimentary canal. The trunk was then completely divided, the mesentery cut through and the alimentary canal straightened out to its full extent by pulling gently. A condition is soon reached where no farther stretching occurs and the length was measured when this point was reached. This character may be considered as little subject to error in determination as any visceral character, since a number of tests showed that this method of measurement gave a very reliable datum, much more reliable than was expected. A résumé of the data will be given first as briefly as possible, discussion of their significance being deferred until they have been completely presented. 3. Numerical Ratio Between Sexes Preliminary examination of the data shows at once that the sexes must be treated separately. Of the total number measured 173 were males and 252 females but these numbers are not quite indicative of the actual ratio, as sixteen additional individuals were collected and used for purposes that might have affected the values of some of the internal characters and which therefore were not measured. Of the total 441 collected 175 were males and 266 females, giving a ratio of 658 males to 1000 females in the entire colony. 4. Variability The means and standard deviations of all the characters in the male and female series, calculated according to the usual methods? are given in lable I and the coefficients of variability in Table II. ?Formule and methods from Davenport (04). eg ee f) Ey = + 0.6745 — Vas aw n — i ~ (?. f) Eg -= a: 06745) = Vo ce o G Cc 243 Cc =— . Ico Eg = + 0.6745 Sy it all (oa M V 2n Wee Correlation and Variation in the Toad TABLE VE Means and standard deviations of all characters 581 MEAN STANDARD DEVIATION oO n=173 2 n=252 On=173 9 n=252 Motaleweiphtoramistcimcl.)1 ears | 33-54ob0.364 52.560+0.617 7.105-£0.257| 14.526+0.436 Mo talilemethe mars crreyeicletercrsrelos lee 160.405+0.518 183.690+ 0.628) 10.107 0.365) 14.78310.444 Length of body,mm. ............ 67 .8710.221) 78.1130.292) 4.319-0.156| 6.877+0.207 Length of thigh, mm ............. 29.5750. 101 33-690 0.098 1.97040.071| 2.538-0.076 Length of shank, mm.............. | 23.020+0.080 26.325-+0.094 1.5600.056, 2.224: 0.067 Wenethiofioot pam) | 40.840+0.133) 46.506+0.155| 2.5940.094) 3.657-0.110 Wenpthiotiler. name. tte eee 92.8120. 309| 105 .gg0-+0.355, 6.035+0.218 8.359-0.251 Menpthorhead mm. sec)-1 cele oe 18.000+0.045| 20.052-+0.051| 0.872-0.032) 1.205-+0.036 Width of mouth, mm.............. | 23,.202+0.074| 26.650+0.087| 1.4370.052) 2.044-+0.061 Length of alimentary canal, mm .. 353-873 I.779)429 .603+2.253) 34.6704 1 .255| 53-01841.593 Weight of gastrocnemii, decigrams... 6.832+0.086 10.6350.131 1.6730.060 3 -0770.092 Weight of liver, grams............. | 1.5900.027, 2.233-0.034, 0.530+0.019 0.794-10.024 Weight of ovaries, grams.......... | | 5-2530.209 | 2.79§0.145 | \ | TABLE II Coefficients of variability | on=173 9 n=252 Cc Be G 1B ANCESTORS cA bene doa oomodion Oocke neg cols Coe ook 21.180.80 27.64+0.89 ALOGEII NER, 6oA be Bean pb pcobmcacaanssanoe boo so near 6.300.23 8.o5+0.24 Iban OH ChPes5cunsoeocpedodoonocudnoovondccgont 6.36-0.23 8 .800.27 lbeaveqinl Gi idak(els Rou os Gad eodoR Bee nn epeCouOoD cae 6.66--0.24 AGRE Os 28 Wengthiolyshan kame ere csc tasers: ekesrerlsteceel ay | 6.780.25 8.45-0.25 VARIN! WODE, 5o go gonovooddonngAuonOnAObOnH Au.D. 6.350.23 7.8610.24 [Lali rios ieee eh oe a aon copa SuSmmeemoe ere boo sociate 6.50+0.24 , 7-89+0.24 Wenlecinoishieadasetrte ctryrtere sue etreierst terse pensiaalenaeh tenet 4.820.17 6.01+0.18 Wirdthionmo wthisteeee senate eco ola co eieietsracereerosele | 6.20-0.22 | 7.6710.23 Wenethtotalimentany Gana lemese- lalalaliii ck | 9.810. 36 12.34+0.38 Nietehitiotica strOGn enilivee pe titers cise acre) sera 24.500.94 28.94+0.94 Wierghtotwliventasscaercryss saci ti teicic ae ceicerreir 33 -34E1-33 | Srlolspae Haze) WIGiEle OO Fin Secccodqausccundagd madegudaocos mob | 53 -233.