ae gO Sea eke rat al Lee ee ar te ph t i t as ee | ngs t +e § ; ‘f ft 4 sf oaca 2 tee ee Sob Way er i ‘, ik RwCt te Sete he . e. fae we shehok eseiptgt cit 8) #9 4 ak pe “Sale Tal ba st c~e— Sata 3 te 6-6. ° wasictatratn's ate cass ststnerscemee coe a Be Oi s " 2 i S — ee . 5 “3 ot pe hear he ee “ eseses ome ee gases Po.» 2 ‘ ¢ eh Sad “Ved to -.2 wo = bua. o8-e8 ee on : fA hl ND et ns hh Aeon ea EY A ie an o-4 058. 8272070288 8k - al” lle Nae el é “ seteesas +8 8.4.8.4) “ nl Stott = #edn@secocee ms Mek eh ee eter a 0268 - = S = iis 3 saree, = te Boa os x #55882 aks . > acegist soon Bee ; sf 3° : ee} pe ren ete a oe. a 3) oe * 3 m. : ; : > he ae ah ©, . W595 % Foe 3 o" 5 +e. : : >. oe eeeeene raters Fag Pin Bn eee ee ett =: Pt a8 --8, —— Pa 8% Bo THE JOURNAL OF EXPERIMENTAL ZOOLOGY EDITED BY Wiuiiam E. Caste Jacques LoEB Harvard University The Rockefeller Institute Epmunp B. WILson Epwin G. ConKLIN Columbia University Princeton University Tuomas H. MorGan CHARLES B. DAVENPORT Columbia University Carnegie Institution H g Grorce H. PARKER ERBERT 8S. JENNINGS Harvard University Johns Hopkins University RayYMOND PEARL FRANK R. LILLIE Maine Agricultural University of Chicago Experiment Station and Ross G. HARRISON, Yale University Managing Editor VOLUME 21 1916 THE WISTAR INSTITUTE OF ANATOMY AND BIOLOGY PHILADELPHIA, PA. CONTENTS NO.1 (JULY Lewis R. Cary. The influence of the marginal sense organs on the rate of regeneration in Cassiopea xamachana. Eleven figures.............. Bae S. O. Mast anp F. M. Root. Observations on ameba feeding on rotifers, nematodes and ciliates, and their bearing on the surface-tension theory. IBV CRITE CS een eee et ley reece ein. hme one en ane eh Myctee Ural A 3 OnerA A. Merritt Hawkes. The effect of moisture upon the silk of the hybrid Philosamia (Attacus) ricini Boisd. o X Philosamia cynthia CLEVER a) VS? 2A ae Ra 78 a CO Cae eo a H. B. Goopricu. The germ cells in Ascaris incurva. Eleven text figures BUTE UHMGE MO LADES ets sete yeas pie a ate archi cn haynes WRIA cia eR al ec eee Ie C. M. Curnp. Studies on the dynamics of morphogenesis in experimental reproduction and inheritance. IX. The control of head-form and head frequency in Planaria by means of potassium cyanide. Ten figures.... A. FRANKLIN SHULL AND SonrA Laporr, Factors affecting male-produc- Pomme Livgamnae, (One. fisupean. ss. oc ooh 2c shee eee Wats +00 -- 0 55> Al soe A. R. Moore. The mechanism of orientation in Gonium. ‘Two figures.... ul 33 ol 61 199 213 281 295 347 lv : CONTENTS NO. 4 NOVEMBER . G. H. Parker and E. G. Trrus. The structure of Metridium (Actinoloba) marginatum Milne-Edwards with special reference to its neuro-muscu- lar mechanism. Seven figures (one plate)......... REAR Tt ois 5 eee 433 G. H. Parker. ‘The effector systems Of actinians....:.......a.7eeeeees . -- 461 E. I. Werser. Experimental studies on the origin of monsters. I. An eti- ology and an analysis of the morphogenesis of monsters. Eighty-nine THE INFLUENCE OF THE MARGINAL SENSE OR- GANS ON THE RATE OF REGENERATION IN CASSIOPEA XAMACHANA LEWIS R. CARY Princeton University ELEVEN FIGURES INTRODUCTION The previous studies on the influence of the nervous system upon regeneration have given very divergent results which can hardly be reconciled even when the fact that widely separated groups of animals were used as the material for experimentation is taken into consideration. While on the one hand certain students of this problem (Herbst, Goldstein, Walter, Wolff) have taken the position that the nerv- ous sytem in general, or some portion ofit (sensory ganglion Herbst, Walter), exerts a stimulus necessary for the complete regeneration of normal structures; other workers have attributed less and less importance to these influences. The intermediate position that the influence of the nervous system is indirect, being exerted mainly through the controlling of motor activity is well expressed by Child (05a) in the statement concerning anterior regeneration in Leptoplana that, ‘‘as’in posterior regeneration there is a close parallelism between the rapidity, amount and com- pleteness of anterior and lateral regeneration and the charac- teristic motor activity of the part concerned.” Goldfarb (’09) coneludes from his experiments on newts, earth- worms and planarians that ‘“‘these experiments should make one cautious about accepting the view of the direct or even indirect influence of a nervous influence on regeneration.” In all these studies the point at issue has been whether or not complete regeneration of typical structures is possible in the ab- 1 THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 21, No. 1 JuLY, 1916 2 LEWIS R. CARY sence of any influences exerted through the central nervous sys- tem. An affirmative answer to this question is apparently held, by certain at least of these investigators, to settle finally the question of nervous influence without any consideration being given to the comparison of the course of the regenerative process in animals in which the nervous system was removed from the regenerating area and those in which the nervous system had been uninjured in the portion of the animal left to regenerate. In only a relatively few animals can the nerve centers be removed without bringing about the destruction of, or degenerative changes in, other intimately connected portions of the nervous system, so that this type of operation has not been frequently undertaken. Zeleny (’07) and Stockard (08) removed the marginal sense organs from the disk of Cassipoea xamachana to determine the influence of these structures on the rate of regeneration. Both report that there was no evidence of any regulatory influence. In Zeleny’s experiments the entire margin of the disk with its sense organs was removed and the rate of regeneration in these individuals compared with others in which the bell margin and sense organs were intact. In Stockard’s experiments the re- sults obtained from specimens prepared as above described were supplemented by those obtained with individuals from one-half of which the marginal sense organs were removed while from the other half an equal amount of tissue was cut from be- tween the sense organs. The two halves were insulated by the removal of two diametrically opposite strips of subumbrella ectoderm. In both experiments the rate of regeneration was measured inward from the periphery of a cavity in the center of the disk from which a circular piece of tissue had been removed. In both these researches the experiments were carried out with the view of ascertaining the influence of muscular activity and thus indirectly of the nervous system on the rate of regeneration. In each case it was held that there was no constant difference in the rate of regeneration between the active and inactive individuals. . RATE OF REGENERATION IN CASSIOPEA 3 In the course of my studies, which were taken up primarily to reexamine the work of Stockard and Zeleny upon this point, I discovered the marginal sense organs influence regeneration, inde- pendent of their control of muscular activity. Such an influence of the sense organs can be accounted for either on the ground that metabolic activities, not expressed by muscular activity, are under the control of the sense organs, or that a direct trophic influence is exerted by the sense organs on the regenerating tissues. A series of determinations with the ‘biometer’ of the rate of CO: production by specimens under different experimental conditions, for which I am indebted to Dr. 8. Tashiro, shows that the first of the two alternatives just mentioned offers a satisfactory ex- planation of the observed facts. It is a pleasure to acknowledge my indebtedness to the author- ities of the Carnegie Institution of Washington for putting at my disposal the facilities of the laboratory at Dry Tortugas, Florida, and especially to Dr. A. G. Mayer the director of the laboratory for many helpful suggestions and constant interest throughout the course of the work. TECHNIQUE As Mayer, Stockard and Zeleny have pointed out the disk of Cassiopea will live for an indefinite period after the removal of the oral arms and retain its full capacity for regeneration. These nearly flat circular disks with their sixteen equally spaced sense organs and with the ectodermal musculature entirely on the sub- umbrella surface offer exceptionally favorable material for the study of the phenomena of regeneration as they will withstand practically any type of operation. The medusae can be procured in great numbers from the moat at Fort Jefferson at Dry Tortugas, Florida, so that specimens of any desired size can be selected for experimentation. In pre- paring material for my experiments specimens of about 100 mm. in diameter were chosen, the oral arms and stomach removed as soon as they were brought into the laboratory and the operations, of whichever type, performed as soon as convenient, so that in 4 LEWIS R. CARY every case the experiments were started within two hours after the medusae had been removed from their normal surroundings. As the normal habitat of Cassiopea is in rather stagnant quiet water, the disks retain their vitality indefinitely when a single pair are kept in a medium sized battery jar of sea water. In- deed Stockard found that the benefit derived from the daily change of water was more than offset by the harmful effects of the agitation attendant upon the changing of the disks from one jar to another. Since my experiments necessitated the daily measuring of the regenerated tissue which could be done only by removing the disks from the Jars and placing them upon a back- ground of colored glass, the water was changed daily. In specimens in which the halves were insulated it was neces- sary to scrape over the denuded strips at least once every forty- eight hours, as within that space of time new nerve fibers would be regenerated connecting the old nerve fibers of the insulated halves and consequently transmitting the stimulus necessary for pulsation from the sense organs of the half disk on which they were retained to the muscles of the half disk from which they had been removed. In a few of the experiments in which regeneration was slowest, particularly some of those where the sense organs were removed from both halves of the disks, new functional sense organs were developed in the course of an experiment so that a second re- moval of tissue from the portion of the bell margin originally occupied by the sense organs was necessitated. In order to determine the influence of the sense organs on the rate of regeneration the following experiments were carried out. First those in which the influence of the sense organs was removed from an entire disk (Zeleny’s operation, figs. 1 and 2) or from one- half of a disk (Stockard’s operation, fig. 3) by the removal of an appropriate number of sense organs and insulation of the two halves (figs. 1 and 2). Second, disks from which all but one of the sense organs were removed, with the disk either left with its subumbrella ectoderm continuous or with the halves insulated (figs. 4 and 5). Third, disks prepared according to Stockard’s method were subjected to the action of anesthetics to eliminate RATE OF REGENERATION IN CASSIOPEA 5 muscular activity. Fourth, disks prepared in the manner just mentioned were treated with solutions of oxalic acid which will destroy the sense organs without seriously injuring the muscular system or the conducting portions of the nervous system. Fifth, specimens from which all sense organs had been removed and with the halves insulated in one of which muscular activity was initiated by means of an induction shock and maintained as a circuit wave of contraction in an endless labyrinth of muscle Figs.land2 Experiments of type 1. Fig1. Active disk with sense organs intact, and with pieces of tissue cut from the margin of the disk between them. Fig. 2 Inactive disk from which all of the sense organs have been removed. tissue, instead of being controlled, as normally, by nerve im- pulses arising from the marginal sense organs (fig. 6). Sixth, specimens with insulated halves in one-half of which muscular contraction was maintained by means of a circuit wave of con- traction while the pulsation of the other half was under the normal control of its sense.organs (fig. 7). EXPERIMENTS WITH ENTIRE DISKS When the rate of regeneration of a series of active and inactive entire disks is compared it is found that in about 75 per cent of all the experiments the regeneration is most rapid in the active : 6 LEWIS R. CARY specimens. In the remaining disks the amount regenerated at any given time is, in about 10 per cent of the pairs, found to be equal within the limits of accuracy of measurement, while in about 15 per cent of the pairs of disks regeneration was greatest in the inactive specimens. The results of many different kinds of experiments upon Cassi- opea have shown that there are wide variations in the sensitivity and metabolic activity in this animal. It therefore seems evi- Fig. 3. To illustrate the type of operation (Stockard’s) used in experiments of types3 and 4. The sense organs were removed from one-half of the disk, and an equal amount of tissue from the margin between the sense organs of the other half of the disk. The two halves were then insulated by scraping off two diamet- rically opposite strips of subumbrella ectoderm. The insulatingstrips are shown stippled in the figure. dent that the conflicting results obtained from specimens sub- jected to this type of operation are to be explained as individual variations in physiological activity. More dependable results may be expected from specimens prepared according to Stockard’s method (fig. 3) where individual variations in physiological activity are eliminated. After this operation the inactive half of each specimen is moved about by the pulsation of the active half so that there can be little difference in the degree of aeration of any two parts of the disk. In all experiments of this type (2) where large numbers RATE OF REGENERATION IN CASSIOPEA 7 of specimens were used two difficulties were met in making the measurements. Frequently the disk became folded backward at the point where the subumbrella ectoderm was removed some- times even bringing the exumbrella surfaces in contact. While this seemed in no way to interfere with regeneration it frequently made accurate measurements impossible unless the specimen was first narcotized, as any attempt to unfold the active disk usually resulted in tearing the delicate regenerating tissue. As this procedure involved the expenditure of so much time all badly folded specimens were discarded. If the folding took place some days after the start of any series of experiments the specimen was discarded and the figures for the earlier stages re- tained in the record. The other most common source of difficulty in making the measurements arose on account of the tendency of the edge of the regenerating tissue to fuse with the edge of the old cut sur- face or with a more proximal part of the sheet of regenerating tissue. Whenever the edge of the thin sheet of new tissue be- came folded back sufficiently to touch any of the more proximal tissues fusion took place so that a tube would be formed from the new tissue. When the folding involved only a small area sepa- ration could be easily accomplished, but if a considerable por- tion of the regenerating sheet was involved the specimen was rendered useless for further study. The results of two typical experiments are shown in table 1. The measurements are in millimeters. The upper figure for each date shows the width of the sheet of tissue regenerated from the active half disk, the lower figure the width of that regenerated from the passive half disk. When the sheet of new tissue had entirely closed over the cavity in the center of the disk the point of closure remained recognizable for at least a day so that the measurements could readily be made for those disks that had be- come closed since the time of the last measurements. By the end of twenty-four hours after the new sheet of tissue was com- pleted the point of closure would be shifted until it came to lie in the center of the disk. 8 LEWIS R. CARY TABLE 1 Record of experiment 4a started September 30, 1913. 40 disks each with all sense organs removed from one side and with the halves insulated by removal of 2 strips of subumbrella ectoderm DAYS AFTER OPERATION NUMBER OF SPECIMEN 1 2 3 4 5 , ae RON SHOT atx GA dole 3.25) 4N7S 9.50, Half without S. O......... 2.00 | 2.50 6.00 § fe ELAS OO fa eRe ee 2.50 | 4.00 5.00 9.00 Half without S. O......... 1.50 | 3.00 3.75 6.25 Haltiqvatnis Oe... 8h. oe 2.95 | 3.00 5.50 8.25 (6 : rae Sh net CRO OIG a 1.75 | 2.00 | 3.25 5.00 alt with s. OF. cose es 2.75 | 4.00 5.75 14.75 ‘ fee ont: OF i 1.75 | 3.00 | 4.50 | 10.50 ee ait NCI 6 etek eee Pao" |) 3275 5.50 7.25 Cl 9.00 Don inltamat out Gea. 2016o 1.50 | 3.00 | 4.75 6.00 | 7.50 : ree WSS Oc cae ae 1.50 | 3.00 3.75 5.00 6.75 C Half without S. O......... 0.80 | 2.25 2.75 3.25 C] 4.50 7 eae walt SiO, 6... eee 3.75 | 5.00 5.75 7.50 Half wathout S.'O......2008 1375 2.00 3.00 5.00 4 ee ens Ol. < - . eae 3.50 | 4.00 4.00 9.50 ¢ Half without S. O......... 2.50 | 3.00 3.00 7.00 9 (f Halt with 8.O>.. . : 2:95 3.25 4.50 6.25 7.00 | 8.25 Half without S. O...| 1.75 2.75 4.00 5.50 6.25 | 7.25 Wie oan al Ese, 2 a). 2300 S20 4.75 6.00 7.50. 138250 Half without S. O...| 1.75 2.75 4.00 5.50 6.50 | 7.25 a ee ale ek 2.75 4.00 5.00 6.50 8.00 C Half without S. O... Deo 3.25 4.25 5.50 7.00 3 ee alts 4o% 3.00 4.00 5.25 6.50 8.00 C Half without 8. O...} 2.50 3.00 4.50 5.75 7.25 ests: BR 1.00 2.25 3.50 5.25 6.50 | 8.00 Half without S. O...} 0.75 1.75 3.00 3.75 5.75 | 7.25 a fea halt, <2 1.75 3.00 4.25 6.00 7.50) #8425 Half without S. O...| 1.50 2.50 3.50 5.00 6.25 | 7.00 7 eee half... oa 2.50 3.25 4.50 6.25 7.25. 3200 Half without S. O...| 2.25 2.75 4.00 3.75 6.75 | 7.50 ie ee te ali... 3.50 4.25 5.25 6.75 8.25 CG Half without S. O...} 3.25 3.75 4.75 6.00 7.50 iG ree halts oe 1.25 2.50 4.25 6.00 7.50 | 8.50 Half without 8. O...} 1.00 2.25 4.00 5.50 6.75 | 7.25 14 aes halt ee ae 2.00 3.25 4.50 6.25 7 25) | S825 Half without S. O...| 1.75 2.75 4.00 5.75 6.50 | 7.25 1b os ITED heen 2.25 3.50 4.50 6.50 8.00 C Half without S. O... 2.00 3.00 4.00 6.00 Te25 ic Wen ca Healt: , see 1.00 2.50 4.25 6.00 7.25 \"8i25 Half without 8S. O...| 0.75 2.25 4.00 5.00 6.25 | 7.00 17 enn on Inailiges. ae 1.75 3.00 4525 6.00 7.25 8.00 Half without 8. 0...) 1.50 2.50 3.75 5.50 6.75. lh ae50 8 eee ralf?)))../'. 2.50 3.50 4.50 6.50 8.50 C Half without S. O...| 2.25 3.00 4.00 6.00 7.75 o a ine re 1.25 2.75 4.00 6.00 7.00 | 7.75 Half without S. O...;} 1.00 2.50 3.75 5.75 6.75 | 7.50 Bp { Activated half...... 2.00 Si) 4.50 6.00 7.25 "| 8.00 \Half without S. 0...) 1.75 2.75 4.00 5.75 6.50 | 7.25 RATE OF REGENERATION IN CASSIOPEA NUMBER OF SPECIMEN 21 Half without S. O... 93 ee halts... Half without S. O... {Activated half...... \Half without S. O... f Activated half...... \ Half without S. O... 26 arene raliee.. doc: Half without S. O... Activated half...... Half without S. O... Activated half...... Half without S. i sien i th { ctivated half...... 24 25 bo 7 iw) 8 iw) Ri) Half without S. O... Activated half...... Half without S. O... Activated half...... Half without S. O... Activated half...... iow) oO ise) _ ist) 2 33 Activated half...... Half without S. O... Activated half...... Half without S. O... Activated half...... Half without 8S. » ivated half...... { ctivated half...... 30 40 TABLE 3—Continued Be EE EEE _ {Activated half...... | Half without S. O... 99 Rca half. =... Half without S. O... { Activated half...... \ Half without 8. O... SA aie ali, S28 ce Half without S. Half without S. O... Activated half...... Half without S. O... Activated half...... Half without S. Activated half...... Half without S. ba | Oromo =! nSeeoeu ao —=) 00 rk FST URS) SITS HT OES SS SSO RC) NTE FT ENS FH OSS) CS TN TSS) IRE) TSO he Fae ae ak eae aS eel a, Dae arty ane Or mae OO (SS) (2) Ss Sas 0.75 2.00 1.75 2.00. 1.73 DAYS OF REGENERATION 21 - Mean of all observations 2 3 4 5 6 3.00 | 4.50 | 6.25 | 7.25 | 8.00 2.75 | 4.00 | 5.50 | 6.50 | 7.25 3.25" | 950. 1. 6200...) 7.00, |2825 2.75. | 24800 |. 5.50) (66.25 0) 225 3.00 | 4.50 | 6.00 | 7.00 | 8.00 2:75, | eezs00 | 5.25) | 906.25. 1) 7.25 3.751) meeO0) 196.25) +), B00). 1, 3125 eee: ae SSO! bye 725 2.50 | aoe |) 575 | 67.00 | 8.00 2.25 | ~AO0) |) 5.50 | 6.50 | 7.50 2.50 | 4.50 | 6.25 | 7.00 | 8.00 205. | AeO0i et /5,50 (|, 6.25.4) 728 3.25 | 4.50 | 6.50 | 7.25 | 8.00 37008, | 3275, (iS Th 6.50» eh. 25 3.25 | 4.50 | 6.25 | 7.00 | 8.00 2.75 | 4.00 | 5.50 | 6.25 | 7.25 3:00 | 4.75 |996c25 | 7.25 7) 8:00 27s | £00 |) 5 50yth 6.25 WPx-00 2.50 | 4.00 | 6.00 | 7.00 | 8.25 2255) 3.75 || 5.500) “6.50: WES.25 3.00 | 4.50 | 6.00 | 7.25 | 8.00 2.50 | 4.00 | 5.25 | 6.25 | 7.00 3.50 | 4.75 | 6.75 |- 7.50 | 8.50 3100 4 4:00. Sago le 6:75 |p mees BETS, VO.75 |, Sve C A765. gO | 75 3.50 | 5.00 | 6.50 | 8.00 | 4% shoo | 74895" |, - 5.7. 27.50 2,50 | 4.00 | 6.00 { 7.50 | 8.50 225 «| 3.7 | 575) 6.50, 7.25 2.50 | 4.25 | 6.00 | 7.25 | 8.00 Boo | Se | eraoe 6.75" | 7.50 200 |, sera | Gree 7.25 | 800 2075 | fe | Seo |) 6.50 | 7.25 B00 | 4025 | G00) |p2 7°25 || 8.25 2.75 | 400 | 5.50 | 6.50 | 7.25 2.25 | 4.00 | 6.00 | 7.00 | 8.25 2.00 | (875° | 5.25 | 6:25 |) 7.50 3:00 | 4.50 | 6.50 |" 7.50 | 8.50 2.50. | 4:00) | 5.50 |, 6:25 |. 7.00 3 apr) eeGr elt -eGsipe | 7eas. | S12 S67 eae | 5.57 |) 6/36: | 7:27 22 LEWIS R. CARY influence of its sense organs. The half disk with its sense organs, although pulsating much more slowly than its activated mate shows a considerably higher rate of regeneration than the latter. 9 Fig. 9 Curves to show the comparative rates of regeneration between-the halves of 40 disks from which all of the sense organs have been removed, while the subumbrells muscles of one-half of each disk are activated by a circuit wave of contraction. The upper (solid) line shows the regeneration of the activated half disks, while the lower (broken) line shows the regeneration of the half disks with- out sense organs or activation. The divisions along the ordinate represent the amount of regeneration in millimeters. Those along the abscissa the time of regeneration in days. That the influence of the sense organs upon the rate of regen- eration differs in no essential respect from that exercised upon the general metabolic activity of the entire disk is shown by comparing the amounts of CO, produced in a given time by half ame RATE OF REGENERATION IN CASSIOPEA 23 TABLE 4 Record of 40 specimens, one half of each one pulsating under the contro! of its sense organs; while a circuit wave of contraction was maintained in the sub- umbrella muscles of the other half (fig. 10) DAYS OF REGENERATION NUMBER OF SPECIMEN 1 2 3 4 5 1 its TU OMe anton ace 3.00 4,25 5S as 7.50 9.00 Activated half............. Do 3.25 4.50 6.00 7.50 9 le ‘oll ise (Oe oo denis ae eri: 3.50 5.50 220 9.00 Activated Nalfi.5..o5.<1.+.s 1.50 3.20 5.25 7.00 8.00 3 oe TALEO tS KO) tay eaols ABER 4.00 5.50 6.75 8.25 C Activated Nhalfis..5... 562. .+ 3.50 4.75 5.50 7.50 4 ear Nuatilst tal 0) See ene se 2.50 4.25 5.50 7.50 9.50 IActivacedshalte 3. \....c 2. 2.00 auto 4.75 6.75 8.00 5 fee WiltiimosiOeer 4 Saale 2.00 4.00 5.50 7.50 9.00 Activated) half. <.... 0. 2.c- 1.50 35: 4.50 6.00 7.50 6 ae Wilner Ose lornenta ty < 2.75 4.25 SE dele 9.25 AC HIVAteO malic. seme cs a2 2.00 3.50 4.75 6.50 coe 7 toe TALS HMO) ee 6 See 2.50 4.00 5.50 420 8.75 Nctiveted Naliis.c. see eel 2.00 3e25 4.75 6.25 vii 8 ne WAGON Ol. cht 225 4.25 Selo 7.50 9.00 Wetivabed: Half.) 5.c.'