-39 These tables show several facts of interest. First, the absolute distinctness of the male and female series as regards size, the female showing the larger values in every character measured. On the whole the female is about 24 per cent larger than the male, the actual 582 Wm. E. Kellicott percentages varying from 11.4 in length of head to 56.7 in total weight. This same distinctness between the sexes was found in the calculation of all constants and consequently they were treated separately throughout. Second, there is a corresponding distinctness between the males and females with respect to the amount of variability. “The coeffi- cient of variability may now be accepted as a thoroughly reliable measure of comparative variability. As expressed by this coefh- cient the females show a uniformly higher variability. On the whole the females are about 23 per cent more variable than the males, the actual percentages varying from 6.6 in weight of liver to 38.4 in length of body. Finally, inspection of the coefficients of variability shows that the characters are separable into two quite distinct groups. ‘The external dimensions show not only a comparatively low degree of variability but a remarkable uniformity, the limits of C being, with a single exception, 6.0 per cent and 8.8 per cent. ‘The other characters, which we may call “internal” are several times more variable and show much less uniformity, the limits of C being 9.8 per cent and 53.2 per cent. Excluding the total weight, total length and length of leg as composite characters, we find that the average coefhicients of variability of the external and internal char- acters are as follows: of g Per cent Per cent External characters’. ;asemE soe 2 ese ocoSaSmGA be Goose Ddsdc dha SuccckrosedodUs. 743 IDeniai (dita NGcareur saree ndvercdn ta sch G005 00 oducsioo en onoeHdacntdagoor 857 IbGits lib \leqnmpann secorodo bode canoe Opp acads oc cba dc aso GaapoucuoDsar S755 IDEM Gop oroga cee codec agodoemadsesucnobar Uoeorne odoccercuansduaKoes -910 Weneth headmasanss ee ean ts see oe ete | eet ttn ater 711 Wid thimo thins. vate sparserbscs Steensie eters ay oleh shale reenete yates tar ennieyatetaie onto cowed re -788 Weight pastrocnemli... .. 20... secs cece nce e tect eee ence essa ceceenenesece .866 \iGanehSerqenonuesercpooasenconbs sor adbbcmpanaconuetebopodsococone6t 539 enpthtalimentanyicamalllers «we sete 1elstec Pet oles oled toll ote ore fe Yet cecal sketa tater 452 Wreightiovaries ser tis-teyecrsciata tela & -1elaie ofetetote ap oteeteied en rsch Peheeot- Pols a a Averape’all characteris criel-1cte cite «ce ete e eleiet yphe etal toto) ier ielepelol=)= tet ateiat- telat -747 Average external characters. .... 0... 2 cesses ese teen eee net seer ee 803 Average internal characters... ........-- 22sec cece eee e ete e te e eee 495 TABLE X Indices of correlation between length of foot and Weight total... ao decautans saw sence tavate cr ayeioves alate one aah se apres en aie aiete 864 Db Sorta beigao leadot ed obe ooasod ds oped DoD dace neeD cab ooemmaEarrD yoo 0S 942 ISNA Oh Pansndeavane Tac es adaSoongpae onc odes naGGoUsSaoRcSasraonoC 863 [enim etd Anon oo odonu aoe JabEoddougosd saneL ode snadan saueD ono xcddo oo 865 Ib Guy aAGueNl aaa Gou sop aoide bouU br SonoGS cUcAd dD oeeenasSqnncovoodeUuICGUt 755 [beryl uses aa dnsbouspocesoacoonannenan soos och cououasssuGn ue cee eleteienss -954 Wen cthihea diy. semetyeatcrctie teria stet reste leveteke rate (aletedete selec tetanic ne ealatniototpate -713 Wiser) inna goeimonb GOT AD OG Meador boasts Obs dno suebob snap socnoaase 764 Weight gastrocnemit.. 5c \ 2 an). ci seer ole eo ciel nial wie reins nieimioie vinnie eisies os -813 WIGS tynon catodoouee dose bre cooneamopoAooansegGae —dovucono0uDt 643 Mengthialimentary/camal). 5. -je.-1c\c «ce [es ewe aye ne oom = lee eee) ial =) ese -366 IWietaht OvaTdeSti facie jets « eronsistere spe hese ol eaageraieletetes homers hse toluene vel oh fens epee 2) late talers Average all characters... atsyoiae leis eet ale odeto othr .700 WiviemanGialimentaryscanaleneoti ics chery cio eer teem Soca erase eine oe . 500 MI y Cla TIO Valle Saye pet eaters i cele rate Aad Shen farewete evcaha es Sec a Se recat .702 Alimentanyacanal andlovianl Sancta simran cient stele citer serene 465 Mariousiexternal chiaractensy avicha ver tiesto (cherie ite eee eee 875 Human— Strenethiot pulliand hetpht @Pearsonlg9))saaei eases pe eee ne eee ae -216 to .303 Strengthiof pullland) weight (Pearsom99))s2.-/-).2-eey oe eiieieee acess -338 to .545 Mental abiiity and various physiological characters (Pearson ’06)............ .06 to .45 Mental ability and various structurai characters (Pearson ’06)............. -00 to . II skeletal characters. ‘This indicates that the characteristics of the skull are more dependent upon the brain characters than vice versa. The relation between variability and correlation as indicated by these and other data may be mentioned here. We have seen that in the toad conditions of high variability and high corre- lation are associated in the female sex. The same relation has been noted in the male of Eupagurus by Schuster (’03), and among 610 Wm. E. Kellicott plants a similar relation has been found between variability and correlation. (See e.g., Weldon’s (’o1) recalculation of MacLeod’s data in Ficaria.) On the other hand, we have seen also that in all the data available in internal characters the higher variability is associated with the lower correlation. As a specific instance may be mentioned some of the figures given by Greenwood (’04). In her study of the human viscera she found that the normal con- dition is one of low variability and high correlation as contrasted with a relation of high variability and low correlation among those diseased or in “general poor health.” Pearson (’o2b) in one of his Mathematical Contributions to the study of evolution demon- strates by a purely mathematical method that natural selection must determine the amount of correlation, that indeed, it is probably the chief factor in the production of correlation throughout the constitution of the organism even though it may operate occasion- ally upon single characters. This certainly agrees with the bio- logically determined facts. He farther establishes mathematically the general principle that “intensity of selection connotes a lessen- ing of correlation,” that a condition of lower variability must be associated with lower correlation. For example supposing the correlation between the lengths of tibia and femur to be .70, if selection reduces the variability of the tibia by 50 per cent the correlation will then be but .44—a reduction of 37 per cent. The correlation between organs related only indirectly would be simi- larly reduced though not to quite the same extent. Pearson does not make it sufficiently plain that this relation would hold only for the same pairs of characters before and after a period of selection. It is evident that, speaking in general terms, a condition of higher variability is frequently associated with a higher correlation, and that also a lower variability is associated with a lower correlation. But that this is not a necessary relation is clear when we compare the internal and external characters on this basis. That the relation mentioned by Pearson 1s not a neces- sary one is again evidenced by Greenwood’s data. ‘The following table summarizing some of these data shows that the ae of the weights of ce viscera is higher but the correlation at the same time he er in a general ase population than in a popula- Correlation and Variation in the T oad 611 tion