.0 50 ce 1.75 3.00 4,25 6.00 7.50 9 rae Wilh Oko sac e tae 3.00 4.50 DD he 9.25 Metivated Nalf-jnes. sc. ss: 2.50 205: 5.00 6.50 7.50 10 hes Wilh Omens. 1.50 3.00 5.00 7.00 9.00 ING Hivahe dell iteys act sishstels iss 1525 PAS 4.50 6.25 8.00 ee with: Sa Oeyaceces .. 2 74 AS 4.00 5.50 7.00 9.00 ‘area ali ee techs cis 3 1.75 3.25 4.75 6.25 7.50 12 ice Wil libisen Oleic sae syste ois 3.00 4.75 5.50 Ato 9.25 Activated half......... sone 7475, oe25 4.50 6.00 7.25 13 ha: Vat Cups ek Oe ele ee ieee 2.50 4.00 Dee e225 9.00 Activated hali-......44.6. 2 2.00 3.50 4.75 6.25 “20 14 ee: Vali] sets CO) ate te ee eae ATES 4.25 5D LUE 9.25 J NCUTIK EHe) 0 Ml Ob 2.25 3.50 4.50 6.50 TATE 15 res Wilts OMe eee ie, 3.00 4.25 5.75 8.00 9.25 Activatede Hal. ..% <5. =< 6. Zoos 328 4.75 6.25 7.50 16 fase Wii Nwoe On. cee ae. 4.00 6.00 to 9.50 C Activated haliic.s...5!-..- O25 Is: 7.00 8.25 7 fae UGE Oe ee picts 2.50 4.25 5.50 7.25 9.00 NCuIiVvAtTed Naliet ces cen. ts 2.00 See 4.75 6.50 tbo 18 ate WAG ONO seco se: 2.00 4.00 oto de2D 9.25 Activated half. .s ... i405 %.. 1.50 3.50 4.75 6.00 7.50 19 We ABUT oe \O)s Nem as See ae PD AES. 4.25 eo Ue 8.75 Activated half. 1. 4...2:<6. 2.00 3820 4.50 6.50 8.00 20 a: WALES eROe, see ciptete cect Deh 4.00 ae 7.50 9.00 Metivatedehaltiv.. .asccn.ck. 150 3.20 4.75 6.25 7.25 24 . LEWIS R. CARY TABLE 4—Continued DAYS OF REGENERATION NUMBER OF SPECIMEN 1 2 3 4 5 A ee “ou rces Cer nn 3.00 4.50 6.00 9.00 C Activatedubalften-2s...5 <:- 2.50 4.00 4.75 7.50 59 ee SGN 0 a 2.25 4.00 6.00 8.00 C Activated half............. 2.00 3.75 5.50 7.75 93 Le “abi, SEN Cae 2.75 4.25 6.00 7.50 9.00 Wetivated half)....:....-.. 2.25 3.50 5.00 6.25 7.50 oe te DASE Oe 2.50 4.00 5.75 7.75 9.25 Netiveited mealft.s5.0 esac 2.00 O25 4.75 6.25 a0 a5 tee BOMBS OO) Noches eee bets 1.25 3.50 5.50 7353 9.00 Netiveredehallinn ccc 26 ” 0.005 x 0.0023 per centimeter = 383.18 dynes per centimeter or the minimum tangential tension in the rim of the food-cup required to cleave the paramecium on the basis of the first of the two methods to be considered. But in cutting the paramecium with the glass fiber the maximum pressure was applied when the length of fiber in contact was several times reduced. It is therefore evident that the tangential tension required is greater than 383.18 dynes per centimeter. But in a cup, if the constriction of the rim is = 391 grams OBSERVATIONS ON FEEDING OF AMEBA 45 due to surface tension, it is only the difference between the sur- face tension at the rim and elsewhere that can be effective. If the paramecia then were actually cut by the contraction ofthe rim of a food-cup due to the action of surface tension, there must have been a difference in the surface tension of the rim of the cup and that of other regions of the ameba equivalent to more than 383.18 dynes per centimeter; and to make this difference possible the surface tension of ameba must have been much larger. If the cutting of paramecium by ameba occurred in accord with the second method mentioned above, that is, by the ap- proach of the distal ends of two pseudopods, as the bulk of the evidence seems to indicate, and if the cutting quality of the pseu- dopods was the same as that of the glass fiber, then the pressure exerted by each of the pseudopods must have been approximately equal to 9 mgm. Since the pseudopods and the glass fiber were nearly all of the same size their cutting quality was probably practically the same. We may, therefore, assume that the pres- sure exerted by them was, if the paramecia were cut in this way, 9 mgm. If this pressure was the result of the action of surface tension it must have been due to a reduction of the tension at the end of the pseudopods equal to 9 mgm., and the width of the surface involved must have been equal to the circumference of the pseudo- pods. These were approximately 0.025 mm. in diameter having a circumference of 0.078+ mm. The difference in surface ten- sion between the end and the base of the pseudopods must there- fore, if our postulates are valid, have been 9 mgm. or 8.72 dynes per 0.078+ mm., which equals 1118+ dynes per centimeter; and to produce this difference the surface tension of the ameba must have been much higher. Consequently to cut a paramecium in two in accord with this method a much higher surface tension is required than to cut it in accord with the first method. The process in question would, therefore, require, at the very least, a surface tension considerably higher than 383 dynes per centimeter and in all probability it would require a surface ten- sion higher than 1118 dynes per centimeter, that is, practically ‘46 S. O. MAST AND F. M.. ROOT as high as that of molten platmum, the highest yet discovered. But the surface tension of protoplasm is, according to Czapek as previously stated, only approximately 50 dynes per. centimeter and with a surface tension of this magnitude in ameba it is ques- tionable whether a greater local reduction than 25 dynes per centimeter could be produced without destroying the organisms. It is therefore evident that surface tension plays a very insig- nificant rdle in the process of feeding described unless the proto- plasm of these organisms consists of some sort of a structure that makes possible a great magnification of the effect of the surface tension. But since there is no evidence of such a structure the required power must, for the present, be sought largely in con- nection with other phenomena, gelation pressure, absorption pressure, adhesion, cohesion, diffusion, etc. There is as yet little or no experimental evidence which di- rectly bears upon the relative importance of these different fac- tors or upon the mechanics of their regulation, although it is clear that the reactions in Ameba do not depend solely upon changes in the environment. Movement and changes in move- ment (responses) may undoubtedly occur without any affective external changes, such responses being entirely due to internal processes. It is also obvious that while some of the responses which are dependent upon external conditions are directly re- lated to the environment, being local responses to local stimula- tion, others are not; and we are unable to conceive how some of the latter can be explained without assuming that the entire animals are involved as organized systems of considerable com- plexity, that there are impulses transmitted from one part of the body to another and that there is a regulatory center, in which impulses may originate and in which those originating elsewhere may be modified and controlled. This is especially true regard- ing much that occurs in the process of feeding on rotifers. It is also true regarding the peculiar mushroom-shape assumed when feeding on infusoria and regarding a considerable number of responses described by Jennings (’04), Kepner and Taliaferro (13) and others. Moreover, the facts that enucleated parts of ameba do not respond at all or respond in a haphazard fashion, OBSERVATIONS ON FEEDING OF AMEBA 47 indicate, as Hofer (’90) concludes, that the nucleus acts as a regulatory center. This conclusion is also supported by the work of Mr. H. S. Willis carried on in our laboratory and now in press in the Biological Bulletin. Gruber (12) and others, however, oppose this construction, although they obtained results similar to those obtained by Hofer and Willis. The conclusion reached above regarding surface tension is in harmony with that reached by Jennings (’04) in his observations on the rolling movement in ameba and by Dellinger (’06) in his observations on the walking movement. It is also supported by Kepner and Taliaferro ’(13) in their interesting observations on the process of feeding. SUMMARY 1. Certain amebae at times feed almost exclusively on rotifers, at others they feed largely on paramecia. 2. They capture the rotifers by flowing around the foot at mile point of attachment to the substratum. After they have sur- rounded the foot they begin to flow out over the body. The roti- fer responds by contracting and forcing the ameba back, after which it extends again and the ameba again begins to flow out over it, etc. In the meantime the foot begins to digest and grad- ually the rotifer weakens. , Thus they continue sometimes for days before the rotifer is swallowed. 3. When amebae are feeding on paramecia they assume a sort of mushroom shape with a serrate edge consisting of numerous short pseudopods. The paramecia tend to come to rest between and under these pseudopods by which they are usually sur- rounded, but sometimes the ends of the pseudopods approach each other before they are fully extended and cut the parame- cium in two. 4. To cut a paramecium in two with a fine glass fiber it re- quires a pressure of approximately 9 mgm. If the pseudopods have the same cutting quality as the glass fiber and if their move- ment is due to a change of surface tension, it requires to perform the work involved, a reduction in surface tension of at least 1118 dynes per centimeter at the tips of the pseudopods. 48 S. O. MAST AND F. M. ROOT 5. If the ends of the pseudopods fuse so as to take on the form of a ring around the paramecium and if the cutting is due to con- striction in this ring, and if the constriction in this ring is due to a change in surface tension, the work involved requires a minimum reduction along the inner surface of the ring of at least 383 dynes per centimeter. 6. The bulk of evidence at hand seems to indicate that the paramecia are divided by the approach of two pseudopods. To account for the process on the basis of the surface tension theory therefore, the surface tension of the amebae would have to be, at the very least, much higher than 383 dynes per centimeter and in all probability considerably higher than 1118 dynes per centimeter. The surface tension of protoplasm is, however, only approximately 50 dynes per centimeter. It is, therefore, prob- ably at best an insignificant factor in the process of feeding in ameba. OBSERVATIONS ON FEEDING OF AMEBA 49 LITERATURE CITED BiocuMaNNn, F. 1894 Kleine Mitteilungen iiber Protozoen. Biol. Centralb. Bd. 14, 8. 82-91. Burscutr, O. 1892 Untersuchungen iiber mie -opische Schiume und das Protoplasma. Leipzig, 234 8. CzapEK, FREDERICK 1911 Chemical phenomena in life. London and New York, 148 pp. DELLINGER, O. P. 1906 Locomotion of amoebae and allied forms. Jour. Exp. Zo6l., vol. 3, p. 337-357. GREENWOOD, M. 1887 On the digestive process in some Rhizopods. Jour. Phys., vol. 8, pp. 263-287. GRUBER, K. 1912 Biologische und Experimentelle Untersuchungen an Amoeba proteus. Arch f. Protistenkunde. Bd. 25, S. 316-376. Horer, Bruno 1890 Experimentelle Untersuchungen iiber den Einfluss des Kerns auf das Protoplasma. Jen. Zeitschr. f. Naturw., Bd. 24, S. 105-176. JENNINGS, H. S. 1904 Contributions to the study of the behavior of lower organisms. Carnegie Inst. of Washington, Pub. No. 16, 256 pp. JENSEN, P. 1905 Zur Theorie der Protoplasmabewegung und iiber die Auf- fassung des Protoplasmas als chemisches System. Anatom. Hefte (Merkel u. Bonnet), Bd. 27, Heft, 83, S. 831-858. Kepner, Wn. A. AND TALIAFERRO, W. H. 1913 Reactions of Amoeba proteus tofood. Biol. Bull., vol. 24, pp. 411-428. McCienpon, J. F. 1912 The osmotic and surface tension phenomena of liv- ing elements and their physiological significance. Biol. Bull., vol. 22, pp. 113-204. RuuMBLER, L. 1905 Zur Theorie der Oberflachenkrafte der Amében. Zeit. f. wiss. Zool., Bd. 83, S. 1-52. 1910 Die wemchicdenaviieen Nahrungsaufnahmen bei Amdben als Folge verschiedener Colloidalzustinde ihrer Oberflichen. Arch. f. Entw.-Mech., Bd. 30, S. 194-220. Ryper, J. A. 1893 Dynamics in evolution. Biological Lectures, Woods Hole, pp. 63-81. Verworn, Max 1909 Allgemeine Physiologie. Fiinfte Auflage, Jena, 742 8S. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 21, No. 1 Wh, a oa ' U ra Pie! iva 7 Pr ws ig rg pe leat | a P vO ’ eh 'y i me \ ‘ TA. i) : alae 1. ae THE EFFECT OF MOISTURE UPON THE SILK OF THE HYBRID PHILOSAMIA (ATTACUS) RICINI BOISD. «# xX PHILOSAMIA CYNTHIA (DRURY) ¢ ONERA A. MERRITT HAWKES Philosamia ricini and of P. cynthia from Ning-po were crossed in 1914 in order to study the method of inheritance of certain spots on the larva, the colour of the cocoons, and the ar- rangement of the long white hairs on the abdomen of the moth. The results of the preliminary work on cocoon colour are given in this paper, no reference being made to the still unfinished work on the larva and moth. When the cocoon colour work'was undertaken, it was hoped that the results would help to explain the lack of coincidence be- tween the work of Kellogg and Toyama on the inheritance of colour in cocoons. The breeding, thus far, has thrown no light upon the method of inheritance, but, if the cocoons used by them, are as much affected by moisture as the hybrid here discussed, their results may have to be considerably modified. Philosamia ricini is found wild in Assam, but has also been domesticated to a considerable extent in that province, and to a smaller degree, in other parts of India. Ia... stay Ree oo keus {le 133 Factors suggested by Whitney’s experiments..................2...00--00% 137 Bitechrondulmtesooulllonbe. a. 7 eee ec oan s sree ee, byes 139 Hiteciotmeraoolcaproaucis Omasereenalea...... aa ee cee 140 LN ER ROG te) Sa Soak AE Ste ee ee 2 143 WR SetSMOR st te Mere Matec, seh e ys dete sro P 2 OE rs. ce Mel Ae 156 SUELSSDGINY 2 A IED Gan: Oe Rao Ee EERO ak Rete rg wore te 160 Petis lip iia (sli ny eee Oy ok eS Ae NA 3 oan as eo Oe eee 161 INTRODUCTION After it had been well established that certain agents reduce the number of male-producing females in the families of the rotifer Hydatina senta, two important related lines of investi- gation were undertaken. One sought to explain how these agents operated to prevent the appearance of male-producers; the other had for its object the discovery of means of increasing the num- ber of male-producers. The former line of research has proven nearly fruitless; the latter has met with some success. The experiments described in this paper were aimed at the solution of these two sets of problems. 1Contributions from the Zodlogical Laboratory of the University of Michigan. ; . 127 128 A. FRANKLIN SHULL AND SONIA LADOFF EXPERIMENTS ON THE MECHANISM OF THE PREVENTION OF MALE-PRODUCTION Osmotic pressure The list of substances which effect a reduction in the number of male-producing females includes ammonium salts, sodium hydroxide, beef extract, manure solution, creatin, urea, and some others. In substances of such widely different properties it is difficult to select a common feature to which their com- mon effect upon the life cycle of the rotifers could be attributed. One possibility that early suggested itself was that the osmotic pressure of the solutions produced the effect observed. Were this the only cause of the reduction in the number of male-pro- ducers, it would be expected that those solutions whose osmotic pressure was the highest would reduce male-production the most. Unfortunately it is probably quite impossible to say whether a solution produces osmotic effects in living protoplasm unless that protoplasm changes volume. Tables of osmotic pressures are therefore of little value in determining whether, in the ex- periments referred to, the reduction in the number of male- producers was proportional to the osmotic effect of the agent employed. As methods of detecting change of volume of the living rotifers or their tissues were found impracticable, it was necessary to resort to conjecture. In the experiment first to be described, a substance was selected which, if the tissues of the rotifers behaved as theoretically perfect semipermeable membranes, would give an osmotic pressure considerably higher than that which obtained in the other solutions used. The sub- stance selected was cane sugar. Experiment 1. Two sisters isolated January 13, 1911, became the parents of the two lines of this experiment, one of which was reared continuously in sugar solution, the other in distilled water. Food and other conditions were the same for both lines. A ™ solution of cane sugar was kept in stock. It was heated daily to prevent fer- mentation, and was tested at intervals for inversion to reducing sugar. Once when Fehling’s solution was reduced, the stock solution of sugar was rejected and a new one prepared. This stock solution was diluted to ™. for use in the experiments. MALE-PRODUCTION IN HYDATINA 129 The effect of the sugar solution, as shown in table 1, was to reduce the number of male-producers. But the amount of reduction was less than was expected on the theory that osmotic pressure was responsible for that reduction. It was less than the effects of certain ammonium salts whose effects should have been, theoretically, less than those of the sugar. TABLE 1 Two lines of the rotifer Hydatina senta derived from sisters, one line reared in dis- tilled water, the other in a ® solution of cane sugar. The number of male-pro- ducing (72) and female-producing (2 2) females is recorded for each line. The cane sugar reduces the number of male-producers DISTILLED WATER = CANE SUGAR SOLUTION Number Daan: Number | Number | Number Date ofthear Number | Number eae young ae ae eae YOURS go ae Jan. Jan. 1 15 3 20 1 15 2 16 2 17 4 16 2 MZ 1 22 3 19 3 17 iy 0 9 4 21 4 PAL Si” 19 3 33 5 22 5 29 19 je 25 6 24 2 26 4 21 4 20 Tf 26 7 37 21 6 12 8 27 26 5 5 2o% a 27 9 29 6 18 23 2 10 10 31 17 34 6 24 0 13 Feb. 25 1 21 11 2 23 31 7 26 3 15 . PA 9 14 8 29 2 29 29 5 23 9 31 4 14 31 7 19 Feb. 10 2 9 14 2 1 10 TUG | ae oe Sh tie ll 100 254 67 346 Percentage of 7 @.. 28 .2 16.2 THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 21, No. 1 130 A. FRANKLIN SHULL AND SONIA LADOFF Acidity In one of Shull’s (’11) earlier experiments it was found that sodium hydroxide reduced the proportion of male-producers to a slight extent. At the same time experiments with acids were performed in the hope of obtaining the opposite effect, but it was found impossible to rear the rotifers in even a very dilute solution of the inorganic acids used. Although weak solutions of hydrochloric acid were used, by the time the slightly alkaline food was added, the solution appeared neutral. When he used solutions of the acid strong enough to remain acid after the food was added, the rotifers died. With the expectation that organic acids might be less deleteri- ous, the following experiment with butyric acid was begun. Experiment 2. A 1 per cent stock so'ution of butyric acid was kept in a glass stoppered bottle, and diluted for use to 0.03 per cent (the diluent being Great Bear spring water). Two lines of rotifers derived from two sisters isolated September 10, 1911, were reared, one in spring water, the other in butyric acid solution. For want of a satisfactory indicator, it was not known that the latter solution remained acid after the food was added. The characteristic odor of butyric acid remained, but it was to be expected that the butyrates would possess the same odor. Contrary to our hopes, the acid reduced? the proportion of male-producers, as shown in table 2. Delay or inhibition by manure solution The non-occurrence of male-producers while the rotifers were being reared in strong manure solution might be attributed to delay, rather than to inhibition. So long as a line were reared continuously in manure solution the delay would be continuous; but if only one or two generations were reared in such a solu- tion, might not male-production which could not occur in these 2 It may be pointed out that the families are larger in butyric acid (mean size, 47.1 daughters) than in spring water (mean size, 38.7 daughters). According to Mitchell’s conclusions (’13) based on Asplanchna but applied to rotifiers in gen- eral, the higher nutrition evidenced by larger families should have been accom- panied by greater male-production, instead of less. MALE-PRODUCTION IN HYDATINA 13h TABLE 2 These two lines of Hydatina senta, derived from sisters, one line reared in spring water, the other in dilute butyric acid, show that butyric acid reduces the propor- tion of male-producers. It also increases the size of family SPRING WATER 0.03 PER CENT BUTYRIC ACID Number Rateot Number | Number | Number Dia of Number | Number f f f f pe naratian first young 3? 2 o Q ponbration first young o 2 ° 2 Sept. Sept. 1 12 0) 46 1 12 3 51 2 14 4 19 2 15 0 42 - 3 16 6 23 3 Li 0 52 4 iN%/ 0 26 4 18 0 45 5 19 0 28 5 20 2 46 6 20 7 45 6 22 1 45 a 22 11 44 a 23 4 39 8 23 @ 44 eNO tals cess tials 35 275 10 320 Percentage of 72 .. 11.2 3.0 Average size of ANY TM livery ec eco) eissetce 38.7 47.1 generations occur in greater degree in the generations immediately following them, which were reared in spring water? The possi- | bility was tested in the following experiment. Experiment 8. The general plan of the experiment was to rear, in spring water, females whose grandmothers had been kept in manure solution, and whose mothers were reared (except for a few hours after hatching) in spring water. The details of the method were as follows: Six or eight daughters of one female were divided into two lots of three or four each, one being put into spring water, the other into manure solution. Eight to 24 hours later these lots were transferred to new dishes, one in spring water as before, the other in manure solution. The transfer was made in order that the first young to hatch in the second dish of manure solution might be known to have hatched from eggs that underwent their maturation and part of their previous de- velopment in manure solution. These first young, if females, were transferred to spring water, in which (if they proved to be female- producers) they were kept the rest of their lives. Their entire families weré reared to maturity and recorded. The totals for each lot of families simultaneously reared are shown in table 3. 132 A. FRANKLIN SHULL AND SONIA LADOFF TABLE 3 Records of male-producers and female-producers among females whose grandmothers were reared in manure solution, and females whose grandmothers were reared in spring water. All individuals other than these grandmothers, of all generations, and in both halves of the experiment, were reared in spring water. There is no accumulation of male-producers among granddaughters of females reared in manure solution GRANDDAUGHTERS OF FEMALES GRANDDAUGHTERS OF FEMALES REARED IN SPRING WATER REARED IN MANURE SOLUTION LOT NUMBER LOD NUMBER’. |) saa eee Number of Number of Number of Number of oe Qe of 2 2) A 9 4 A 32 37 B 4 51 B 1 38 C 2 40 C 4 70 D 33 4 D 30 37 E 9 13 E 19 19 F 28 16 F 15 83 otal eccns 85 208 101 284 Percentage of cg? Gm. 29.0 26.2 There is no increase in the proportion cf male-producers among the granddaughters of females bred in manure solution. The figures show an actual decrease, but it is so small as to be ‘probably insignificant. The manure solution does not merely delay, but inhibits, male-production. EXPERIMENTS DESIGNED TO INCREASE MALE-PRODUCTION In an attempt to increase male-production artificially a num- ber of salts were tested, some of them being selected because of their well-known physiological effects in no way connected with sex, others selected purely at random. Only those in which the rotifers could be easily reared are reported in this paper. While this work was in progress Whitney’s (14) paper on nutri- tion experiments appeared. Our energy was then directed to- ward testing the agents which seemed to us to enter into Whit- ney’s experiments but which he had apparently neglected. These two lines of work are discussed separately. MALE-PRODUCTION IN HYDATINA 133 EFFECT OF VARIOUS SALTS Experiment 4. Calcium chloride. Two sister individuals of a line obtained in Nebraska were isolated on November 20 and 21, respectively, 1912. One, with its progeny, was reared in spring water, the other, with its progeny in ,*, solution of calcium chloride. Other conditions were the same in both lines. As shown in table 4, the only male- producers in this experiment appeared in the calcium chloride solution. TABLE 4 The two lines of rotifers in this table were reared, one in spring water, the other in a dilute solution of calcium chloride. The only male-producers (7 Q) in all the families were produced in calcium chloride SPRING WATER = CALCIUM CHLORIDE spamiber Dateor Number Nusrber Humber Datei Number | Number San ceationl first young, oe °) Pi io) generation first young g 2 Pe io) N ov : Nov. 1 22 0 25 1 24 oe) 31 2 24 0 39 2 26 0 42 3 26 0 26 3 28 0 45 4 29 0 24 4 30 2 40 Dec. Dec. 5 2 0 26 5 2 0 23 6 4 0 42 6 4 0 47 a 5 0 44 Uh 8 0 23 8 a 0 24 8 10 0 12, 9 9 0 22 9 13 0 45 10 12 0 il 10 15 0 Pa 14 0 9 11 15 0 50 Motaleeert eee see 0 So 2 329 Percentage of 7 @ .. 