classed as “healthy,” 7. ¢.. the relation between variability and correlation is inverse. General Hospital “Healthy Population Postmortems” Coefficient of variability in (males) (males) Niele bGOmneatbarserrntertr nce ct Aetee el sicher @ elle 32.39 iG 7 VMSA GH INiaete oo ooo edness 00 Ole Gere ain asic Paik ip? 14.80 Wala. oneilaane set 6 hoo tedion ono ares oa 50.58 38.21 Wielchtofileidineys.pcemetrs tssier sie ct helena tatcyors 24.62 16.80 Coefficient of correlation between EV ean tran cuunverscrrrc cers. crerevste se vercr.oienekers Susliayens 1931 .2780 ldleninesinelepal lash, sonobnaduéoe ato abe OOO oo 1827 2654 iEleartandiucid me yperctesteystel s)he 'e clrai ore telelielleraueaie SET . 4004 And again, with increase in age the variability of the heart and spleen was found to remain constant but the correlation coefficient increases steadily from .08 (25 to 35 years) to .25 (45 to 55 years). How this relation between variation and correlation can be explained mathematically is not clear but doubtless a mathe- matical “description” will be forthcoming once the biological fact is established. The last point to be mentioned is the relation of the groups of larger and smaller individuals, 7. ¢., those above and below the average, to the regression line. We have seen that among those above the average in any respect the “scatter” about the regression line is markedly less than among those below the average. The full significance of this relation needs farther investigation but I take it that this is an indication that the larger, and therefore in general the older, individuals gradually approach more closely the type as measured by the coefficients of correlation. From the charac- ter of the distributions found in this material this would seem to be the result of growth rather than of selection upon the basis of variability. That the coefficients of correlation as well as of variability change with age is well known. ‘That is to say in growth, organs behave as more or less independent units. It seems likely that normal processes of growth may consist to an important extent in the gradual approach toward a condition of more perfect general correlation. Data bearing upon this impor- tant point are still meager; it is unfortunate that the age of these 612 Wm. E. Kellicott toads was not determinable with sufficient accuracy to afford evidence. In conclusion, it seems that the general subject of correlation is one of increasing importance. Attention has been so largely given to the subject of variation that the facts of correlation, which may prove to be the more significant, have not received their due share of attention. The “balance of fitness” which is the result of growth or selec- tion is not a condition merely to be guessed at or estimated; it is measurable by a series of correlation coefiicients among a consider- able number of characters. The “summum bonum”’ 1s a condi- tion of high correlation, be the associated variability higher or lower. ‘To be fit an organism must exhibit throughout its organ- ization a certain degree of proportion of its parts. “This condition of fitness is measurable by a series of correlation coefficients not only of external dimensions and skeletal characters but also of size and strength of muscles, the position of their insertions upon the skeletal elements, the size (though only an approximate indi- cation of efficiency) of viscera and especially of nerve centers, and, were the technique possible, also of the eficiency of sense organs and of the accuracy and intensity of reactions. And lastly, looking in this direction, studies in correlation should include not only the relation of pairs of organs in a large number of individuals but farther of the relation between a large number of representative characters among single individuals. ‘The collection of data upon this last topic is now in progress. The Woman’s College of Baltimore, Md. June, 1907 V LITERATURE CITED Bumpus, H. C., ’97—A Contribution to the Study of Variation. Skeletal Varia- tions of Necturus maculatus, Raf. Jour. Morph., xii, 455, 1897. Cuénort, L., ’99—Sur la determination du sexe chez les animaux. Bull. Sci. de la France et Belgique, xxxii, 462, 1899. Original paper not acces- sible—abstract by the author in L’Année Biologique, xv, 212, IgOI. Davenport, C. B., ’04—Statistical Methods. 