0.0 0.6 Experiment 5. More dilute calcium chloride. The above experiment was repeated, using a more dilute solution of calcium chloride. A {*, solution was used in the first three generations, ;%, thereafter. In table 5, which records the details, it appears that three times as many male-producers were reared in calcium chloride as in spring water. Experiment 6. Calcium chloride used intermittently. In this experi- ment, only one line was reared, beginning Jan. 11, 1913. Instead 3 The results of these experiments with calcium chloride have been published in summary form elsewhere (Shull, ’13) but appear here for the first time in detail. 134 A. FRANKLIN SHULL AND SONIA LADOFF TABLE 5 One of the two lines here recorded was reared in spring water, the other in a solution of calcium chloride more dilute than that used in table 4. Again there were more male-producers in the calcium chloride N N O =, CALCIUM CHLORIDE SPRING WATER 300 7° G00 Number D Number | Number | Number D f Number | Number f ate of f f f. ate o f Senate first young eo Q ° Q pentration first young eo Q ° Q Jan. Jan. 1 13 2 40 1 13 0 38 2, 15 0 40 2 15 0 15 3 17 0 24 5 IV 0 1 4 18 0 35, 4 18 0 29 5 20 0 49 5 20 9 31 6 22 5 35 6 22 if 43 if 24 0 B13 ah 24 ° 0 40 8 26 0 29 8 26 4 44 9 28 0* Ths 9 28 OF 1b “DG YP Sec A a a 7 292 20 252 Percentage of J @ .. 2.3 7.3 * Remainder of family not recorded. of rearing it continuously in one medium, alternate generations were reared in spring water and a ,%. solution of calcium chloride. All members of one family were reared, from the time they were isolated shortly after hatching, in the same medium. That member of each family which was selected to become the parent of the next generation laid her eggs in the medium in which she was bred, and her eggs hatched there. But shortly after hatching, her daughters were transferred to the alternate medium, and there reached maturity. Most of the male-producers appearing in this experiment, | as shown in table 6, were offspring of females that lived and laid their eggs in calcium chloride solution. Since itis the growth and maturation period of an egg in which it is determined whether a male-producer or a female-producer will develop from that egg, as shown by the earlier experiments of Shull (’12), the results given in table 6 are what would be expected if calcium chloride favors male-production. In this respect Experiment 6 is in _accord with the other calcium chloride experiments above. MALE-PRODUCTION IN HYDATINA 135 TABLE 6 In the single line here recorded the first generation was reared in spring water (from a parent reared in ~N, calcium chloride); the second generation was reared in a calcium chloride solution (from a pareit reared of course in spring water). Most of the male-producers are daughters of parents reared in calcium chloride OFFSPRING OF PARENTS OFFSPRING OF PARENTS REARED IN SPRING WATER REARED IN CALCIUM CHLORIDE NUMBER CF DATE OF GENERATION FIRST YOUNG oe oe a ee : Number of Number of Number of Number of ‘op! 19) Se) i) <9) ie 12) Jan. 1 13 0 26 2 15 0 14 3 ily/ 0 25 4 19 1 15 5 PALL 4 4} 6 23 0 23 7 25 6 oo 8 27 0 Defy “(Dai | Foes Ole Geleda ee pares cua 1 79 10 124 | Percentage of op'@ ........ thee 7.4 Experiment 7. Calcium chloride and a New Jersey line of rotifers. The effect of continuous rearing in calcium chloride of two concentra- tions was tested in a line of rotifers from New Jersey, to check the results in the three preceding experiments, in all of which the Nebrasla line was used. The method of conducting the experiment was the same, except that different concentrations were used, but, as may be seen from table 7, the results were different. There is a reduction in the number of male-producers in calcium chloride in this line. We are unable to explain the difference between the Nebraska and the New Jersey line in this respect, though inherited physio- logical differences of other kinds are well known. The constancy of the results obtained with calcium chloride in the Nebraska line, even if the differences produced were small, lead us to attribute these effects with some confidence to the action of the calcium chloride. Experiment 8. Magnesium chloride. The physiological effect of magnesium upon other organisms is In some ways in strong contrast to the effect of calcium. We hoped to find this difference extended to their effect upon the life cycle of Hydatina. A =, solution of magnesium chloride was used for one line, spring water in another 136 A. FRANKLIN SHULL AND SONIA LADOFF TABLE 7 Three lines of rotifers of a stock obtained in New Jersey were started from two sis- ters. One line was reared in spring water, the other two in two concentrations of calcium chloride. There are fewer male-producers in the calcium chloride solutions SPRING WATER 0 CALCIUM CHLORIDE =x CALCIUM CHLORIDE Numbervot Num= |) Num=t i mber of Num- | Num- Number of um- | Num- Numberot | ber | barof | Numberof | ber of | berot | Number of | bora | ber of 1 0 4 il 0 23 1 1 30 2 3 38 2 6 31 2 0 31 3 17 Pa 3 0 38 3 0 34 4 0 27 4 4 14 4 2 31 5 0 27 5 0 15 5 0* Gy 6 Oz ae 6 0* 10* 6 OF 13* ies 0* 14* a 0* Ile 7 0* 8* 8 0 42 8 0 36 8 0 26 9 @ 29 9 0 3 9 0 2 10 0 2 0 4 0 3 10 0 9 Mo tales. case : 20 224 10 194 3 190 Percentage of GOES ae 8.1 4.8 5 *Remainder of family not recorded. line. Although both lines were reared for ten or more generations and over three hundred individuals were obtained in each line, not a single male-producer appeared in either line. It seems unnecessary to state such uniform results in detail, hence no table for this experiment is published. There is no way of interpreting the results of this experiment. Magnesium chloride may reduce male-production. Or it may increase male-production, but could not overcome a strong ‘tendency’ to female-production in the line used. The experi- ment has neverbeen repeated with a line in which male-producers appeared more frequently. Experiments 9, 10, and 11. Potassium sulfate, tron chloride, and ammonium chloride. The success obtained by using very dilute cal- cium chloride suggested that very dilute solutions of other substances might also increase male-production, even in the case of substances, MALE-PRODUCTION IN HYDATINA 137 more concentrated solutions of which were known to have the reverse effect. Among the substances tested were the three named above. Since all of them reduced male-production or showed no effect, the results of the several experiments are for brevity summarized in a single table (table 8). : TABLE 8 Summaries of several experiments with dilute solutions of potassium sulphate, iron chloride, and ammonium chloride. The number of male-producers is not increased by any of these substances, but in most cases is reduced EXPERIMENT CONTROL Num-|} Num-| Per Substance centra- ; genera- 9 9 a genera- 9 29 Potassium sulphate a 8 19 343 | 5.2 8 31 326 | 8.6 Potassium sulphate a5 8 28 269 | 9.4 8 31 326 {| 8.6 Iron chloride....... 007% 5 8 158 | 4.8 5 43 157 | 21.5 Ammonium 1 ghiorides. 2.055: 7000 9 0 240’ 0:0°) 40 3 356 | 0.8 FACTORS SUGGESTED BY WHITNEY’S EXPERIMENTS Whitney (’14) reared Hydatina on a diet of Chlamydomonas; and obtained a striking increase in the proportion of male-pro- ducers. He freely attributed this effect to the (qualitative) difference in nutrition, without mentioning several agents whose possible effects had not been eliminated. For example, his food cultures were started in different media; Chlamydomonas was reared in a solution of bouillon cubes, the Polytoma used in the control was reared in manure solution. Although only small quantities of the liquids of the food cultures were introduced with the food, nevertheless there must have been some initial differences of a chemical nature between the experiment and its control. The possible effects of these substances were not tested, nor were they mentioned in Whitney’s paper. Further- more, it was probable that the organisms in the food cultures 138 A. FRANKLIN SHULL AND SONIA LADOFF produced, as a result of their metabolic activities, differences between the food cultures that did not exist at first. Thus, it might be expected that Chlamydomonas, a green organism, would carry on photosynthesis, with the resultant liberation of oxygen. It seemed not improbable that oxygen might produce part of the effects in the experiments with Chlamydomonas. This was the more plausible since, as Whitney states, marked effects were produced only when large numbers of Chlamydo- monas were present, and only when the dishes were kept in direct sunlight. Both of the conditions named should result in the production of relatively large quantities of oxygen in solution. If it should appear that any or all of these agents present in the Chalmydomonas cultures have any considerablé effect in increasing male-production, Whitney’s conclusion that his experi- ments gave evidence of an effect of nutrition would lose much — of its support. If, on the other hand, these obvious factors could be shown to have no effect whatever, the belief that nutri- tion effected the increase of male-production noted would be greatly strengthened. It may be stated in advance that our results may be inter- preted as being largely in support of Whitney’s contention; for, while one of the suspected agents (oxygen) gave positive re- sults, its effect was much smaller than that which Whitney reported. Our experiments are the more significant because we have worked with a line of rotifers kindly sent to us by Dr. Whitney in January, 1915, from the line used in his own experi- ments. We can only regret that our experiments were not con- ducted by Dr. Whitney himself, as we have had some difficulty in duplicating his conditions merely from his published accounts. In the account which follows, we separate the experiments aimed to test the effect of dilute bouillon from those designed to test the effect of oxygen. MALE-PRODUCTION IN HYDATINA 139 Effect of dilute bouillon Experiment 12. The bouillon solution used in this experiment was made from one cube of Armour’s beef bouillon in 21,000 ce. of water. As nearly as we could compute from Whitney’s paper, this was the concentration of bouillon present.in his cultures after the rotifers had been fed with Chlamydomonas. ‘Two lines of rotifers, derived from sisters, were reared, one in spring water, the other in this dilute bouillon solution. Table 9 shows a slight reduction in the number of male-producers in bouillon, rather than an increase. f TABLE 9 Showing the effects of continuous rearing of Hydatina in a dilute bouillon solution. The number of male-producers (7 9 ) is slightly decreased SPRING WATER BOUILLON Number | Number of Number of Number Number of Number of of generation a9 29 of generation rot a 1 0 15 1 0 11 2 6 17 2 3 32 3 2 19 z 0 3 4 0 al7/ 4 0 l7/ 5 0 29 5 0 8 6 1 24. 6 1 29 if 0 14 ii 0 20 8 1 25 Motale: 22% 9 135 55 145 Percentage 1 ite ae 6.2 3.3 Experiment 13. Bouillon was tested in this experiment by rearing only the parents in bouillon, not their offspring. Thiswas Whitney’s method in working with Chlamydomonas, and was based on Maupas’s statement (1891), subsequently confirmed by Shull (’12), that the de- termination of the sex of an individual takes place in the body of its grandmother. The concentration of the bouillon was made less than that in Experiment 12, being one cube in quantities of water varying, on different days, from 30,000 to 80,000 cc. Two lots of females of approximately equal size and from the same source were set aside on the same day, the one in spring water, the other in bouillon solution. The parents were removed after 24 hours, The offspring from all eggs laid in that time were reared to maturity in spring water, and recorded. In table 10 the nature of these offspring is summarized. Here again there is a slight reduction in the number of male-producers in bouillon solution. The consistent results of these two experiments seemed to make it not worth while to look for a male-producing factor in the dilute bouillon solution used by Whitney. 140 A. FRANKLIN SHULL AND SONIA LADOFF TABLE 10 Showing the effects, on the offspring, of rearing the parents in dilute bouillon. The eggs from which the offspring hatched were laid and hatched in bouillon in one- half of the experiment, those of the other half laid and hatched in spring water. In both cases the offspring were removed to spring water a few hours after hatching PARENTS REARED IN SPRING WATER PARENTS REARED IN BOUILLON Na Number | Daughters proved to be] Number | Daughters proved to be aa f Ro ee parents 32 29 parenta Fre) 29 Jeimibiny 2licoa dep ous 3 2 10 4 0 11 RXUoe aaeduat 4 0 59 4 2 42 Rebnuanyan one cece: 3 0 39 4 0 65 Qe ace shaih eet 3 0 28 3 0 27 I ergeieen saree 3 18 20 3 9 38 Dati eye ste 3 0 54 3 0 52 March ee icy atae 2 3 ve 3 0 8 PROG aa ena d ine 21 23 217 24 11 243 Percentage of 1 @. 9.5 4.3 Effect of metabolic products of a green alga In the hope of approximating the conditions of Whitney’s experiments except the nutritive conditions, we reared the roti- fers in water containing a green alga that the animals could not eat. Spirogyra was selected, though it was not known whether the metabolic products of Spirogyra are similar to those of Chlamydomonas or not. Experiment 14. In this experiment, one of two lines derived from sisters was reared in water in which Spirogyra was kept, the other in water without Spirogyra. The same food was used for both. The Spirogyra was obtained from a spring, was kept in dishes in direct sunlight in the laboratory, and was washed out several times in Great Bear spring water before using, to prevent the introduction of foreign water in one part of the experiment. In the Spirogyra line only the parents and the first three or four daughters were kept in dishes with Spirogyra. The young females were removed daily to Great Bear water, in which they were reared to maturity. All dishes containing the parents, whether with Spirogyra or not, were in direct sunlight a part of the day. The temperature was not high when these experi- ments were performed (late winter and early spring), hence it was not necessary to take any precautions to reduce the temperature of dishes MALE-PRODUCTION IN HYDATINA 141 TABLE 11 Showing the effect of rearing rotifers in the presence of Spirogyra. In one line the parents of each generation were reared in dishes containing Spirogyra; in the other line there was no Spirogyra in the parents’ dishes. All offspring in both lines were reared in the absence of the alga PARENTS REARED WITHOUT SPIROGYRA PARENTS REARED WITH SPIROGYRA Number Number of Number of ce Number Number of Number of of generation oh {e) 92 of generation ae? kopte) 1 8 42 1 14 41 2 11 42 2 ibZ, 28 3 1 38 3 2 18 4 2 18 4 7 43 5 3 22 5 6 36 Aoitallieayys tae 25 162 41 166 Percentage Oiie OLS aes 13.3 19.8 set in the sunlight. Bubbles, presumably of oxygen, were usually present in the Spirogyra dishes, indicating that Phe yaeiieas was taking place. As indicated in table 11, there is a noticeable increase in the number of male-producers in the Spirogyra cultures. Experiment 15. The preceding experiment was repeated, this time with two lines in Spirogyra cultures, and but one control. The meth- ods were the same as described for Experiment 14. Table 12 gives the results. In these lines there is practically no difference between the Spirogyra cultures and the control. Experiment 16. Adopting for this experiment Whitney’s method of rearing only the parents under experimental conditions, we used the following procedure. Two or three mature, egg-laying females, members of the same family, were placed in each of two dishes. In one was placed some Spirogyra, in the other none. Both dishes were put into a covered dish, which was floated on the water in a small aquarium, and kept in the sunlight at a south window during the major part of each day. A thermometer was put into the floating dish, and the temperature (at this time often quite high out of doors) was found to vary only two or three degrees during the day. After 24 hours the parents were removed. The eggs laid during this 24 hour period hatched where they were laid, and the young were removed to spring water to grow to maturity. These offspring are recorded in table 13. In this experiment there is a slight reduction in the number of male-producers, though it is so small as to be probably insignificant. 142 A. FRANKLIN SHULL AND SONIA LADOFF TABLE 12 Showing the effects of continuous rearing of the rotifers in water containing Spi- rogyra, asin Table 11. Two lines were reared in Spirogyra cultures, with a single control WITHOUT SPIROGYRA WITH SPIROGYRA WITH SPIROGYRA Num- | Num- Number Num- | Num- Number Num- | Num- Number of generation | ber of | ber of of ber of | ber of cf ber of | ber of ae? 29 generation oped O79 generation oune) folte} 1 1 10 1 0 8 1 0 32 B 1 30 Z 1 19 2 8 16 3 2 10 3 0) 33 3 0 16 4 3 32 4 0 19 4 1 23 5) 1 9 5 3 21 5) 4 on 6 3 42 6 2 20 6 0 19 u 0 13 of 0 13 ia 0 3 8 0 10 9 6 3¥/ Robart os 11 146 12 180 13 160 Percentage of ets cae 7.0 622 Cao TABLE 13 Recording the offspring of parents reared in Spirogyra cultures, as contrasted with those of parents reared in the absence of Spirogyra. All offspring were reared in spring water. Spirogyra used in this way does not increase male-production. See also tables 11 and 12 WITHOUT SPIROGYRA WITH SPIROGYRA Daughters Daughters Number of oe EE = Nomber.ofy jl ee ee parents Number of Number of parents Number of Number of fof’) 22 oh’) erie) 3 3 15 2 0 14 3 5 18 3 2 22 2 2) 40 2, 7 35 z 7 28 74 4 22 2 1 19 2 1 18 Wotalermey >... ; 18 120 14 111 oN Urabe o-cte 13.0 . 11.2 MALE-PRODUCTION IN HYDATINA 143 On the basis of the Spirogyra experiments as a whole, we are ~ unable to draw a definite conclusion. While by continuous rearing with Spirogyra an increase in the number of male- producers was one time obtained, at another time the increase was practically zero; while when only the parents were reared in Spirogyra cultures, the offspring were slightly less frequently male-producers than in the control. Effect of oxygen The effect of oxygen was directly tested in several experiments. The method, in general, was to put the rotifers into a dish of water previously saturated with a mixture of air and oxygen containing a known proportion of oxygen, then set the dish in an atmosphere containing the same proportion of oxygen. The two atmospheres used were composed, respectively, of 40 per cent and 60 per cent oxygen. The oxygen was obtained by heating potassium chlorate with manganese dioxide and col- lecting the gas in a gas holder over water. No attempt was made to purify the oxygen. The 40 per cent oxygen atmosphere for saturating the water was obtained by lowering a graduated tube, fitted at one end with a stop-cock and open at the other end, into a vessel of water, admitting first three volumes of air, and then one volume of oxygen frcm the gas holder. The mixture was subsequently driven into a flask of spring water inverted over water, and later shaken repeatedly with the water remaining in the flask. When the rotifers were put into a dish of this oxygenated water, the dish was set under a bell jar made air-tight at the bottom with a vaseline-coated rubber gasket. .By means of a filter pump, one-fourth of the air under the bell jar was removed, as indi- cated by a mercury manometer in connection with the bell jar. Oxygen from the gas holder was then admitted until the mercury fell to zero. The partial exhaustion and refilling of the bell jar occupied usually less than two minutes. The 60 per cent oxygen atmosphere was obtained in the same way, except that equal parts of air and oxygen were mixed. 144 A. FRANKLIN SHULL AND SONIA LADOFF Experiment 17. Continuous rearing in 40 per cent oxygen. Two sisters were isolated March 31, 1915, one in spring water, the other in similar water first saturated with an atmosphere composed of 40 per cent of oxygen. The former dish was kept in air, the latter under a bell jar in an atmosphere containing 40 per cent of oxygen. From each parent was reared a line under the same conditions as those in which the parent female was placed. In the oxygen line, only the parents and several additional members of each family were kept in oxygenated water. Four or five of the first daughters of each family were kept in oxygenated water, and from among them the parent of the next. generation was selected. The remaining daughters in each family were transferred to spring water, where they reached maturity. Whatever effect the oxygen has, therefore, must be exerted in early larval stages, or in the egg, or in the body of the mother. From the previous experiments of Shull (712) it is to be expected that the egg or obgonial stages are the only ones in which the life cycle can be altered. TABLE 14 Recording two lines derived from sisters, one reared in untreated spring water, the other in water which was first saturated with an atmosphere of which 40 per cent was oxygen, then placed in a corresponding atmosphere. The oxygen line yields more male-producers than the control AIR 40 PER CENT OXYGEN Number Date ot Number | Number | Number Datei Number | Number neeees first young 3 °) oe eS first young a % Be Apr. Apr. 1 3 0 16 1 3 Bh 28 2 6 1 36 2 4 1 2h 6 0 29 3 6 2 32 3 8 0 8 4 8 4 46 9 0 6 5 10 7 43 4 10 3 48 6 12 4 42 5 12 11 43 Tl 14 10 39 6 14 8 33 8 16 2 39 if 16 1 22 9 18 0 8 8 18 0 6 19 0 12 19 0 8 10 20 0 14 9 20 0 23 11 22 12 38 10 22 4 23 22 4 35 11 23 1 35 12 23 2 44 12 25 1 35 13 25 9 25 MLOGS Reape. no ce Aa 30 Sil 59 482 Percentage of 9 .. 7.4 10.9 MALE-PRODUCTION IN HYDATINA 145 Table 14 records the results of this experiment. There is a small increase in male-production in the oxygen line. Another feature of the oxygen line that is perhaps of signifi- cance in connection with the next experiment, is the distribution of the male-producers in the family. In any line of Hydatina the male-producers tend to occur in groups; that is, several successive members of a family will be male-producers. This grouping is most apparent in lines in which male-producers are abundant; for, though the most frequent number of male- producers in a group is always one, by far the majority of male- producers in such a line occur in much larger groups. When the proportion of male-producers is 30 to 40 per cent, it is not uncommon to find groups of ten, twenty, or even thirty or more male-producers occupying successive places in the family. The cause of this grouping is unknown, but evidently when the con- ditions are right for the appearance of male-producers, they remain so for a time, then disappear or are inoperative. Were there no such grouping, in a line yielding only 5 or 10 per cent of male-producers, most of these would necessarily occur in groups of one. Groups of two would be very uncom- mon, while larger groups would be almost unknown. It is interesting, therefore, especially in view of the fact that large groups are quite common in lines in which male-producers are abundant, to compare the distribution of the male-producers in the two lines of table 14, one having a higher proportion of male-producers than the other. The daily records of these two lines have been examined, and the number of times that one, two, three, four, or five successive members of the family were male-producers was recorded. There were no groups larger than five. If such records are to be trustworthy, it is essential that the order of individuals in the family be known. ‘There is a possibility of error here, for the order of age was determined each day by the order of size. In the young females the dif- ferences in size usually left little doubt as to their relative ages, but errors were undoubtedly sometimes made. However, errors were no more probable in one line than in the other; and there could have been no bias, either conscious or uncon- THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 21, No. 1 146 A. FRANKLIN SHULL AND SONIA LADOFF scious, on our part, since the order was determined long before it was known which ones would be male-producers. The number of groups of successive daughters, of one to five members, comprising only male-producers, is summarized in table 15. The slight increase in the mean number of male- producers per group in the oxygen line is not in proportion to TABLE 15 Showing the number of times that groups of one, two, three, four, or five successive members of the families in table 14, were male-producers - NUMBER OF MALE-PRODUCERS NUMBER OF GROUPS IN NUMBER OF GROUPS IN OCCURRING IN SUCCESSION AIR LINE OXYGEN LINE 1 15 27 2 4 10 3 1 1 4 1 1 5 0 1 Mean size of group....... /yLe42 1.47 the increase in the total number of male-producers in that line. The additional male-producers called forth by the oxygen ap- pear, therefore, in additional groups of small size, rather than as additional members of the groups which would otherwise appear. Thus, there is an increase in the number of groups, rather than an increase in the size of the groups. This means a more uniform distribution of the male-producers through their respective families in the oxygen line. Experiment 18. Continuous rearing in 60 per centoxygen. This is a repetition of Experiment 17, but with an atmosphere of which 60 per cent was oxygen, instead of 40 per cent. The methods used were otherwise the same. Table 16 gives the results by generations. There is no increase in the number of male-producers in the oxygen line; whether the slight decrease is significant is not known. Although there is practically the same proportion of male- producers in both lines in this experiment, the distribution of these male-producers over the families is more uniform in the oxygen line than in the control. This is shown first by the number of male-producers occupying successive places in the MALE-PRODUCTION IN HYDATINA 147 TABLE 16 Showing number of male-producers (7 2) and female-producers (2 9 ) in two lines ‘ of Hydatina, one reared in water in ordinary air, the other in water saturated with an atmosphere of which 60 per cent was oxygen AIR 60 PER CENT OXYGEN Number Date of ar Number Number Date of Number Bamber f fo) first fo) of of first ° te) generation young a? 99 generation young oe? come] June June if i 1 20 1 1 4 31 PL 3 0 3 2 3 0 9 4 0) 1 3 5 2 22 6 3 13 5 4 25 3 a. 12 10 4 af 3 37 ay 9 10 31 5 9 4 ilf/ 5 11 0 10 6 11 1 3 6 14 0 5 i 14 5 28 14 0 1 8 17 16 16 14 2 28 9 19 3 42 7 15 13 23 10 20 1 8 8 17/ 23 20 21 11 38 9 19 3 26 i! 22 19 31 10 21 15 23 12 25 7 31 iM 23 1 46 13 20 11 23 12 25 15 36 14 28 14 25 13 27 18 34 15 30 2 25 14 29 4 45 July 15 30 6 32 16 pe 6 11 July 17 4 1 30 16 2 24 24 18 6 35 12 17 4 9 39 18 6 24 29 ARO GALS: cere ahets 183 - 499 149 464 Percentage of 7 2 26.8 24.3 families. Table 17 records all these groups of successive male- producers in both lines. The average size of these groups in the air line is 3.26, in the oxygen line only 2.40. The male- producers are distributed more uniformly through the families in the oxygen line than in the control. In this connection, see also table 15, and the discussion of it under Experiment 17. This difference in the distribution of the male-producers is also seen if, instead of recording the generations separately, as 148 A. FRANKLIN SHULL AND SONIA LADOFF TABLE 17 Showing the number of groups (of various sizes) of successive male-producers in the families of the two lines recorded in table 16 NUMBER OF MALE-PRODUCERS NUMBER OF GROUPS IN NUMBER OF GROUPS IN OCCURRING IN SUCCESSION AIR LINE OXYGEN LINE 1 32 33 2 4 18 3 4 2 t 3 3 5 5 2 6 2 1 7 0 2 9 3 0 12 1 0 19 1 0 24 1 0 32 0 1 Mean size of group of successive male-pro- CUCEES 37 0 been, iets tae 3.26 2.40 in table 16, one examines the daily output. In each day’s product occur members of several successive generations. As Shull (1915) has pointed out, rhythmical production of male- producers may often be more easily detected by an examination of daily records than by a study of family records. In table 18 is recorded the number of male-producers and female-producers on each day in each line of table 16, regardless of the families to which they belong. The percentages of male-producers are computed for tri-daily periods, in order to smooth some of the enormous fluctuations in both lines. These tri-daily percentages are graphically shown in figure 1, where it is seen that the oxy- gen line is plainly more uniform than the control line. If it be objected that the daily records should not be combined for comparison, that combination may be rejected, and the dif- ference between the two lines is still plainly visible. The num- ber of male-producers appearing on successive days in the control line is subject to greater extremes of fluctuation than in the oxygen line. Of the 39 days through which the experiment extended, there were twelve days on which no male-producers ap- TABLE 18 Showing the number of male-producers and female-producers appearing each day in the families recorded in table 16, regardless of the families to which they be- longed AIR LINE 60 PER CENT OXYGEN LINE DATE | Number of Number of Per cent of Number of Number of Per cent of ros 29 a9 fos’) 29 (ope) June 1 0 2 0 1 2 0 5 5.8 1 11 10.7 3 1 9 2 13 4 0 0 0 6 5 0 8 15.7 1 11 12.5 6 3 8 4 18 i 1 8 1 23 8 4 4 50.0 1 16 10.7 9 11 4 5 19 10 6 ii 2 6 1S 0 13 15.0 0 10 14.2 12 0 14 1 2 13 0 6 0 1 14 0 6 0.0 0 a 0.0 15 0 6 0 6 16 2 18 2 3 17 0 17 22 5 10 37.9 18 10 7 4 5 19 vi 13 7 19 20 10 12 44.2 5 16 17.6 21 10 9 0 21 22 3 22 2 15 23 6 13 26.2 7 22 27 ea 24 7 10 10 14 25 1 21 4 17 26 1 23 10:2 5 13 28.5 27 6 26 . 9 15 28 9 21 4 20 29 17 13 29.5 9 15 22.0 30 3 85 4 25 July 1 2 29 6 21 2 0 ; 11 11.9 4 7 25.0 3 6 19 2 8 4 5 11 2 Zz ) 5 14 11 : 44.0 5 6 26.6 6 14 20 1 9 7 10 16 9 18 8 14 15 38.7 13 6 56.6 9 0 7 1 2 149 150 A. FRANKLIN SHULL AND SONIA LADOFF peared in the control line, only seven such days in the oxygen line. On the other hand, there were nine days on which the control line yielded ten or more male-producers, only three such days in the oxygen line. The second of the above-mentioned dif- ferences may be partly due to the greater absolute number of male-producers in the control line; but if allowance be made for this fact, the former difference becomes all the more striking. Percentage of og 1-3 4-6 7-9 10-12 12-15 16-18 19-21 22-24 25-27 28-30 1-3 4-6 7-9 June Jul Dates Fig. 1 Graphic representation of the proportion of male-producers, computed for three-day periods, in the two lines recorded in tables 16 and 18. The light curve represents the oxygen line, the heavy curve the control. The fluctuation is less in the oxygen line. A further difference between the two lines is the existence of four ‘fairly distinct waves of male-production in the control line, each separated from the others by periods of few male-producers. Such a rhythm in the oxygen line is less distinct, orin part wanting. These differences are all due to a more uniform distribution of the male-producers over the families of the oxygen line, and over the period of the experiment in the oxygen line, than in the control. Thus, although the oxygen did not cause an in- crease in the number of male-producers, it was not without its effect upon male-production. MALE-PRODUCTION IN HYDATINA 151 Why the 60 per cent oxygen in this experiment did-not in- crease male-production, whereas the 40 per cent oxygen of the preceding experiment did, is not known. Different concentra- tions of the same agent may have different effects. Or oxygen may increase male-production only when the other conditions present are rather unfavorable to male-production. That the other conditions were right for high male-production is shown by the fact that the control line produced 26.8 per cent of male producers, as against 7.4 per cent in Experiment 17. The stock of rotifers used was, at the time of the experiment, in one of its waves of high male-production. This may be the reason why the oxygen could not still further increase the number of male- producers. Experiment 19. Oxygen counteracting bouillon. This and the fol- lowing three experiments were suggested by the possibility, mentioned in the preceding paragraph, that oxygen couldincrease male-production only when other conditions were rather unfavorable to male-production. The several experiments immediately following show the influence of oxygen in counteracting the effects of agents known to reduce the num- ber of male-producers. In this experiment, a one-seventh per cent solution of Armour’s beef bouillon cubes was used. Only the parents were reared in the bouillon, not the offspring (this being Whitney’s method). The parents were put into bouillon for 10 to 14 hours, then transferred to a new dish of bouillon. This transfer was made to insure that all the eggs laid in the second dish went through their maturation stages in the bouillon. The bouillon in the second dish was first saturated with an atmosphere con- taining 40 per cent oxygen, and the dish was then placed under a bell jar, in a 40 per cent oxygen atmosphere. The parents remained in the second dish for twenty-four hours, at the end of which time they were removed. All the eggs laid in the second dish of bouillon were allowed to hatch there, after which the young were transferred to spring water, where they grew to maturity. The control parents were kept only in spring water which was kept in ordinary air. Every time the parents in bouillon were transferred to a new dish, the control parents were likewise transferred to fresh spring water, to prevent the unequal accumulation of metabolic prod- ucts in the two lines. Table 19 shows the results of the experiments. The oxygen appears not only to have counteracted the effect of the bouil- 152 A. FRANKLIN SHULL AND SONIA LADOFF lon, but to have added a substantial balance of male-producers on the bouillon side. TABLE 19 Showing the effect of oxygen and bouillon on one line, as contrasted with a line not subjected to either. Bouillon, even in dilute solutions, has been shown to reduce the number of male-producers (see table 9). In this experiment, oxygen coun- teracts the effect of the bouillon, and actually increases the proportion of male- producers above that in the control line SPRING WATER, WITHOUT OXYGEN BOUILLON, WITH OXYGEN Number of experiment ee ve Number of experiment Seer pias A 3 23 A 1 24 B if 11 B 10 4 C 1 P23) C 0 8 D 0 14 D 0 3 E 0 16 E 0 23 F 0 26 F 0 12 G 0 13 G 0 9 H 1 wD, H 6 12 I 0 3 I 0 11 J 0 22 A 14 20 K 0 6 K 0 2 een | Mo tal? eee 32 6 179 31 128 Percentage of & @ on 19.5 Experiment 20. Oxygen counteracting strong manure solution. In this experiment only the parents were reared in the manure solution, the offspring being in all cases reared in spring water. The parents to be used for both experiment and control were placed in manure solution in the afternoon; the next morning one lot (the control) was transferred to a new dish of manure solution, the other lot was transferred to manure solution which had been saturated with an atmosphere of which 40 per cent was oxygen. The former lot was kept in air, the latter under a bell jar in an atmosphere of 40 per cent oxygen. At the end of 24 hours the parents were removed from both dishes. All eggs laid in the 24 hour period were allowed to hatch where laid, the young females being then removed to spring water. The young females from each lot of parents were transferred to spring water on two successive days. When the parents were removed, those eggs that were laid in the first seven to ten MALE-PRODUCTION IN HYDATINA £53 hours had already hatched. These females were transferred to spring water at that time; in table 20 they are recorded as ‘older daughters.’ On the following day all the remaining eggs TABLE 20 The parents were reared in strong manure solution, but in one-half of the experi- ment this solution was saturated with an atmosphere 40 per cent of which was oxygen. All offspring were reared in spring water. The oxygen approximately doubled the proportion of male-producers MANURE SOLUTION, AIR MANURE SOLUTION, 40 PER CENT OXYGEN ant aa ee He ene meee umber ae crs Aue ae a aes | ae Baran 2 Ie a? 92 ose) 22 fos) 22 Apr. 28 5 0 17 0 12 4 0 13 0 13 29 3 1 1 3 17 3 0 4 0 8 30 3 0 a 0 9 3 0 f 0 1 30 2 0 2 0 6 2 1 5 0 5 May 4 2 0 0 0 15 2 0 4 1 12 4 2 0 3 0 13 D, 0 4 2 ig 5 2 0 5 0 iL7/ 2 0 3 0 6 5 2 0 5 0 24 2 0 9 0 17 6 2 0 5 I 13 3 0 18 0 15 6 4 0 12 1 37 3 0 1l7/ 0 19 7 2 0 6 0 14 3 0 11 0 18 a 4 0 9 0 29 3 0 8) 0 16 9 4 0 19 2 48 + 0 17 3 33 10 3 1 8 0 26 4 0 25 1 38 12 3 0 22 0 23 2 2 16 1 18 14 3 0 6 0 25 3 0 8 1 16 16 3 0 11 3 ly 2 0 a 1 11 18 1 0 7 1 8 3 0 4 2 5 20 4 2 8 0 7 f 2 13 0 22 22 3 0 14 0 a 2 0 a 3 10 24 _4 3 16 2 22 L 6 23 6 23 26 3 0 1 0 17 3 1 5 0 21 28 5 0 13 0 50 5 3 21 6 33 30 3 0 2 0 20 3 2 4 0 10 Total.. 72 a 194 13 476 71 17 251 27 381 Percentage of 154 ' 4, FRANKLIN SHULL AND SONIA LADOFF had hatched; the young females were removed to spring water, and appear in table 20 as ‘younger daughters.’ The sex-ratio for the older and younger daughters is given separately to show that the effect of the oxygen is as marked at first as it is later. The proportion of male-producers is approximately doubled by the oxygen. Experiment 21. Oxygen counteracting weak manure solution. In this experiment two lines were bred continuously in weak manure solution. From the first several daughters of each generation one was selected to become the parent of the next generation. One line was reared in a dilute manure solution saturated with a 40 per cent oxygen atmos- phere, the other line in a similar (but not oxygenated) solution. In the former line, only the parents and the first few daughters of each family were kept in the oxygenated solution, the other daughters being transferred to spring water. The method of conducting the TABLE 21 Two lines of rotifers, both reared in dilute manure solution, are here recorded. One line was kept in manure solution saturated with an atmosphere of which 40 per cent was oxygen. The oxygen increased the number of male-producers DILUTE MANURE SOLUTION, AIR DILUTE MANURE SOLUTION, 40 PER CENT OXYGEN Number Date of sale pect eeree Date of Number Number Beneration first young a 2 29 senacuitian first young 3 9 ° fe) July July 1 9 i 40 1 9 3 - 14 2 11 0 42 2 il 0 33 3 13 1 18 3 13 0 33 4 14 0 21 4 14 13 14 5 16 0 31 5 16 1 14 6 li 0 6 6 17 (Oh on 7 19 3 40 18 0 11 8 20 5) 19 tf 19 il 40 9 22 2 45 8 21 13 37 10 24 2 32 9 22 2 33 aa 25 0 31 10 24 4 40 12 20 0* 23m 11 25 5 25 12 20 ie 25* LO Galea a boise 3: 14 348 43 322 Percentage of o'Q@ 3.8 TANS 4 * Remainder of family not recorded. MALE-PRODUCTION IN HYDATINA Td experiment was the same as in Experiment 17, except that a manure solution was used instead of spring water. Table 21 shows that the oxygen increased the number of male-producers. Experiment 22. Oxygen counteracting creatin. In the experiments of Shull (11) it was shown that creatin was one of the most effective agents in the reduction of the number of male-producers. A more dilute solution of creatin was used in the following experiment. The parents alone were reared in the creatin solution. On one side of the experiment this solution was saturated with an atmosphere of which 40 per cent was oxygen. The parents on both sides were kept only 24 hours. All eggs laid in that time were allowed to hatch, and the young were reared to maturity. Table 22, which records this experiment, shows that oxygen increased the number of male-producers. TABLE 22 The parents of the ‘rotifers here recorded were reared in solutions of creatin of vary- ing concentrations; but in one-half of the experiment, the creatin solution was saturated with an atmosphere of which 40 per cent was oxygen. All offspring were reared in spring water. The oxygen increased the number of male-producers Rea CREATIN, AIR CREATIN, 40 PER CENT OXYGEN DATE OF CREATIN, ma eae Sh oe ee | ie ne ar IN Number Number Number Number Number Number PER CENT fo) o of of fo) oO parents one) O28 parents a? 29 Dec. 7 0.01 3 1 21 = 2 20 8 0.01 3 0 7 3 0 8 9 0.005 3 1 24 3 6 20 11 0.005 3 13 10 3 12 16 12 0.005 3 9 8 o 25 1 13 0.005 3 1 15 3 12 17 14 0.005 2 2 13 3 7 10 16 0.005 + 6 19 aa 19 15 18 0.0083 3 0 16 3 1 10 19 0.005 4 1 15 4 2 22 20 0.0083 4 3 12 4 0 9 PAL 0.0083 5 1 12 5 3 De MO Gal eaer es sees sc ts 40 38 172 41 89 177 Percentage of J Q.......... 11 33.4 156 A. FRANKLIN SHULL AND SONIA LADOFF DISCUSSION The more recent investigations upon Hydatina, conducted chiefly by Whitney and Shull, left no doubt as to the main prob- lem which those investigations were designed to solve. Exter- nal and internal factors both determine the amount of male- production. But the evidence of the operation of external factors was so abundant that new problems were at once created. All such agents at first discovered, and they were not few in number, had the same effect; they reduced male-production. It became important to discover the methods, or preferably method, by which these very diverse agents produced their common result; for by that means it appeared most likely that the solution of the second new problem would be reached. This latter problem was to discover a method of increasing male- production. Only in this way, it seemed, was it likely that the ultimate aim of these studies, the discovery of the physiological phenomena accompanying changes in the mode of reproduction, would be attained. The former question, namely, that regarding the modus oper- andi of various chemical substances in retarding male-production, is still unanswered. The experiments described in the first pages of this paper, on osmotic pressure, acidity, and possible after effects of manure solution, led us to no conclusions. . Fail- ure to solve this problem has made the attack of the second prob- lem, discovery of means of increasing male-production, largely a matter of trial and error. Even by this method some success has been attained; experiments 4, 5, and 6 show that calcium chloride has the desired effect upon certain lines of rotifers. Whitney’s fortunate discovery that feeding the rotifers on a green flagellate increased male-production gave us our only clew. We have discussed Whitney’s experiments above on pages 137 to 139, and will not repeat here. Repeated experiments with practically uniform results have. demonstrated, to our satisfaction at least, that part of the in- creased male-production following the use of Chlamydomonas as food, in Whitney’s cultures, was due to the oxygen liberated MALE-PRODUCTION IN HYDATINA ESZ by the green flagellates as a by-product of photosynthesis. Our results show marked effects of oxygen. However, the in- crease of male-production with oxygen alone is by no means as great as Whitney obtained with Chlamydomonas. We sus- pect, therefore, that although all the factors obviously associated with Chlamydomonas in the cultures should be separately tested before any residue of influence is assigned to nutrition, Whitney’s conclusion that the food conditions influence male- production is correct, though that influence is less than he be- lieved. We are convinced, nevertheless, that this influence of nutrition is dependent chiefly, if not wholly, upon quality, not quan- tity, of food. Although large quantities of Chlamydomonas in Whitney’s cultures produced more marked results than small quan- tities, this is to be attributed, we think, to the greater quantity of oxygen evolved. It can not be assumed that any ‘law of mass action’ holds for the ingestion of food by organisms. That is, after a certain optimum (probably moderate) quantity is reached, further ‘concentration’ of the food does not necessarily increase the nutrition of the devouring animal. But such con- centration of a green flagellate does increase the concentration of oxygen in solution. If experiments with moderate quantities of Chlamydomonas show an increase of male-production, it is probable that such increase more nearly represents the effect of nutrition itself. _ With the discovery that oxygen increases male-production are we any nearer a knowledge of the fundamental causes of changes in the life cycle? We are inclined to answer in the affirmative. It may be recalled that male-production is ordinar- ily subject to marked periodicity (Mitchell, 713; Shull, 715). Periods of many male-producers are also usually periods of rapid growth and reproduction. Lines that yield many male- producers are also usually vigorous. Metabolic processes are going on at arapid rate. It is hard to avoid the suspicion that in some way this speed of reaction in the protoplasm is related to the production of males. Probably not all metabolic proc- esses, but only certain ones, are thus related to the sex-ratio. For, while, as just stated, periods of male-production are usually 158 A. FRANKLIN SHULL AND-SONIA LADOFF also periods of rapid growth, not all periods of rapid growth are accompanied by many male-producers. And while lines producing many males are usually vigorous lines, there are equally vigorous lines (judged by all known standards) that produce comparatively few males. These facts may be har- monized with the view that speed of reaction within the proto- plasm makes for male-production, if we assume that only cer- tain reactions bear this relation to sex. If we adopt the view that the rate at which certain chemical events proceed determines the form of the life cycle, the posi- tive results of our present experiments have some meaning. The dilute solution of calcium chloride merely provides a medium that is slightly more favorable to the processes concerned. Oxygen obviously provides for accelerated oxidation in the protoplasm. The effect of oxygen aside from male-production is not merely inferred; for the rotifers in oxygenated water, and those in Spirogyra cultures, were almost invariably healthier than the control. Sometimes the families in the oxygenated water were larger, but always the animals were more easily reared to maturity. If qualitative differences of nutrition affect the sex-ratio, as we must on the basis of Whitney’s experi- ments assume that they do, these may likewise be conceived to affect the speed of the metabolic reactions. Even quantita- tive differences of nutrition, should these eventually be found to alter the sex-ratio, could conceivably alter the metabolic processes; indeed, there are no a priori grounds for believing that they do not. The suggestion that mere speed of reaction is responsible for varying degrees of male-production