2d ed., New York, 1904. Correlation and Variation in the Toad 613 Davenport, C. B. and Butiarp, C.,’96—Studies in Morphogenesis. VI. A Contribution to the Quantitative Study of Correlated Variation and the Comparative Variability of the Sexes. Proc. Amer. Acad. Arts and Sciences, xxxil, 85, 1896. Dona.pson, H. H., ’98—Observations on the Weight and Length of the Central Nervous System and of the Legs in Bull Frogs of Different Sizes. Jour. Comp. Neurology, viii, 314, 1898. Fu.ton, T. W., ’91—Observations on the Reproduction, Maturity and Sexual Relations of the Food Fishes. Annual Report of the Fishery Board for Scotland, x, 11, 232, 1891. GreENwoop, Jr., M., ’04—A First Study of the Weight, Variability and Correla- tion of the Human Viscera, with Special Reference to the Healthy and Diseased Heart. Biometrika, 11, 63, 1904. GriEsHEIM, A., ’°81—Ueber die Zahlenverhaltnisse der Geschlechter be: Rana fusca. Arch. f. d. ges. Physiol., xxvi, 237, 1881. Howe 1, W. H., ’05—Text-Book of Physiology, p. 38, Philadelphia, 1905. LonnBeERG, E.,’93—Ichthyologische Notizen. Ueber die Variabilitat bei Petromy- zon. Bihang till K. Svenska Vet.-Akad. Handlingar, xvii, Afd. iv, no. 2, 1893. Minot, C. S., ’91—Senescence and Rejuvenation. First Paper: On the Weight of Guinea Pigs. Jour. Physiol., xi, 97, 1891. Montcomery, T. H., ’96—Organic Variation as a Criterion of Development. Jour. Morph., xu, 251, 1896. Peart, R., ’05—Variation and Correlation in Brain Weight. Biometrika, iv, 13, 1905. Pearson, K., ’97—Variation in Man and Woman. Inthe Chances of Death and Other Studies in Evolution. London, 1897. ‘99 Data for the Problem of Evolution in Man, III. On the Magni- nitude of Certain Coefficients of Correlation in Man, etc. Proc. Roy. Soc., Ixvi, 23, 1899. ’o2a—On the Correlation of Intellectual Ability with the Size and Shape of the Head. Proc. Roy. Soc., Ixix, 333, 1902. ’o2b—Mathematical Contributions to the Theory of Evolution, XI. On the Influence of Natural Selection on the Variability and Correlation of Organs. Phil. Trans., A., cc, 1, 1902. °06—On the Relationship of Intelligence to Size and Shape of Head and to other Physical and Mental Characters. Biometrika, v, 105, 1906. Pricer, E., ’81—Einige Beobachtungen zur Frage tiber die das Geschlecht be- stimmenden Ursachen. Arch. f. d. ges. Physiol., xxvi, 243, 1881. 614 Wm. E. Kellicott RosENTHAL, M. J., '67—Note sur la force que le muscle de la grenouille peut dévellopper pendant la contraction. Comptes Rendus, Ixiv, 1143, 1867. Scuuster, E. H. J.,’03—Variation in Eupagurus prideauxi. Biometrika, u1, 191, 1903. Srrone, R. M., ’or—A Quantitative Study of Variation in the Smaller North American Shrikes. Amer. Nat., xxxv, 271, 1901. Vernon, H. M., ’95—The Effect of Environment on Development of Echinoderm Larve. Phil. Trans. Roy. Soc., B., clxxxvi, 1895. ’03—Variation in Animals and Plants. New York, 1903. We pon, W. F. R., ’o1—Change in Organic Correlation of Ficaria ranunculoides during the Flowering Season. Biometrika, 1, 125, 1g0I. a =X oetetetetete= : etesreenrestsreteterssece ects = — oss St = SHERI See esseee Serre teteteer eters Stoertot re pistes tt i } i, = tie sehe te rHeiehe dd! tetas ANH nie) 1 4 i t 5 =tae es oe ne ye) sete iste las seis p—o~ eee ets 7 Se phets tt i if eis bode she] | 13) Ssessersestieitttssstesebeeers th} >. Gre Oreo tse aoe ee erer es eto ts