becomes more plausible if its operation can be visualized in terms of known processes. It is well established, we believe, by the investigations of Shull (12, pp. 3802-308), that male-production is either caused or prevented at some time within the growth and maturation pe- riod of every parthenogenetic egg. When an egg has passed its maturation stages, the fate of the female which will hatch from that egg is sealed. She will be either a male-producer or a female-producer, according as one or another series of events MALE-PRODUCTION IN HYDATINA 159 has taken place in growth or maturation, and her nature is no longer subject to alteration. What happens to decide this fate is unknown. It may be the failure of some chromosome to divide. The male-producing female may have fewer chromo- somes than the female-producer. The apparent variability of the number of chromosomes (Whitney, 09) may be due, not entirely to difficulty in counting them, but partly to actual differences. , If such behavior of the chromosomes results in the develop- ment of a male-producer, the rate of formation of the spindle, or of the division of chromosomes, may be the cause of the chromosome change. A chromosome dividing a little later than its fellows may be drawn (?) to one pole without completing its division. The cytology of the germ cells of Hydatina should be re-examined, with a view to discovering the difference be- tween male-producing and female-producing females. But we emphasize that our theory of the speed of reaction is not bound up with chromosomes, to stand or fall with future discoveries regarding the chromosomes of these rotifers. Other phenomena than chromosomes which have, with present technique, no visible expression, mdy as conceivably be influenced by the speed of metabolic processes. If male-production is related to the rapidity with which cer- _ tain physiological processes occur, it is not surprising that a change of environment more often reduces male-production than increases it. A high degree of male-production, on our view, depends upon an efficient mechanism working at top speed. Any unskilled workman may ruin a delicate machine, it takes an inventor to improve it. Many changes of environ- ment may retard metabolic processes, only a few accelerate them. We should expect, therefore, that most agents which affect the life cycle at all would reduce male-production, and this has. been the case. It may be suggested that our view is very near that of Whit- ney (14), Mitchell (13), and Nussbaum (’97), that nutrition is the controlling factor in the life cycle of Hydatina. If nutri- tion be re-defined to include all chemical processes in proto- 160 A. FRANKLIN SHULL AND SONIA LADOFF plasm, the suggested factor may be classed under nutrition. But by so defining nutrition the term metabolism becomes super- fluous. Furthermore, it is clear from the context that none of the writers just named meant anything more by nutrition than quantity or quality of food devoured. We care not about the terminology, the ideas are distinct. SUMMARY The common effect of numerous substances upon the life eycle of Hydatina senta (diminution of male-production) is not due to their osmotic pressure, acidity or alkalinity, nor to mere delay of certain processes. Calcium chloride, in very dilute solutions, repeatedly increased male-production in one parthenogenetic line, not in another. Magnesium chloride gave results that could not be interpreted, while potassium sulfate, iron chloride, and ammonium chloride all reduced male-production. Dilute bouillon also diminished male-production. Oxygen in the water increases male-production. Its effect is most marked in the counteraction of agencies which diminish male-production, such as bouillon, manure solution, and creatin. Whitney’s experiments with Chalamydomonas, in which male- production was greatly increased, an effect which he attributed to nutrition, are partly explained, therefore, as dependent upon the oxygen evolved in photosynthesis. Our results, however, were not as marked as Whitney’s, and experiments in which the rotifers were reared with a green alga too large to be eaten, gave negative results. It is probable that nutrition has some effect, as Whitney supposed, but to what extent can not be known until the other agents which can not be eliminated are separately tested. MALE-PRODUCTION IN HYDATINA 161 BIBLIOGRAPHY Maupas, E. 1891 Sur la déterminisme de la sexualité chez l’Hydatina senta. Comp. Rend. Acad. Sci., Paris. T. 113, pp. 388-390. MircuHEett, C. W. 1913 Sex determination in Asplanchna amphora. Jour. Exp. Zodél., vol. 15, no. 2, pp. 225-255. Nusspaum, M. 1897 Die Entstehung des Geschlechtes bei Hydatina senta. Arch. f. Mikr. Anat. u. Entw., Bd. 49, pp. 227-308. -Suutu, A. F. 1911 Studies in the life ae of Hydatina senta. II. The rdéle of temperature, of the chemical composition of the medium, and of internal factors upon the ratio of parthenogenetic to sexual forms. Jour. Exp. Zoél., vol. 10, no. 2, pp. 117-166. 1912 Studies, etc. III. Internal factors influencing the propor- tion of male-producers. Jour. Exp. Zoél., vol. 12, no. 2, pp. 283-317. 1913 Eine kiinstliche Erhéhung der Proportion der Mainnchenerzeuger bei Hydatina senta. Biol. Contralb., Bd. 33, Nr. 9, Sept. 20, pp. 575-576. 1915 Periodicity in the production of males in Hydatina senta. Biol. Bull., vol. 28, no. 4, pp. 187-197. Wuitney, D. D. 1909 Observations on the maturation stages of the parthe- nogenetic and sexual eggs of Hydatina senta. Jour. Exp. Zodl., vol. 6, no. 1, pp. 137-146. 1914 The influence of food in controlling sex in Hydatina senta. Jour. Exp. Zodél., vol. 17, no. 4, pp. 545-558. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 21, No. 1 a" aan -, Bi. WAT. te im + aed pon’ Git Mig? pinta ye sith oe i me ewe i pibalrenmea “al! “i ae pc CHEMICAL CONTROL OF RHEOTAXIS IN ASELLUS W. C. ALLEE Marine Biological Laboratory and Lake Forest College} TEN FIGURES CONTENTS Me ATierawOEKsana MetnOUS. se ceaacie ees ft oes sso 0ry ARE eee hls 163 ENC CATON PS eres 7s fae te Heats SWE cdg Seid 6.8 Gd Ss Rb > Cs. When the amount of increase in positiveness is compared N/5 sodium chloride is less effective than N/10 lithium chloride but if N/5 solutions of both salts are compared the former is more effective and the above order holds for the amount of increase as well as for the number made more positive. If this series be written in the ordinary physio- logical style itis Cs < Li < Na < Rb < K witha sharp division between the sodium and rubidium. This reminds one of the usual physiological order of these cations except that caesium which usually lies between lithium and sodium in its effects on living tissues and colloids (Héber ’14, s. 487) here is less effective than either. 168 Wine: ALEEE The chlorides of potassium and rubidium were found to be most efficient in increasing the positiveness of the rheotactic reaction. These salts are chemically closely related and must represent the optimum of the series of conditions that are found in the alkali metal series. The effect of these two cations (with the chlorine anion) is so striking that it is worth while exhibiting the result of their action. The effect of rubidium chloride may be seen in table 2 and that of potassium chloride in table 5, p. 190. TABLE 2 Showing some diagrammatic reversals in the sign of the rheotactic reaction caused by n/10 RbCl are RHEOTAXIS BEFORE TREATMENT | RHEOTAXIS AFTER TREATMENT Leer a Kae TEMP. + = a El ar = a El min. 514 100 2.25 | 100 LEO} kG: 7.0.| > 2p 525 30 70 | 0.90 80 | 20 1.0) 500 6.0.; 21 526 100 2.65 60 |; 20 20 Mee || ~ LMS BOS) 2 527 100 3.25 80} 20 0.5 | 16 SHON Zt 528 90:) 20) 2) 1365 60 |} 20 20 0.8 8 6.0; 23 529 100 3.00 | 100 Bay fal eG 6.0 | 23 531 10 90 2.55 | 100 13) | 26 6.0 | 23 532 20 20; 60 | 1.00; 100 1.4] 15 %j0%| 23 533 10 90 | 1.10 80 20 1.25) 5 G40) 223 Ave... 8 67 | 25 | 2.04 84 | 9 7 12a el 6.7 1 The figures in these columns give in a standardized from the amount of movement of the isopods during their rheotactic test. For detailed account see 713, p. 261. The cases listed in the rubidium table are selected and are the most diagrammatic obtained with that reagent while those shown for potassium include all the tests for a given experiment which explains the apparently greater efficiency of rubidium. The stimulating effect of potassium and rubidium chlorides was noticeable in the general behavior and particularly in the increased rapidity of movement of the isopods. One of many similar laboratory notes for rubidium (N/10 solution) is: ‘‘Isopod became nervously active immediately although it had been slug- gish before.”” Another for potassium (N/10) is “All nervous, excited, although losing power of codrdination.’’ Continued CONTROL OF RHEOTAXIS IN ASELLUS 169 exposure even to these stimulating salts results in a decrease in the positiveness of the reaction but at times, particularly in potas- sium chloride, this decrease comes only when the isopods lose coordination through the toxic action of the solution. They may either lose power of positive orientation while still strongly stimulated or may orient repeatedly but be unable to hold the position. While in this state the isopods have a tendency to run in small circles of about a centimeter in diameter. Similar circular reactions have been observed under natural conditions but not to the extent produced in these solutions. Sodium chlor- ide, the next most stimulating salt, gives none of these phenomena and its stimulating action is slower as well as less pronounced. It has been shown that a relationship exists between the posi- tiveness and the efficiency (the distance covered during a minute reaction period) of the rheotactic reaction (13). This has since been amply confirmed in the study of reactions under natural conditions or under such depression as may be caused by calcium chloride or cane sugar. That the two phases of the reaction are not correlated at all times is shown by the ex- periments with rubidium and potassium chlorides just mentioned and is typically illustrated by the results shown in table 2. In the trials listed there the positiveness of the isopods was in- creased markedly in all cases but in all save one the efficiency of the response decreased, and while the positive rheotactic reaction was increased 76 per cent the efficiency was cut al- most in half. In the very beginning of the treatment the activity increased as has been noted but the subsequent depression often came before the isopods lost the power of positive orientation. This toxicity-depression ratio (dotted line figure 1) which was high in the case of potassium and relatively so with rubidium was quite low with lithium and caesium both of which depress the rheotactic positiveness long before the toxic effect is appar-. ent. The ratio was not found for sodium because the toxicity was not accurately determined. In considering the relative toxicity of the different alkali metal cations it should again be noted that the comparison in the figure (unbroken line) is between N/5 sodium as compared 170 W. C. ALLEE with N/10 solutions of the other cations. Even with this dif- ference sodium is least toxic, followed after a long interval by lith- ium, caesium, rubidium and potassium in the order named. The favorable effect of sodium chloride is shown by the fact that ten isopods were reacting after eleven hours and at fourteen hours six of these were more positive than atthe beginning. Three were still reacting after sixteen hours and lived for five days in N/5 solution. The cation NH; is chemically closely related to the other members of this group and is often similar in physiological action to potassium which it particularly resembles. In its effect on rheotaxis, however, ammonium chloride is less effective than caesium chloride so that the complete series would be NH; < Cs < Li < Na < Rb < K. The toxicity effect of ammonium is like that of potassium and rubidium as shown by the series Na < li < Cs < Rb < NH. < K. This is in the same order as has been found to preserve the irritability of fresh frog’s nerve (Brodsky ’08, vide Hober s. 511); to cause recovery of irritability after loss in cane sugar solution (Hoéber, s. 497) namely: Na < Li < Cs < NH, < Rb < K. The only dif- ference is that in the foregoing instance ammonium stands next to rubidium while with the isopods it is next to potassium. I am unable to state why ammonium should act in its usual manner as regards toxicity and not as regards rheotaxis.: The rheotactic reaction may be influenced by the well developed nervous system of the isopods or by direct action on the muscles or by a combination of the two. It may be we have here a differential effect in that potassium stimulates the muscles more than it depresses the central nervous system while ammonium acts in the opposite manner. Mathews (’07) found that these two cations have similar effect on muscles and on motor nerves but are opposed in their effect on the central nervous system, which potassium depresses while ammonium stimulates it to the point of tetanic convulsions. The action may also be ex- plained by assuming that ammonium produces a greater per- meability than potassium such as Lillie (’09) discovered with Arenicola larvae. It is known (Mathews ’07) that the am- CONTROL OF RHEOTAXIS IN ASELLUS 17. monium salts may be eccentric in their action and Mathews interprets this as due largely to their hydrolytic dissociation but this explanation is unsatisfactory so: far as the present work is concerned because of the slight amount of this kind of dis- sociation in the solution strengths used. CHIR ane Bees ae nas map bee naam coe aE Peete el bl Hp LAN Yee ese HBS ake eee eee) HA A LL Sel sees lae pine i oo - oo, HNL # Ere tha f fa 2s Be o SRD MARSA eR ea BRS Shae oie Ree eewaeoe Fig. 2 Showing relative toxicity and the relative effect of different anions with the cation potassium on positive theotaxis of Asellus. For details see text Dali2e Different anions with the cation potassium Some work was carried on to find whether different anions with the most effective cation, potassium, also affected rheotaxis. The studies were only carried far enough to demonstrate that the anions are effective and to suggest in a tentative way the order of their relative effect. The findings with N/10 solutions, which was the strength most often used, are shown in figure 2. Figure 2 is charted in much the same manner as the preceding figure. The anions are arranged in their usual relative colloidal 172 WwW. C. ALLEE and physiological effect (Héber, s. 308). Again the simple broken line shows the percentage of isopods that were made more positive at some time during the treatment with each ordinate space worth 5 per cent. The dot-dash line gives the _ average difference between the percentage of positive rheotaxis before and at the most positive point after treatment with each | ordinate space equal to 2 per cent. The unbroken line gives toxicity, each ordinate space representing one minute, and the dotted line shows the toxicity-depression ratio with the ordinate spaces worth 5 per cent. Experiments with potassium sulphocyanide in N/5 solution showed that it was more toxic than any other anion tried at this strength. In N/20 solution it caused some increase in positive rheotaxis but the evidence of the experiments at hand is contradictory and there are too few instances to locate this anion definitely. Potassium sulphate (2N/7 and N/15) increased the positiveness of the response, especially in the weak strength where it was almost as effective as potassium chloride in N/10 solution but more toxic. As the figure shows the relative toxicity was Br > NO; > I > Cl > Ac, ClO;, and the effect on positive rheotaxis was Br < NO, < I < CIO; < Cl < Ac if one considers the effectiveness in causing some increase in the positiveness. The same order holds true when the extent of the increase is considered except that Ac is less effective than ClO; and Cl. I am not prepared to state why the order found here differs from the usual relative effect shown by the order of arrangement in the figure or why the more toxic salts should vary in their toxicity and their effect on rheotaxis. It is possible that more experiments might change the order in one or two places, but enough tests have been made to demonstrate conclusively that potassium chloride is the most effective salt so far as amount of change is concerned, and the regularity with which bromine takes the usual place of iodine at this concentration indicates that the order found is not a mere accident. A number of the anions that failed to show a marked increase in N/10 solution were tested in N/20 and N/40 concentration. No marked change occurred except in | CONTROL OF RHEOTAXIS IN ASELLUS LS the bromide-iodide relation just mentioned. At the weakest concentration the bromide became more effective than either the nitrate or the iodide and took its usual place in the anion series viz: NO; < I < Br < Ac. Kt is noteworthy that the toxicity-depression ratio isuniformly high which means that the potassium cation, regardless of the anion, tends to cause the isopods to give a positive reaction well after the amount of activity is markedly depressed. The varia- tion in effect of these anions on rheotaxis must be due in part to some other factor than relative toxicity because the toxicity- depression ratio does not vary sufficiently to account for the rheotactic results obtained. Chlorine salts of the alkali earths Although any chemical at the concentrations used will ulti- mately cause a decrease in rheotactic positiveness, calcium and strontium chlorides were found to be particularly effective in causing this depression without preliminary stimulation. Ba- rium and magnesium chlorides also cause an ultimate depression but they often give a preliminary period of stimulation. Some of the experimental data on which these conclusions are based are shown in table 3 which shows the toxicity of M/5 TABLE 3 Showing the toxicity and effect on rheotaxis of chlorine salts of alkali earths. Stron- tium chloride experiments were run in N/5 and the others in M/5 solutions. Iso- pods with an initial response of more than 71 per cent positive were not consid- ered TOXICITY RHEOTAXIS i re B Depression - Per cent, Amount of ora ape eee time toa, ae gaa waerenke (4) (6) (7) IY (oa SI CUS ao 10 8 hrs.! 21 32 31 —9 Oar ih. Vie it 6 3 hrs. 6 21 0 —36 Sine ee tae ea 11 58 min. 26 7 14 —ll lhr. | ee ae) aa 5 1 a 12 \30 | Hels lal] We Eee Fig. 3 Showing antagonistic action of N/10 potassium chloride and m/5 cal- cium chloride on positive rheotaxis of three isopods. For explanation see text. After about two minutes they were placed on filter paper and in- troduced into the other solution. Experiments showed that this handling had a stimulating effect but this was not sufficient to mask either the antagonistic effects here recorded or the depressing effects of most salts tried. The method of charting is the same in all three figures. One space in the ordinates is worth 5 per cent, one in the abscissae equals five minutes. The lines show the variation in the per- centage of positive responses given and in all cases at the be- ginning of the treatment, the low points mark the response when treated with calcium, the high ones when treated with potassium chloride. The only exceptions are found in figure 176 W. C. ALLEE 5 after the isopods had been treated with alternating solutions for over two hours. The reactions of the different individuals are shown by the different lines and are duly labeled in the figures. No. 188 is repeated in figure 4 for the sake of com- parison. Figures 3 and 4 are from work done in 1913 with N/10 potassium against M/5 calcium chloride and figure 5 is from 1914 ‘results with both solutions 1/10 normal. E/fvanae 1H ERaE: |S eee Lae? SASR a SUMAN JAE |_| \_ iY |_| He al E REARE | Brie - ial EPiBa B E Ban, HT | yf 6 Rh PEAY h_{/ | IAI |_| vi | | B ca i & || af ig o a ai et AE LA \ | A\ a | \ i WP RABE? GRASS Le PAC ACE ERVER ry BR |_| Go VIP ESEY WER ARAN eas ER GRHARPEL UWE MoV Yigal Aidt tT AN SEAGER ANE Bo LIsTA TT Ty do A ~I4GRRRR BASRA af | Fig. 4 Showing antagonistic action of N/10 potassium chloride and M/5 cal- cium chloride on positive rheotaxis of three isopods. For explanation see text [ok JWGy GMB S SSReR EERE T CASA RSL ARE EA 4BEREERE (USERRA ER ASRS ER EESERY. UN SRRR EP Ri eEREREeE Lo. fA af BREE O's0 PERERA RELEEREEES REESE OE USERRA SOS SESES ta ea ALA EL BReliNe A | ial BED se BID RRR CRS WEDGE UBER ISSR PE SACI Wik 42 SRUES US. LER Bee Ra GBR RRL Lee Rae Vie AA pe AN ee a Bch RRR LEERE LD a BEE BEAD | Sot | AD at Ge iS faa | alate ala al aE Tt oat a Ne Vasa ea tt ee lela] ULL fal N L LA di cese5 aan Fig. 5 Showing antagonistic action of N/10 solutions of potassium and cal- cium chlorides on positive rheotaxis of three isopods. For explanation see p. 175. CONTROL OF RHEOTAXIS IN ASELLUS Lie Tests were made for similar antagonisms between sodium and magnesium chlorides which are less effective in influencing the rheotactic response than the ones just given. No evidence of marked antagonism was found until the isopods were left in the different solutions for several hours when a definite an- tagonism was exhibited as is recorded in figure 6 where each space in the abscissae represents an hour. Acids (H tons) It is particularly hard to summarize in brief form detailed results of the experiments with acids. Preliminary tests were run with acetic, N/100, hydrochloric; N/50, N/100, N/500 BSED TE tee AOE eee eee SGGRS "CUS S ASSES EER ees LAURE PRA e ARBs Bes ae ees SUBNE SURE Sees Pee eee BEARS SOTERA ENO ee BR Bees koe SoS Re ae eee eee ae ER RE Lae VA SED BRRE Uh saR Nf htt ee Anes TOS ne ae Ae ils Son eee S HE VOR eee eis eS Fig. 6 Showing antagonistic action of n/5 sodium and magnesium chlorides on positive rheotaxis of one isopod. and N/1000; sulphuric, N/100, N/500, and N/1000; and with nitric acid N/100. There was some slight evidence of stimula- tion with N/1000 hydrochloric and with N/100 acetic acids. These reagents accordingly were tried again the following season and 3 isopods with N /100 acetic acid showed complete depression while 15 trials with 6 isopods using N/1000 solution of the same acid showed depression in all cases save one. A similar repeti- tion with N/i000 hydrochloric showed only one stimulation. To test the matter further a fairly complete series was run with N/4000 hydrochloric acid the results of which are shown in figure 7. In this series 36 trials were made with 14 isopods “LLU ‘d uo syreqop r90yyan ‘UOTPOVII OTJOVJOOYA VATpISOd 9Yyy UL osUvYD yuUdd Jad Gg ‘soyeUIPIO Oy uo puv Moy uv syusserdor oVSsTOSqe oY} UO doVdS YORUM “sIxejooyA aATyISOd uO plow oIMOTYooIpAY CYOOF/N JO Yooyo oy} BSurmoyg 2 ‘Sy 178 CONTROL OF RHEOTAXIS IN ASELLUS 179 whose initial response was under 71 per cent positive with one exception where the reaction was 90 per cent positive at the start. This individual was made more positive by the action of the acid and so may fairly be included (p. 178). Each space in the abscissae represents one hour and in the ordinates 5 per cent change in the positive rheotaxis. The location of the cross gives then the increase or decrease in the positive rheotaxis and the time that the isopod had been subjected to the treatment. The crosses on the base line indicate in this experiment that neither before nor after the treatment did the isopods give a positive response. In brief the results show that 14 tests of 5 isopods showed an increased positiveness averaging 15 per cent, 6 tests showed no change in the positive reaction though considering the change in negative, indefinite and zero responses each indicated a de- pression; and that in 16 tests 9 isopods showed a decrease aver- aging 21 per cent. Asa final summary 39 per cent of the trials showed some increase but the average amount of change in the positive reaction was a decrease of 13 per cent. This is similar to the results obtained with magnesium chloride and indicates a depression. In the more concentrated solutions acids are decidedly toxic as is shown by the following records for hydrochloric acid: Solution Time before strength all are on back Li Donscoe ates olor c Bios SOU DAL Sea Rea Re DEG era ee ait Oe a 1 minute S/O are E Oe OO ne, LOO OTOL OES AL | Coen: 8 minutes N/T) 3 doa led dale b EP UE DU ACTER Oe taba oon Ge 120 minutes n/100 Out of three tried two were dead in 24 hours The toxicity was greater than with any other chlorine com- pound tried and by the usual reasoning in such cases this must be ascribed to. the hydrogen ion and the depressing action. of acids on rheotaxis must also be ascribed to this ion because each anion tried had the opposite effect on rheotaxis when combined with some other cation. By this I do not mean to maintain that the action of the cation is necessarily direct but that the cation here is the important part of the combination either through its direct action or because it fails to entirely THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 21, No. 2 180 W. C. ALLEE neutralize the effect of the anion. As to which of these possi- bilities is correct my work offers no direct. proof. The depressing effect of acids on nervous tissue is well known to physiologists (Mathews ’04). They antagonize sodium chloride (Osterhaut ’14) but less than calcium chloride and they depress permeability as also does calcium. The effect on rheo- taxis is in accord with this work but with it opposes the ob- servations of Bohn (’12) that N/1000 sulphuric acid increases the oxidation rate of lobster larvae and thereby causes them to become positive to light, and those of Lillie (’09) that N/400 and weaker solutions of hydrochloric acid increase permeability and stimulate Arenicola larvae. Alkalies (OH tons) Mathews (’04) found that hydroxyl ions are most efficient nerve stimulants. Bohn (’12) found that lobster larvae were made positive to light by N/1000 sodium hydroxide and states that the action is similar to that of sulphuric acid but less rapid. Loeb and Wasteneys (713 a) report that sodium hydroxide has — no effect on the rate of oxidations of fertilized Strongylocentrotus eggs unless the concentration is above N/1000 and (’15) that this minimal concentration varies with different species. More concentrated solutions cause a marked increase in the oxygen consumption but the eggs are injured. Osterhaut (14) found that in the above strength sodium hydroxide makes practically no change in the permeability of Laminaria. Kanda, (’14) using N/20 potassium and sodium hydroxides obtained no reversals in geotaxis with Arenicola larvae although in common with Bohn and earlier workers he did obtain reversals in the reaction to light. Lillie (’09) using N/25-N/100 sodium hy- droxide found no stimulation of Arenicola larvae and suggests that it decreases permeability. Potassium and sodium hydroxides were tested for their effect — on rheotaxis, the former in N/50 and N/200 solutions and the latter in N/200 and N/500. With the weaker solution of potas- sium hydroxide two of six isopods were slightly stimulated and CONTROL OF RHEOTAXIS IN ASELLUS 181 the average rheotactic reaction was depressed by 25-37 minutes treatment from 40, 32, 28 to 23, 25, 52 per cent positive, negative and indefinite. The isopods lost correlation in this solution in about two hours making the toxicity-depression ratio approxi- mately 31 per cent which is the lowest ratio shown by any potassium salt. None of the six isopods tested with sodium hydroxide were made more positive and the average rheotactic reaction was also markedly depressed. These results with alkalies obviously are. based on too little work to conclude that hydroxy] ions will not stimulate positive rheotaxis but since there was no evidence of such stimulation the tests were discontinued. Distilled water To what extent are the results just recorded due to osmosis? It will be remembered that Asellus is a fresh water isopod inhabit- ing in this case pond water that must exert some osmotic pressure, especially during the summer season when the solutes present are more concentrated. Obviously however the effects recorded cannot all be due to osmosis because as has been shown equi- molecular solutions of certain of the salts had opposite effects. For example N/10 potassium chloride with an osmotic pressure of about four atmospheres is very efficient in causing isopods to give a more positive response. The same strength solution of calcium chloride with an osmotic pressure only slightly greater has the opposite effect. In all probability these were more concentrated than the pond water to which the isopods were accustomed. By the use of distilled water it is possible to obtain a condition where the osmotic pressure is less than that of the pond water though the action of distilled water may be due to some other factor than osmosis. Water redistilled in glass is quite toxic for these isopods but the water furnished by automatic laboratory stills will allow the isopods to live as long as five days. Bullot (04) obtained similar results with the fresh water Gammarus. W. C. ALLEE Each space on the abseissae represents two lurther details on p. 183. Showing the effect of distilled water on positive rheotaxis. 8 Fig. hours and on the ordinate . 5 per cent change in positive rheotaxis. ~ e CONTROL OF RHEOTAXIS IN ASELLUS 183 The results of experiments with 11 isopods are shown in figure 8. The charting is the same as in the preceding figure except that the abscissae spaces represent two hours in place of one. As usual, responses obtained from isopods with an initial reaction of over 70 per cent positive are disregarded. Of the isopods tested 55 per cent were stimulated at some time during their treatment. The 17 of the 35 tests which showed some increase in the positive response averaged 49 per cent more positive than before treatment. Six showed no change in the positive reaction and the 12 trials that were less positive gave an average decrease of 24 per cent. Altogether there was an average increase in positiveness of 16 per cent for the 35 trials. Peters (’04) found that ‘very pure’ distilled water causes Stentor to lose salts but that the animals do not swell, i.e., there is no intake of water. Garrey (vide Mathews ’04) reports that treating Chilomonas with distilled water increased their irritability and Mathews suggests this is because the protoplasm is brought more nearly to the neutrality point. However it is possible that there is an intake of water and there is some evidence (Riddle ’14) that such an increase would result in an increased metabolic rate which could account for the increase in’ positive rheotaxis. This subject will be considered further on p. 186. III. NON-ELECTROLYTES Cane sugar Cane sugar (rock crystal sugar) in M/2 solution exerts an osmotic pressure of about 12 atmospheres (see references to work of Morse and his students in literature list)? which is near three times that exerted by N/10 potassium or calcium chlorides and if the depression or stimulation results given by these salts were due to osmosis cane sugar should give greatly increased effects. That the sugar at this strength causes depression is shown in table 4, which exhibits results obtained by exposing 24 isopods to M/2 cane sugar for 52-75 minutes. As _ usual 3 Garrey 715 found that 1/2 G. M. cane sugar solution gave —1.15 to—1.155 which would give an osmotic pressure of about 14 atmospheres. 184 W. C. ALLEE the reactions of isopods with a high initial positive response are not considered. The averages show that from a rheotactic reaction of 13, 73, 12, 2 per cent respectively positive, negative, indefinite and zero the isopods were changed to 11,19, 63, 7. The decrease in positiveness is insignificant but the decreuse in the negative reaction and the accompanying increase in indefinite responses shows depression in animals with a low initial positive reaction TABLE 4 Showing the effect of M/2 cane sugar solution on rheotaxis of isopods with an initial response of less than 71 per cent positive. In all cases reported the treatment was from 52 to 75 minutes. I indicates an increase and D a decrease in positive rheotaxis caused by the treatment BEFORE TREATMENT AFTER TREATMENT mee | cob (ae (etal in|. ‘Sige + = a O E. | Effect 237 20 | 80 Pl 6.5 50 40 10 1.0 I 239 | 70 10 | 20 1S ae 80 | 20 0.7 D 244 90 | 10 7. 6.0 20 80 1.4 I 245 | 10 80 | 10 Weel 6.5 40 50 10 W374 ee dD) 246 50 | 20 | 30} 0.9] 7.0 10 90 0.6 D 247 | 70 10 | 20 1.4) 6.0 30 20 50 0.9| D 248 | 40 30 | 20 | 10 Weay | fab 20 30 50 0.9 D 249 90 | 10 2.4) 8.0 10 80 10 0.7 D 260 100 acre area) 20 80 1.0 D 261 100 ZEON G25 30 10 | 60 1.2 I 263 | 30 70 D2 ye fer) 40 | 60 0.7 D 265 100 2.3 6.0 30 30 40 1.6 I 266 90 | 10 2 0 60 |} 40 0.4 D 267 | 10 90 save l ASE 70 | 30 0.7 D 283 | 10 90 2.1 5.0 100 Dera i 3D. 285 100 Donna t sO 100 0.6 D 280 100 ibs Pd |p AO) 60 | 40 0.4 D 281 100 HES Woe 20 10 60 10 1.3 I 282 10 | 90 2.05) Seo 30 40 30 1.6 I 277 | 50 50 PAA | CKO) 70 | 30 Vs D 278 100 | 4 2cdA) 8320: 50 10 40 1.2 I 286 | 10 90 1267 PROrO 100 026)" DB 287 100 eG | O30) 10 50 40 1.0 D 288 | .10 80 | 10 1295) 720 20 10 70 0.8 I Ave 13 eT aS 1.8 6.5 11 19 63 7 0.9 CONTROL OF RHEOTAXIS IN ASELLUS 185 just as a decrease in positiveness does when isopods are highly positive at the start. Nine isopods or 38 per cent of those tested were stimulated by this treatment for this length of time. The isopods reacted to water currents after fourteen hours exposure to sugar solutions and fully recovered from the depression when placed in tap water. Needless to say the longer treatment caused a greater depression. Although cane sugar depresses positive rheotaxis it does not do so to the extent that would be expected if the effects of the calcium ions reported above were due to osmosis. No experi- ments were run testing whether or not the effects of sugar could be offset by different ions as in muscle or nerve preparations but three attempts at recovery using distilled water were some- what successful. The best case follows: The rheotactic reaction of isopod 290 was changed by three hours treatment with M/2 cane sugar from 80 per cent positive, 20 per cent indefinite to 20 per cent negative, 60 per cent indefinite, 20 per cent zero. After 3 hours 30 minutes in once distilled water the response was 60 per cent positive, 40 per cent indefinite. This is not the expected result if distilled water acts by a differential removal of salts for cane sugar should act in the same manner. Loeb (’03) found with a marine Gammarus that distilled water and cane sugar solutions had approximately the same toxic effect and ascribed this to the loss of electrolytes or ions into each solution. He suggested that the exit of antagonistic salts takes place with unequal rapidity or in unequal relations. If the effect on rheo- taxis were explained on this basis one would have to assume that calcium or strontium salts escape into distilled water and potassium or sodium salts into cane sugar and finally that’ by withdrawing first enough of one set and then enough of the other the rheotactic reaction would be restored to its original condition. This may be what happens but the observed results can be amply and more simply explained on the basis of a change in water content, with the distilled water allowing an increase in water which should increase the metabolic rate as suggested above and the cane sugar removing water which should decrease the rate ‘of metabolism, and this is what actually happens under treat- ment with cane sugar (p. 193). 186 WiC) ALLEE IV. METABOLISM AND RHEOTAXIS Child (15 and citations) has established a relationship be- tween the metabolic rate of many lower animals and plants and their resistance to potassium cyanide. Hyman (’16, 16a) obtained similar results for certain annelids and sponges and I have found that this relationship holds for Asellus ’14). Gep- pert (’89) for certain mammals and birds, Warburg (’10), Loeb and Lewis (’02), Loeb (’06) and Loeb and Wasteneys (710, 713) for sea urchin eggs found that the addition of cyanide decreased markedly the oxygen consumption; Hyman (’16 a) gives more detailed results of this relationship with sponges. It appears more than probable that the cyanide acts by affecting the oxida- tions but as Child (14) says: Susceptibility to cyanide in concentrations which are lethal within a few hours varies with the general rate of metabolic activity or of certain fundamental metabolic reactions. This conclusion holds whether the cyanide acts more or less directly upon oxidations or upon the condition of the metabolic substratum or certain of its constituents and so indirectly upon metabolism in general. Does resistance to sodium cyanide measure the metabolic rate of Asellus? Certain observations that isopods much depressed by treat- ment with calcium chloride became more active after being put into N/1000 potassium cyanide and that isopods from potas- sium chloride did not, led to a fear that the cation even in this dilution might have some effect on the death point, particularly with isopods previously treated with either of the above salts. For this reason it was thought desirable to substitute sodium cyanide with its less toxic and less stimulating cation. The expectation was not entirely realized for animals treated with calcium chloride until they had lost power to move regained it slightly in sodium cyanide. It is possible that the cyanide itself causes a slight initial stimulation (Loevenhart 706), Hyman (16 a). The use of sodium in place of potassium cyanide required a retesting to determine what strength, if any, would measure’ 187 IN ASELLUS CONTROL OF RHEOTAXIS "Ss ‘d 90s syrvyop Jog “uo aod Z Y}I0M doeds Yous YAIM MOY Yyove Sut -Ap o8vjusd1ed oY} MOYS SoYVUIPAO ‘soyNUTUI OATOM} YIIOM doVds YORI YIM OUT} VAIS OVSSTOSqY ‘snTjosy JO 94BI DI[OqQvyouL oy} Sonsvott opruBAdD UINIPOS ONG-OOF/N 0} 9OUBYSISOI OY} IOYJOYM IOAODSIP 0} 8489} Jo sy[Nsor oyy Surmoyg 6 ‘BLT 188 W. C. ALLEE the metabolic rate of Asellus. The results of this inquiry are shown in figure 9. In this figure the broken line gives the sur- vival time of 7 small isopods, averaging 4.3 mm. long, in N /400 concentration. This is to be compared with the unbroken line which gives the resistance of 5 isopods that averaged 7.1 mm. long. The temperature in both cases was 24°C. The vertical lines give averages and the arrows show the extent of the prob- able error which here is less than a fourth of the difference. The dotted line gives the resistance of 10 isopods 5.6 mm. long in N/500 solution whose temperature had been raised 6 to 8°. This is to be compared with the dot-dash line which gives the survival time of the same number of isopods 5.8 mm. long in the same solution strength whose temperature had been lowered 16°. The dash-three-dot-dash line represents the survival time of 10 isopods which were stimulated by shaking. The average length of these animals was 5.8 mm.; the temperature 24°; the solution strength N/500. Shaking greatly hastens the loss of equilibrium but after the isopods quit moving, the effect is almost nil. This curve is to be compared with the cross barred line which shows the resistance of 20 isopods under conditions identical with the last but which were not stimulated. The probable error of the averages in this connection is about the same as the difference and taken alone would be meaningless but it supports the other curves at least to the extent that the difference is in the same direction. We have here good evidence that young isopods and those at a higher temperature have a shorter survival time in cyanide than older animals or those at a lower temperature, and in- dications that stimulation by shaking also increases the sus- ceptibility. If the cyanide is a measure of the metabolic rate of Asellus these results are logical. Effect of potassium chloride on resistance to sodium cyanide Since potassium chloride was the most efficient reagent found for increasing the positiveness of the rheotactic response the effect of this salt on susceptibility to sodium cyanide would CONTROL OF RHEOTAXIS IN ASELLUS 189 give the best evidence obtainable as to whether or not a change in the metabolic state accompanied the increase in positive- ness. The extent of rheotactic change is shown in part I of table 5. Of the 49 isopods tested only one had its positiveness decreased by the 7-28 minute treatment with N/10 solution. The average rheotactic response was increased from 10 to 74 per cent positive. Comparison of the survival time with that shown in part II gives the effect of the treatment with potas- sium chloride upon the susceptibility to N/400 sodium cyanide. The resistance of 19 highly negative, untreated isopods was 6.1 (+0.3) hours and of the 49 untreated isopods 4.12 (+0.2) hours. The difference is approximately two hours which is four times the probable error and therefore statistically significant. These results are also graphically shown in figure 10 in which the solid line gives the resistance of the 49 isopods after being made positive with potassium chloride and the broken line gives that of the 19 untreated animals. The height above the base line gives the number of isopods dying each hour. The vertical lines again show averages and the double pointed arrows give the extent of the probable error. Effect of caleiwm chloride upon resistance to cyanide and upon carbon dioxide production Opposed to potassium chloride, calcium chloride was the most effective depressing salt found. Its effect on the metabolic rate may be judged both by its effect on resistance to cyanide and on carbon dioxide production. Seventy-one highly positive isopods were treated with cal- cium chloride long enough to cause a reversal in rheotaxis and then gave a resistance to N/400 sodium cyanide of 6 hours 34 minutes (+ 4 min.) while 27 highly positive, untreated isopods gave an average resistance to the same strength of cyanide of 5 hours 32 minutes (+13 min.) The difference here is 3.7 times the probable error. It was not thought necessary to carry this series further because of work done with Dr. Tashiro (Allee 190 W. C. ALLEE TABLE 5 Showing the effect of N/10 KCl upon rheotaxis and survival time in N/400 NaCN I. Isopods treated with KCl until highly positive to a water current. RHEOTAXIS AT START + ee eee bo bd bo Or Or Ot Or 20 lor) 20 60 80 20 60 20 40 80 ature 23 to 60°C. AFTER TREATMENT 20 50 20 80 20 34 20 bo or TIME IN KCl IN| sizE IN MIN- MM. a) (=) eS) (SS) (SS) Se eee Sey) (1 ea (SS) aS) () ) (Si ep SS) (ay SS) ©) (SS) SS) (S) coon oo 10.0 Temper- SURVIVAL TIME IN NaCN. CONTROL OF RHEOTAXIS IN ASELLUS 19] TABLE 5—Continued RHEOTAXIS AT START AFTER TREATMENT KCl Seats SURVIVAL + = a 35 - a oO seen) Te NaCN. 100 | 100 12.0| 6.0| 3:52 100 60 | 40 PAD) Me) 3 :52 100 100 PATO) ess 6:47 100 80 | 10 10 NOP le (Geo 3:30 40 40; 20 1) 1g 20 IPR OM 7a 3:17 100 60 | 20 20 ZO mIe to: 2:47 100 100 2A XY) |) ayes 3:47 100 60 | 40 15.0 5:02 100 710)" 20 10 25.0 13 :00 100 100 BSA) Caw) 5:30 Ave. 10 82 8 74) 15 9 2 13.8 | 6.4 |4:07 +12 II. Negative, untreated isopods. Temperature 17 to 23°C. See ee | ecg. | | a = a 100 0 t5 4:30 100 0 6.0 5:30 100 0 5.0 5:30 100 0 Hag 6:35 100 0 5.0 7:05 100 0 6.5 7:05 100 0 5.0 7:30 100 0 10.0 7:55 100 0 @0 10:05 100 0 8.0 10 :55 20 80 0 6.0 © 7:50 20 20 | 60 0 7.0 4:18 100 0 6.0 8:32 100 0 6.5 5:45 100 0 7.0 7:20 100 0 Cnet 3:02 100 0 6.5 5:43 100 0 7.0 4:18 100 0 9.0 5:18 Ave. 2 95 3 0 6.7 6:06 +18 “681 ‘d 90s sjivqjop 10g ‘anoy yoRe Susp o3vqueosed oy} MOYS SO}¥VUIPIO OY} PU SUNT} OAIZ ovSSTOSqY “(9Ul] WoYo1q) spodost pozvorzuN jo puv eprlxoTYyo unissejod yyIM yuourywery Aq (9UT] UsyOAquN) pas19AdI UOIJOVOL OIPORYOOYI ITY} peYy pey Yor spodost Jo spruvAd WINIPOS YOF/N OF 9OULASTSEL OY} Surmoyg Of ‘stq Seal =) bi is BRES=G2 ab4ARR B4nam aus ~eEe | | TAI ale teties PRSRS Caras Bawes BESS a ca saaee 192 ee] a Fs | BeeEae eee Pa aS2akhis CONTROL OF RHEOTAXIS IN ASELLUS 193 and Tashiro ’14) upon the effect of calcium chloride upon car- bon dioxide production in Asellus. A part of these results are republished here: In brief the experiments were as follows: Two isopods of approxi- mately the same size were tested for their relative rate of carbon diox- ide production in the Biometer. The isopod having the lower rate of carbon dioxide output was taken as a control and was again tested for the rheotactic reaction and then left in conditions to which it was acclimated while the other was treated. The second individual which had the higher rate of carbon dioxide production was placed in a 0.2 mol. solution of calcium chloride until the tendency to give a positive rheotactic reaction was markedly reduced. Then the rate of carbon dioxide production of the two was again tested in the Biometer. In both pairs tested the isopod with the higher rate of carbon dioxide production at the first test in the Biometer had also given the higher percentage of rheotactic responses, but after being treated with cal- cium chloride for 25 to 36 minutes it came to be less positive in its rheotactic reaction, and also gave less carbon dioxide and was less susceptible to potassium cyanide than the control individual. In other words the calcium chloride (0.2 mol.) decidedly decreased the rate of metabolism of the isopods and also reduced their tendency to give a positive rheotactic reaction. Effect of cane sugar on resistance to potassium cyanide Fifty-four isopods that gave an original average rheotactic response 33, 54, 9, 1 per cent positive, negative, indefinite, and zero were treated with M /2 cane sugar until they gave an average reaction of 18, 24, 52, 6. Forty-eight of these were immediately killed in N/1000 potassium cyanide and gave an average survival time of 8 hours 23 minutes (+31 min.). ' Eighty-six isopods that gave an average rheotactic response of 34, 57, 9 per cent respec- tively resisted the same strength of cyanide for an average of 5 hours (+7 min.). This is a difference of 3 hours 23 minutes which is over five times the probable error and undoubtedly ' significant. Since the cane sugar probably acted by removing water (Peters ’04 for Stentor) we have here excellent evidence that such conditions depress the rate of oxidations in Asellus which corresponds to the results reported by Riddle (’14) for the pigeon’s egg. TABLE 6 Showing the effect of calcium chloride upon carbon dioxide production and rheotaxis in isopods. poses. The survival time in potassium cyanide is added for comparative pur- The isopods were first tested for rheotaxis, then two of approximately the same size were taken for determination of their carbon dioxide output in the bi- ometer. The one of these that gave the least carbon dioxide was taken as a control, its rheotactic reaction was again tested and it was allowed to stand in water to which it was accustomed while the other was treated. The second isopod, the one giving the most carbon dioxide, was placed in a 0.2 mol. solution of calcium chloride until the positive rheotactic tendency was markedly decreased. Immediately afterward the carbon dioxide production of the two was again compared in the biometer. ISOPOD NO. 30 Rheotaxis test, 11:55 a.m. Temp. 20 50% +, 50% —; Efficiency, 2.1 Tested in Biometer 1:47-2:00 p.m. Temp. 23.5 Less CO2 than No. 169 Rheotaxis test 2:00 p.m. 70%+, 20%—, 10%, Efficiency, 2.25 Tested in Biometer 3:44-3:57 p.m. More CO: than No. 169 Survival time in 0.001 Mol. 2 hours, 20 minutes 3, 4.5 mm. long KCN ISOPOD NO. 171 Rheotaxis test 12:25 p.m. Temp. 20 10%+, 90%—; Efficiency, 2.4 Tested in Biometer 2:49-3:05 p.m. Little CO: given off Less CO: than No. 84 Rheotaxis tested 3:50 p.m. 60%+, 40%—; Efficiency, 2.0 Tested in Biometer 4:35-5:12 P.m. More CO; than No. 84 Survival time in 0.001 Mol. 1 hour, 35 minutes o’, 5.5 mm. long KCN ISOPOD NO. 169 Rheotaxis test, 12:25 p.m. Temp. 20 90%+, 10%—; Efficiency, 2.1 Tested in Biometer 1:47-2:00 p.m. Temp. 23.5 More CO, than No. 30 Put in 0.2 Mol. CaCl, 2:05 p.m. Rheotaxis test 2:07 P.M. 80%-+, 20%—; Efficiency, 1.6 Rheotaxis test 2:27 p.m. 40%+, 20%, 40%; Efficiency, .95 Taken from CaCl, 3:43 p.m. In CaCl, 36 minutes Tested in Biometer 3:44-3:57 P.M. Less CO: than No. 30 Survival time in 0.001 Mol. 3 hours, 10 minutes 3’, 5.0 mm. long KCN ISOPOD NO. 84 Rheotaxis test 12:00 Mm. Temp. 20 30%+, 60%—, 10%; Efficiency, 2.6 Tested in Biometer 2:49-3:05 p.m. Little CO2. given off More CO, than No. 171 Put in 0.2 Mol. CaCl, 4:02 p.m. Rheotaxis tested 4:07 P.M. 10%+, 40%, 50%; Efficiency, 0.9 Taken from CaCl, 4:27 p.m. In CaCl. 25 minutes Tested in Biometer 4:35-5:12 p.m. Less CO, than No. 171 Survival time in 0.001 Mol. 2 hours, 55 minutes o, 6.0 mm. long KCN CONTROL OF RHEOTAXIS IN ASELLUS 195 Since the most efficient stimulating and depressing salts found affect the metabolic rate of the isopods as measured by their resistance to the cyanides and (calcium) by carbon dioxide production the results of these experiments support earlier work on this subject which demonstrated that Asellus with a high rate of positive rheotaxis have a relatively high rate of metabolism and those with a low degree of positiveness tend to have a low rate of metabolism. Lillie (09) gives evidence’ to show that the primary action of pure sodium and potassium ‘chloride solutions is to increase, and that of magnesium and calcium chlorides, to decrease permeability in Arenicola larvae and that these stimulate and depress, respectively, the muscular activity of these animals. He concludes that this increase in permeability is in itself sufficient to account for the liberation of energy which is the essential consequence of stimulation. Whether or not the relationship between these salts and rheo- taxis can be fully explained on such a basis is a matter for fur- ther experimentation. V. SUMMARY OF EXPERIMENTAL RESULTS The chlorine salts of the alkali metals affect the rheotactic reaction of Asellus communis in such a way as to suggest that there is a relation between the chemical activity of these cat- ions and their effect on rheotaxis of isopods. Potassium is the most effective in increasing the positiveness of the reaction, with rubidium a close second, p. 167. The relative toxicity of these cations does not run parallel with their stimulating power but resembles the relative favor- ableness in preserving the activity of frog nerves and muscles pz 170. The anions of the most stimulating cation, potassium, affect the rheotactic reaction but their effectiveness does not run parallel with their chemical activity and the relative toxicity, while similar, is not exactly the same as the stimulating power, pe 172: Any chemical in the concentrations used will cause a decrease in the positive rheotactic reaction, but the chlorine salts of THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 21, NO. 2 196 W.-C, ALLER calcium and strontium cause this decrease usually without a preliminary stimulation. Magnesium chloride while in the main similar in action often causes preliminary stimulation and barium chloride is still more stimulating, resembling the alkali metals in its effect, p. 173. In the cations such as potassium which are highly stimulat- ing the depression is a toxic effect while in depressing cations ‘as calcium, rheotaxis is depressed long before toxicity symptoms appear, p. 174. There is a marked antagonism between the effect of potassium - and calcium chlorides and a less marked one between the chlorides of sodium and magnesium, p. 175. Both acids (H ions) and alkalies (OH ions) in the concen- trations used generally decrease the percentage of positive re- sponses given, p. 177. Once distilled water gave some evidence of causing Asellus to become more positive in their rheotactic reaction, though pure water was quite toxic, p. 181. Cane sugar decreased the positive rheotactic reaction probably by extracting water and the action of the once distilled water may be due to water intake, p. 183. The results obtained with salts are not osmotic effects be- cause equimolecular solutions of different salts with approxi- mately the same osmotic pressure may have opposite effects and because cane sugar in M/2 solution was a less effective depressant than N/10 calcium chloride although its osmotic pressure is over three times as great, p. 185. Susceptibility to sodium cyanide N/400 or N/500 measures the rate of metabolism of Asellus probably by limiting the oxy- gen consumption, p. 188. Measured in this way, potassium chloride when increasing positive rheotaxis also increases the rate of isopod metabolism, and calcium chloride and cane sugar decrease both positive rheotaxis and the metabolic rate. Calcium chloride also de- creases the carbon dioxide output, p. 189. CONTROL OF RHEOTAXIS IN ASELLUS 197 LITERATURE CITED ALLEE, W. C. 1912 An experimental analysis of the relation between physio- logical states and rheotaxis in Isopoda. Jour. Exp. Zodl., vol. 13, pp. 269-344. 1913 Further studies on physiological states and rheotaxis in Isopoda. Jour. Exp. Zo6él., vol. 15, pp. 257-295. 1914 Certain relations between rheotaxis in isopods and their sur- vival time in potassium cyanide. Jour. Exp. Zodl., vol. 16, pp. 397-412 AuLEE, W. C. anp TasHrro, Sutro 1914 Some relations between rheotaxis and the rate of carbon dioxide production of isopods. Jour. An. Beh., vol. 4, pp. 202-214. Boun, Grorces 1912 Les variations de la sensibilité en relation avec les vari- ations de l’etat chimique interne. C. R. Acad. Sc., Paris, T.154, pp. 388-391. Bropsky, 1908 Dissertation unter Hober, Zurich. Fide Héber, 714, s. 511. Buuiot, G. 1904 On the toxicity of distilled water for the fresh water Gam- marus. Suppression of this toxicity by the addition of small quanti- ties of sodium chloride. Univ. of Calif. Publ. in Physiol., vol. 1, pp. 199-217. Chemiker Kalendar 1913 Cuttp, C. M. 1914 Axial gradients in the early development of the starfish. Am. Jour. Physiol., vol. 37, pp. 203-219. 1915 Senesence and rejuvenesence. Chicago, 481 pp. GarrREY, W. E. 1915 Some eryoscopic and osmoticdata. Biol. Bull., vol. 28, pp. 77-86. GeppPrert, J. 1889 Ueber das Wesen der Blausiurevergiftung. Zeitschr. f. klin. Med. Bd. 15, s. 208-247; 307-369. Héser, Rupotew 1914 Physikalische Chemie der Zelle und der Gewebe. Leip- zig und Berlin, 808 pp. Hyman, Lispiz H. 1916 An analysis of the process of regeneration in certain microdrilous oligochaetes. Jour. Exp. Zodl., vol. 20, pp. 99-163. 1916a The effect of potassium cyanide on oxygen consumption in certain sponges. Am. Jour. Physiol., vol. 40, pp. 238-248. Kanpa, Saxyo 1914 The reversibility of the geotropism of Arenicola larvae by means of salts. Am. Jour. Physiol., vol. 35, pp. 162-173. Lituir, R. 8. 1909 On the connection between changes of permeability and stimulation and on the significance of changes in permeability to car- bon dioxide. Am. Jour. Physiol., vol. 24, pp. 14-44. 1909 a The relation of ions to contractile processes. IV. The in- fluence of various electrolytes in restoring muscular contractility after its loss in solutions of cane sugar and magnesium chloride. Am. Jour. Physiol., vol. 24, pp. 459-492. Lors, JaAcqurs 1903 Relative toxicity of distilled water, sugar solutions, and solutions of the various constituents of sea water for marine animals. Univ. of Calif. Publ. in Physiol., vol. 1, pp. 55-69. 1906 Ueber die Hemmung der Toxischen Wirkung hypertonischer Lésungen auf Seeiglei durch Sauerstoffmangel und Cyankalium. Archiv. f. ges. Physiol., Bd. 113, s. 487-511. 198 WwW. C. ALLER Logs, J. AnD Lewis, W. H. 1902 On the prolongation of life inthe unfertilized eggs of sea urchins by potassium cyanide. Am. Jour. Physiol., «vol. 6, pp. 305-317. Lores, J. anD WastEeNEYs, H. 1910 Warum hemmt Natriumcyanid die Gift- wirkung einer Chlornatriumlésung fur das Seeigelei. Biochem. Zeitschr. Bd. 28, s. 340. 1913 Is narcosis due to asphyxiation? Jour. Biol. Chem., vol. 14, pp. 517-523. 1913 a The influence of bases upon the rate of oxidations in fertilized eggs. Jour. Biol. Chem., vol. 14, pp. 459-464. 1915 Further experiments on the relative effect of weak and strong bases on the rate of oxidations in the egg of the sea urchin. Jour. Biol. Chem., vol. 21, pp. 153-158. LorvennartT, A. S. 1906 Ueber die Beschleunigung gewisser Oxydations- reactionen durch Blausiure. Ber. deutsch. chem. Gesellsch., Bd. 39, s. 130-133. Martuews, A. P. 1904 Nature of chemical and electrical stimulation. Am. Jour. Physiol., vol. 11, pp. 455-496. 1907 The cause of the pharmocological action of ammonium salts. Am. Jour. Physiol., vol. 18, pp. 58-63. Meyernorr, 1911 Die Atmung der Seeigeleier (Strongylocentrotus lividus) in reinen Chlornatriumlésungen. Biochem. Zeit., Bd. 33, s. 294-302. Morsg, H. N., Frazer, J. C. W. anp Hottanp, W. W. 1907 The osmotic pres- sure of cane sugar solutions in the vicinity of the freezing point of water. Am. Chem. Jour., vol. 37, pp. 424-467. Morse, H. N. anp Mears, B. 1908 The osmotic pressure of cane sugar at 15°. Am. Chem. Jour., vol. 40, pp. 194-213. Morse, H. N. anp Hotitanp, W. W. 1909 The osmotic pressure of cane sugar at 20°. Am. Chem. Jour., vol. 41, pp. 257-276. OstEeRHAUT, W. J. V. 1914 Effect of alkalion permeability. Jour. Biol. Chem., vol. 19, pp. 335-345. 1914a Antagonism between acids and salts. Jour. Biol. Chem., vol. 19, pp. 517-520. Peters, A. W. 1904 Metabolism and division in Protozoa. Proc. Am. Acad. Arts and Sci., vol. 39, pp. 441-514. RippLE, Oscar 1914 The determination of sex and its experimental control. Bull. Am. Acad. Med., vol. 15, no. 5, 20 pp. RINGER, SipNeEy 1886 Further experiments regarding the influence of small quantities of lime, potassium and other salts on muscular tissue. Jour. Physiol., vol. 7, pp. 291-308. WarsurG, O. 1910 Ueber die Oxydationen in lebenden Zellen nach Versuchen am Seeiglei. Zeitschr. f. Physiol. Chem., Bd. 66, s. 305. THE EFFECT OF RADIUM RADIATIONS ON THE RATE OF CELL DIVISION CHARLES PACKARD From the Zoélogical Laboratory, Columbia University The radiations of radium produce two distinct types of effects on living matter, depending on the intensity of the radiation. When the exposure is intense or prolonged many abnormalities result. In the unfertilized or freshly fertilized egg they produce a marked cytolysis, in which the water holding power of proto- plasm is markedly increased; with this condition is usually associated a derangement in the mechanism of mitosis. As a result, the formation of polar bodies is interfered with or sup- pressed, wholly or partially, or multipolar figures are produced. The embryo is always abnormal in the case of Nereis and Ar- bacia. Oscar Hertwig found that if the unfertilized frog egg is radiated up to a certain point the resulting embryo after insemination with normal sperm is abnormal, but if the radia- tion is very intense development is haploid, the egg nucleus having been rendered incapable of playing its usual part in fertiliza- tion. This phase of the subject has lead to no significant results thus far, perhaps on account of the difficulty of interpreting the very diverse effects. The disturbance of the chromatin in division has lead Gunther Hertwig to draw interesting con- clusions as to the functions of chromatin in heredity, but this is aside from the general problem of the effect of radium. In his experiments he employed radium as an effective means of injur- ing the sperm, but his conclusions would probably have been the same had he used any other means of disturbing the chromo- tin without interfering seriously with the protoplasmic activi- ties of the sperm. 199 - 200 .CHARLES PACKARD When the exposure to radium is brief, or the amount used is small, an entirely different kind of phenomenon appears. This consists in a change in the division rate of cells, which is not fol- lowed by any apparent abnormality. Many observers have noted this effect but have made little attempt to analyze the problem. In most instances they have described a retardation in the rate of division but occasionally they have noted an acceleration. Richards (’14) studied the effect of X rays on the cleavage of the eggs of the snail Planorbis. As X rays are similar to the gamma rays of radium his results have a direct bearing on the problem discussed in this paper. He found that brief exposures produced an acceleration in the rate of cleavage, but this effect was only temporary since after the first cleavage the treated eggs were retarded. The greatest effect was produced when exposures were made during the metaphase. During the resting stage there was very little response one way or the other. The experiments described in this paper were carried on to find out more exactly the conditions under which acceleration or retardation may be produced by the action of radium radia- tions on the fertilized eggs of Arbacia. METHODS In conducting experiments on the rate of cell division it is of prime importance to keep the temperature of the sea water constant throughout. To this end I carried out each experi- ment in the following manner. Syracuse watch glasses were kept in running sea water for thirty minutes to bring them to an even temperature. They were then quickly dried without being touched by the hands, and filled with a measured amount of sea water. The Arbacia eggs, which had been freshly fer- tilized were now added in measured quantities. It was found necessary to use only a few eggs at a time for a very little over- crowding itself changes the rate of division and produces ab- normalities. One watchglass was placed under the radium tube EFFECT OF RADIUM ON CELL DIVISION 201 which contained 50 mg. of pure bromide, equivalent to 23.4 mg. of element. The control eggs were placed in a watchglass beside the others but separated from them by a thick sheet of lead. Frequent tests showed that there was no appreciable difference in the temperature of the two lots of eggs during the entire experiment. This point indicated that the radiations from this amount of radium are not sufficient to produce any marked temperature change. The radium tube was held in a device so arranged that the tube could be held at any desired distance above the eggs. As there was no screen of any kind interposed between the radium and the eggs it is assumed that all of the rays emanating from the tube, namely the beta and gamma rays, were able to reach the eggs. After exposure the eggs, with the controls, were observed almost continuously until after the first cleavage had been completed. The amount of acceleration or retard is estimated by compar- ing the time elapsing between fertilization and cleavage in the controls and the radiated eggs. Thus in some experiments the controls began to divide 60 minutes after insemination, and the radiated eggs, in 50 minutes. The difference of 10 minutes is about 16 per cent of the 60 minutes:elapsing before cleavage occurred in the controls. Exposures were made during five periods of the cell’s activity, namely, between 5 and 25 minutes after insemination when the germ nuclei are approaching each other; second, during the prophases which occur from about 25 to 35 minutes after in- semination; third, during the metaphase which appears in about 35 to 50 minutes; fourth, during the telophases (50 to 65 minutes after insemination); and finally during the resting stages fol- lowing the first cleavage. Naturally the temperature of the air hastened or retarded this rate, so that in some instances the exposures had to be shorter than the limits mentioned. In each experiment the distance between the radium and the eggs was varied from 2 inch to 2% inches. 202 CHARLES PACKARD OBSERVATIONS The experiments related below are typical of a large number, and only those are recorded which have been verified by re- peated trials. 1. Exposure during the period in which the germ nuclei are approaching and fusing with each other. Five to twenty-five minutes after insemination. Radium placed 3 inch above eggs Length of exposure Result RTENUE ULES 3.3.2 0:2 ERR G rks AOE ES ee Slight acceleration HOMIES Salts cs 6 Seer RTo RTS ee ORCA, Melee erates 5 per cent acceleration SR MMITIUTES :cka.5 ees Soe ce Ce aah cee tie ner ees Slight acceleration A eminNubeses scsi Chee eae Md dhe) Scr ee eee No acceleration Radium placed 2 inch above eggs RO miTiutes,:..)ektiaeric eee cen eA Mime oo cee sae Me pare Slight acceleration TOMINUCES ze eies Re oe ee ere Slight acceleration AO MINUtes Hee Gelee eee re eine ane We tiae oe Merete ee, oe No acceleration If the radium is placed 14 or 2} inches above the eggs there is little evidence of any change. These results were constant and clear cut, even when the amount of change in division rate was small. Under these conditions it is clear that unless the exposure is intense there is little or no response. ‘These results were an- ticipated for I have found in previous experiments that freshly fertilized or unfertilized eggs are not very susceptible even to intense exposures. However, if the radiation is prolonged the whole organization of the egg undergoes a profound change, and subsequent development is very abnormal. Inasmuch as I was not trying to produce abnormalities, I did not use very intense radiations. 2. Exposure made during the prophases of mitosis. Twenty to thirty-five minutes after insemination. Radium placed 3 inch above eggs Length of exposure Resull Eayeraa) WAY NSS hy RAE eS omer cb the etches Gel a Grocer See EES Slight acceleration LOfmimUttesior:!.. soe eee enero ee ercre ks cries 10 per cent retard IG Si TVEALLUNS: eee ete Ooo ets oo Note ORR ama oe 15 per cent retard EFFECT OF RADIUM ON CELL DIVISION 203 Radium placed ? inch above eggs NO) Tea NOUN ES) so oo cece Hie OO SO So a eaaeEs Ain san siya 5 per cent retard US TiN -.9 ats pos bes > SOD REEIES IS Seco o oeOe 10 per cent retard Radium placed 13 inches above eggs HUI VISSIES RENE Se mae 5 UOeRen Phone» SNe ee 2A pee ae a 5 per cent acceleration A SPMVUMUULC HRS Perso rte Si ee sete Ant 20o 6a wc of oe steer Slight acceleration At a distance of 24 inches the radiations produced no decisive effects. From these experiments it appears that during the prophases the egg is more susceptible than during the preceding period. An exposure which, during the first period produces an accelera- tion, here produces a retard, indicating that the optimum amount of radiation has been exceeded. 3. Exposure made during the metaphase . Thirty-five to fifty minutes after insemination. Radium placed 3 inch above eggs Length of exposure Result HUTOMUMILD CSA ATES OS oe, cS e Le eon ee have 5 per cent retard LCUSEECIS ESOS Cie oil ee hey oe 9 a rr 15 per cent retard (Longer exposures produced some abnormalities) Radium placed 3 inch above eggs CE GERTIORIN AEG set eee Se RR A ee ey Oe 5 Shght retard MID PERINRVOLUES Pia Peas ht) ANA OE MEL RIE i a 10 per cent retard Radium placed 1} inches above eggs aR ISIN TELL LER Oe ise y Sey ta tyke te RSA ct NORE Nephi ec odd) ded oO No change MEPCMIRE ESS Sa cose, tlh Re Nene. ie, Macey eel ahdie 6x, s 5 per cent acceleration TGS STETISTI A No Goan ee a 4 te Se dite! ane yee 10 per cent acceleration At 23 inches exposures up to 15 minutes produced no clear cut effects. It is evident that the results noted here are of the same order as those seen in the preceding experiment, but they are of greater magnitude. It may be said therefore that eggs in this condition are more susceptible than before. 4. Exposures made during the telophases. Fifty to sixty- five minutes after insemination. 204 CHARLES PACKARD Radium placed 3 inch above eggs Length of exposure ; Result ASTON TSS Rea SE ERIE Say es oo Su oko aaa No change 10 ‘minutes:.%.%.)..... en 5 5 rn i 5 per cent retard NSW e0 TSU. a ae eres ae ore cre Sy o.g oberaieet a ois ote 5 per cent retard Radium placed 3 inch above eggs POUL S ae os oss co keen Gan eo ee eee ae Slight acceleration 155 en re Shane erchons ater encmec tare A olp weer 5 per cent acceleration At the greater distances there was no clear evidence of any change. These results indicate that eggs during the telophase are not as responsive as they are during the metaphase, but are more like the eggs in the prophase of mitosis. 5. Exposures made during the resting stage of the duis were followed by no decisive changes unless the radiation was intense and prolonged, in which case the eggs developed abnor- mally, as would be expected. The responsiveness of the egg has now reached its minimum. This result agrees very well with the results of exposures made on tissues in which little cell division is taking place. Unless the cells are dividing rapidly radiation produces little result except when it is very intense. In all of these experiments it is assumed that each egg re- ceives an equal amount of radiation. Probably this is not strictly true, especially when the distance through which the radiations pass is small, for the distance from the tube tothe eggs directly under it was somewhat less than the distance to the eggs further from the center of the watch glass. The fact that the source of radiation is not a single point but a tube 1 cm. long would also make a difference. However the inequality in the amount of radiation received by the eggs is very small and is constant for each experiment, so the results are not obscured. The amount of radiation which falls on the eggs decreases inversely with the square of the distance through which it passes. Thus if we consider 2 inch as a unit distance, the energy recevied at 3 inch is one-fourth of that received at unit distance. At 15 inches it is one-sixteenth, and at 2+ inches, one thirty-sixth of the unit amount. Theoretically we should expect the re- EFFECT OF RADIUM ON CELL DIVISION 205 sults to vary in these ratios, but under the conditions of the experiment it is difficult to estimate in these terms. It was found impossible to expose eggs at 2} inches for thirty-six times as long as at 2? inch for during that prolonged period the eggs pass through all of their phases of division at least twice. A further difficulty is found in the fact that the beta rays, which are very active in producing changes in protoplasm, as I have shown (Packard 715), are easily stopped by air so that at a distance of two inches or more very few of them reach the eggs. A few experiments were made in which the eggs were cooled in a freezing mixture so as to retard their rate of development, and the radium was applied for a longer time. But the results were not satisfactory as it was found difficult to hold the con- trol eggs and the radiated eggs at exactly the same stage of development. It was also impossible to keep both lots at exactly the same temperatures throughout the experiment. Although some of the trials tended to confirm the expectations, the general results were not conclusive or reliable. : DISCUSSION The radiations of radium produce effects in matter only when they are absorbed wholly or in part by it. The beta rays, be- cause of their relatively low velocity are quickly stopped by many substances, the particles colliding with the molecules of the substance, and in many instances producing a marked ionisa- tion of the molecules. As these rays are not homogeneous but are made up of particles projected with varying velocities, it follows that the slower particles are stopped more quickly than the more rapid ones and produce greater chemical effects. The fact that they affect protoplasm more vigorously than the ’ rapid rays was shown in a recent paper (Packard 715). Whether all of the effects noted in those experiments were due to an ionisa- tion of protoplasm is a point yet to be determined. The gamma rays are exceedingly penetrating and are stopped completely only by thick sheets of the heavy metals. Some of them however are stopped by protoplasm, especially if the 206 CHARLES PACKARD layer of living tissue be of considerable thickness. When this stopping occurs, due to the violent encounters of the rays with the protoplasmic molecules, the gamma rays disappear and their energy is converted into that of beta particles which travel on in the direction of the original gamma ray. These secondary beta rays produce effects similar to those produced by the pri- mary beta rays. The great penetrating power of the gamma radiations accounts for the fact that few of them are thus changed into beta rays in protoplasm and hence produce few changes in living matter unless an enormous number is allowed to act on it. X rays which are similar to these radiations produce their characteristic effects by means of the secondary beta rays which they generate. The effects produced by these two types of radiations are therefore comparable. The amount of radiation which is stopped by matter depends on the density of the elements composing it. Roughly, the absorptive power varies with the square root of the atomic weight of the elements, and therefore follows closely the group- ing of the elements in the periodic table. From this it is clear that the light elements composing protoplasm, such asscarbon, oxygen, hydrogen, and nitrogen have relatively low powers of absorption. The absorption of radiations by compounds is equal to the coefficients of absorption of the elements composing them. ‘That is, the stopping power of a compound depends on the number and atomic weight of its constituent atoms. This rule applies to all of the compounds studied thus far. Borodowsky showed that the absorption of beta rays by liquids follows accurately an additive law, and does not depend on the concentration or ‘chemical dissociation of the liquids. The amount of radiation which an organic liquid can absorb can be deduced from a knowledge of its constituent elements. From this it follows that the amount of absorption is not influenced by the molec- ular structure of substances. Applying these facts to the protoplasm of the Arbacia egg, it is evident that the protoplasm will always stop a definite amount of the radiation provided that the chemical make up of the eggs remains unchanged. It EFFECT OF RADIUM ON CELL DIVISION 207 is also evident that in these experiments no profound change took place which involved a dropping out of any of the elements present in the protoplasm at the beginning of the experiment, since the eggs developed perfectly normally after the treatment. This fact is important in the explanation of the results here presented, for at first sight they seem to indicate that the re- verse is true, namely, that at certain periods of their activity the eggs are more absorptive to the rays than at others. In- asmuch as this cannot be true we must look to some physiologi- eal changes in the egg which render it more susceptible at one time than at another. When protoplasm is intensely radiated enough energy is absorbed to bring about a marked physical and chemical change. This may consist in partial ionisation of the mole- cules, or in the breaking down of compounds present. Schwarz has shown that when an egg, rich in yolk, is intensely radiated, the lecithin is decomposed into cholin which acts as a poison. It is doubtful whether this explanation can suffice to account for some of the effects reported, sincenot all of the cells which areinjured contain lecithin in large amounts. This is particularly true in the case of tumor cells which contain, according to recent analyses, no more lecithin than the surrounding normal cells which are uninjured by the radiation, although the tumor cells are obviously affected. Other effects of severe radiation, such as cytolysis, have al- ready been mentioned. When the unfertilized Nereis egg is radiated the cell wall is greatly weakened and the vitelline mem- brane altered to such an extent that it allows many sperms to enter. Prolonged exposures of the sperm cause in the chroma- tin a change which shows itself in the fragmentation of the sperm nucleus after it has entered the egg (Hertwig 712). Gunther Hertwig has suggested that radium acts directly on the chromatin. That chromatin is more affected by equal amounts of radiation is undoubtedly true, but it is also true that protoplasm is actually affected. I have elsewhere (Packard ’14) pointed out some objections to this theory. In terms of absorptive powers it assumes that chromatin is better able to 208 CHARLES PACKARD stop the radiations than protoplasm. This is probably true, since the former substance contains more phosphorus and iron than the latter. But the difference in composition is not so great as to account for the striking differences in response. Furthermore, the changes in the apparent susceptibility of chromatin during the different phases of mitosis cannot be ex- plained by the statement that its chemical constitution has al- tered so as to render it more or less absorptive. The time elaps- ing between the prophase and the metaphase is so brief that we cannot imagine any profound chemical change, involving a dropping out of elements, to occur. It is obvious therefore that to account for the observed phenomena we cannot assume that the chromatin is more able to stop the rays, or that it varies in its ability to stop them, at different periods of its activity. It is necessary to take into account the physiological processes occurring during mitosis. During the periods in which the radium was applied to the eggs two phenomena which are of interest in this connection occur. In the first place the chromatin, which, during the rest- ing stage, or the stage previous to the fusion ofthe germ nuclei, is diffuse and semi liquid, becomes condensed and gelatinous in the metaphase. After this time it gradually becomes more and more diffuse up to the period of the resting stage. On ac- count of this increase in its density we might suppose that it becomes more absorptive to the rays. But a consideration of other experiments shows that this cannot be true. The sperm, in which chromatin is condensed to the greatest degree, is far more resistant to radiations than is the egg in which the chroma- tin is diffuse. I have exposed Nereis sperm for five hours yet at the end of that time they were not only actively motile but were able to bring about normal fertilization of the fresh egg. The fact that in the egg and the sperm there is an equal amount of chromatin, yet the egg is much more easily affected than the sperm, argues that something beside chromatin is involved. The second phenomenon which occurs during this period is the breaking down of the nuclear wall with the liberation into the egg protoplasm of the nuclear sap. This is followed by a EFFECT OF RADIUM ON CELL DIVISION 209 marked increase in the rate of oxidation taking place in the egg. This increase is due to the increased activity of the oxidative enzymes, such as nuclease. The continuance of the process of mitosis depends on these enzymes, for if oxygen is withdrawn cell division stops. Mathews has suggested that Autolytic enzymes also evidently become active, either because they are set free from the nucleus, or because the nuclear materials activate, directly or indirectly, the inactive enzymes of the cytoplasm. Since during cell division these enzymes are set free and at the same time the chromatic elements are plainly losing substance, it is possible that these two facts should be correlated and that the conclusion drawn that in the resting condition of the nucleus enzymes of various kinds stick to, or combine with, the nucleic acid and are thus accumulated, made resistant, more stable and inert, and that dur- ing caryokinesis and possibly at other time also, they are split off from the acid, become free in the sap, enter the cytoplasm and rejuvenate the cell by digesting its accumulated colloidal material. If it is granted that the phenomena of mitosis are dependent on the activity of intracellular enzymes, then it is clear that if these enzymes can be stimulated or retarded in their activity the result will be an acceleration or a retardation in the rate of cell division. I believe that the radiations of radium are able to change the rate of enzyme action, and that this is at least a partial explanation of the results recorded in these experiments. I have already reviewed briefly the literature on the effect of radium on enzyme action. There is some lack of agreement in the results obtained by various observers. The radiations are stated to retard the digestive action of pepsin and trypsin, to accelerate it, and to have no effect whatever. Similar state- ments have been made regarding other enzymes. Richards (14) has shown how these contrary results may be explained. Using X rays on pepsin and diatase he found that a short radiation has the effect of accelerating their activity, while a longer radiation inhibits it. “Between these two strengths lies a point at which radiation is non effective.’”’ These results on extracted enzymes are exactly similar to the results given in this paper. A short radiation brought about a stimulation, while a longer one produced a retard. Between these two limits 210 CHARLES PACKARD there was a strength of radiation which produced no noticeable affect, i.e., the initial acceleration was overcome by a subse- quent retard. “These similarities strengthen the view that in living eggs the: changes in division rate are brought about by changes in the rate of enzyme action. Assuming this hypothesis to be the correct explanation, it remains to explain why an acceleration or retard ofthe enzyme action can account for the differences in response under similar exposures during different phases of mitosis. Mathews sug- gests that while the nuclear wall is intact the enzymes are re- stricted in their action, but that they are able to produce oxida- tions as soon as the nuclear wall breaks down, that is, just before the metaphase. Then a mild radiation, during the pro- phase, although it activates the enzymes, results in little change in the egg since they are unable to act to advantage. But when the nuclear wall disappears and, under normal condition, active oxidation takes place, a slight radiation activates them as before, but because they are able to react with the protoplasm more vigorously than before, their activation leads to a more marked acceleration. If the radiation is more intense a retard results, that is, the optimum radiation has been exceeded and the enzymes are injured. These results are analogous to those obtained by treating protoplasm with poisons. A small amount of CO: accelerates muscular action while a larger amount re- tards it. This hypothesis throws light on the well recognized fact that actively dividing cells are more susceptible than those which are not undergoing mitosis. Slow growing tumors are not susceptible (unless they are superficial and can be so intensely radiated that the protoplasm is injured) while rapidly proliferat- ing tumor cells are very susceptible. EFFECT OF RADIUM ON CELL DIVISION fast SUMMARY Arbacia eggs exposed to a brief but intense radiation during the period when the germ nuclei are approaching each other are accelerated in their rate of cell division. Less intense radiation produces less acceleration. Exposures made during the prophase result in an acceleration unless they are prolonged, when a retardation ensues. During the metaphase the same phenomena appear but to a greater degree. During the telophase the effects are much the same as in the prophase. Eggs exposed during the resting stage are not easily affected. The power of the protoplasm and chromatin to absorb the radiations does not change during these periods. The differences in the density of the chromatin during the different phases of mitosis do not affect its absorptive power. During the metaphase when the eggs are most responsive to radiations oxidations take place through the activity of enzymes. If these enzymes are accelerated or retarded the effect is to accelerate or retard the rate of cell division. Experiments indicate that radiations produce these effects on extracted enzymes. It may be inferred therefore that the endoenzymes are af- fected in the same way and that changes in the rate of cell division, following radiation, are due to the direct action of the radiations on them. THE JOURNAL OF EXPERIM«NTAL ZOOLOGY, VOL. 21 gNo. 2 212 CHARLES PACKARD LITERATURE Boropowsky, W. A. 1910 Absorption of beta rays from radium by solutions and liquids. Phil. Mag., vol. 19. Hertwic, G. 1912 Das Schicksal des mit Radium bestrahlten Spermachro- matin in Seeigelei. Arch. f. mikros. Anat., Bd. 77. Matuews, A. R. 1915 Physiological chemistry. PackarD;, C. 1914 The effect of radium radiations on the fertilization of Ne- reis. Jour. Exp. Zoél., vol. 16. 1915 The effects of beta and gamma rays of radium on protoplasm. Jour. Exp. Zoél., vol. 19. Ricuarps, A. 1914a The effect of X rays on the rate of cell division in the early cleavage of Planorbis. Biol. Bull., vol. 27. 1914 b The effect of X rays on the action of certain enzymes. Am. Jour. Physiol., vol. 35. RuTHERFORD, E. 1913 Radioactive substances and their radiations. Woop anp Prime 1915 The action of radium on transplanted tumors of ani- mals. Annals of Surgery, December. CHROMOSOME STUDIES ON THE DIPTERA II. THE PAIRED ASSOCIATION OF CHROMO- SOMES IN THE DIPTERA, AND ITS SIGNIFICANCE CHARLES W. METZ Station for Experimental Evolution, Cold Spring Harbor, N.Y. EIGHT PLATES CONTENTS | SEG UTOXGRT OE TNS UR Don Oe I ee oa a ne Ca a Rn ee 213 PEAOtia SAITO EMeLHOUS t+ aee, Se MRTee se a: te wee oh gues SRG ods 1 eS 217 Reality of chromosome pairing in the Diptera......................00 0000 221 Details of chromosome behavior during one cell-generation................. 226 Pairing in different tissues and during different stages in ontogeny......... 230 Difierent species and families’ compared’......4) 0.0.0 ok oe ene ed wl ee 231 ) USS Sao ie Se Sed Ne PS 245 Summary and conclusions............... BET me Re LA Nee ich: Sc NM cfu sla acc eae 258 TOMO CLAD iver peers Vere oar trees cent = Aes ReM Ry Chath 4), ons See Oh ocular ics ote 260 INTRODUCTION Attention was first called to the pairing of chromosomes in the Diptera by Miss N. M. Stevens during 1907 and 1908 in connection with studies upon the heterochromosomes of insects (Stevens ’07, ’08). Although primarily concerned with the heterochromosomes and maturation phenomena, Stevens never- theless found the paired association of chromosomes, in the nine species she studied, so conspicuous as to warrant the state- ment that, ‘perhaps the most interesting point in the whole study is the pairing of chromosomes in cells somewhat removed from the sphere of the reduction process. This was found to occur in the ovarian follicle cells, the spermatogonia and some embryonic cells. This is not an occasional phenomenon, but one which belongs to every oogonial and spermatogonial mitosis’’ 213 214 CHARLES W. METZ (Stevens ’08, p. 372). In a later paper on ““The chromosomes in the germ-cells of Culex’ (Stevens ’10, p. 215), correspond- ing phenomena called forth a similar statement to the effect that ‘perhaps the most interesting point in the history of the germ-cells of Culex is the fact that, as in the Muscidae, pairing or synapsis, occurs in connection with each spermatogonial and oogonial mitosis as well as in anticipation for maturation.”’ Although only able to study somatic. mitoses to a very limited extent, Stevens surmised that, “it may therefore be true that pairing of homologous chromosomes occurs in connection with each mitosis throughout the life history of these insects’’ (p. 215). Now this would be a very important point to establish, as Stevens realized, and she doubtless would have followed it up had it not been for her untimely death in 1912. Most un- fortunately, however, her work on the Diptera was stopped at its very beginning and many promising questions suggested by it have remained uninvestigated. Nothing further appeared on chromosomes of the Diptera until 1914 when three papers were published, one by the author on Drosophila chromosomes; the others on the chromosomes on Culex pipiens, one by Miss Taylor, and one by Lomen. Both of the latter took exception to Stevens’ conclusions that the chromosomes are paired in Culex and other Diptera, on the ground that the chromosome pairs which she described were — really only precociously split univalent chromosomes. ‘Their evidence on this point, however, is very inadequate, and their conclusions are surely erroneous (see pp. 244 and 245). The purpose of the present paper is to describe in some detail the phenomena involved in ‘chromosome pairing’ in the Diptera, and to consider their bearing on current theories re- specting the nature of the chromosomes and their role in hered- ity. Because of their remarkably definite paired association the chromosomes of the Diptera are especially suitable for studies on the relationships between individual chromosomes and on the qualitative characteristics of chromosomes as indi- cated by their behavior, but as I have mentioned in a previous paper (Metz ’14) the technical difficulties involved in an extensive ASSOCIATION OF CHROMOSOMES IN DIPTERA 75) WS cytological study of these insects have caused them to be gener- ally avoided by cytologists. These difficulties, however, may very largely be overcome by care and persistence. Although certain principles must be observed in making preparations, the task is mainly one of securing and preparing enough speci- mens to get material in the proper stages and in sufficient quan- tity for study. No more difficulty is experienced in studying the nuclear phenomena, when the proper material is secured, than is the case in other insects; indeed the chromatic elements in the flies, when well prepared, appear with a brilliancy that is surpassed by very few objects.! The observations included here are concerned chiefly with chromosomal behavior in somatic cells and in germ-cells outside the sphere of maturation. These cells I shall briefly term ‘dip- loid’ cells, in distinction to oocytes and spermatocytes. Since all of the ‘diploid’ cells agree in respect to the phenomena dealt with, no confusion should arise from such a terminology. Phe- nomena associated with the maturation processes are considered only in so far as they bear directly upon those in‘diploid’ cells. Likewise the relationships between the chromosomes in dif- ferent species of flies are only briefly considered. I hope to re- turn to both of these questions in subsequent papers. In order to facilitate the treatment of the subject matter I will outline at once the main points considered in the paper, and will indicate in advance some of the conclusions attained. This may best be accomplished by taking account of certain genetic hypotheses which intimately involve the chromosomes and which have furnished the occasion for this investigation. These hypotheses are all contained in one comprehensive theory which has recently been brought into prominence by the rapid development of Mendelism. According to this theory the chromosomes are complex, accurately differentiated bodies whose organization and behavior are directly correlated with the genetic factors located in them. In any biparental organism, the diploid chromosome group is composed of two equivalent, 1 Except in maturation stages, which are often very unfavorable for study. 216 CHARLES W. METZ parental series (haploid groups), the individual members of which are respectively homologous and very similar to one another; and this involves the view that the chromosomes are present in bi-parental pairs (Montgomery, Sutton, Boveri). In addition it is supposed, in accordance with the conception of W. Roux that every chromosome contains a definite comple- ment of serially arranged genetic factors, each responsible for one or more inherited characters—the complement of factors being the same or similar in homologous chromosomes (members of a pair) but different in non-homologous chromosomes. In order to explain the perpetuation of this duplex germinal constitution a process (reduction division) is assumed to occur during matura- tion whereby the members of each pair are separated from one another and segregated in different germ-cells. From the cytological point of view the principal questions involved in this theory are as follows: 1) Can definite pairs of chromosomes really be distinguished? 2) If so, are the two members of a pair derived respectively from the male and fe- male parents? 3) Are the two members of a pair actually simi- lar to one another and qualitatively different from the others in respect to their physico-chemical constitution? 4) Do the two members of a pair actually separate from one another and go into different germ-cells during maturation? Three of these questions, together with one other of a more strictly cytological nature—the question of synapsis—form the central points about which most of the facts considered in the present study may be grouped. The nature of the material prevents the detailed consideration of each question in the order given, but so far as possible the evidence is presented in accord- ance with this scheme. The evidence bears especially upon the first question, to which a definite affirmative answer is given. With respect to the second question judgment should, perhaps, be suspended until the genetic continuity of the chromosomes is established, but if this continuity be assumed, this question is likewise answered in the affirmative. Regarding the third question only indirect evidence is furnished, but this evidence lends support to an affirmative answer here also. The fourth ASSOCIATION OF CHROMOSOMES IN DIPTERA PAL question is not directly involved in the present paper. In re- gard to the problem of synapsis the pairing phenomena in diploid cells, including final spermatogonia, clearly demonstrate that a side by side approximation of corresponding chromosomes (the essential feature of synapsis), actually does occur, although in this case it is not connected with maturation. Throughout the course of this study I have profited greatly by the counsel of Prof. E. B. Wilson, under whose direction the work was begun, and to whom I have become increasingly indebted for many kindnesses. MATERIAL AND METHODS My observations are based upon a study of the chromosomes in about eighty species of Diptera, representing thirty-five genera and fifteen families, as given in the following synopsis. ORTHORRAPHA Nemocera Culicidae Culex pipiens Linne. Brachycera Stratiomyidae Ptecticus trivittatus Say. Asilidae Asilus sericeus Say. Asilus lecythus Walk. Asilus notatus Wied. Asilus novae scotiae Macq. Asilus sadytes Walk. Ommatius marginellus Fabr. Leptogaster badius Loew. Eraz aestuans Linne. Erazx rufibarbis Macq. Dasyllis grossa Fabr. Dasyllis thoracica Fabr. Deromyia winthemi Wied. Bombyliidae Anthraz lateralis Say Anthrax sinuosa Wied. . Spogostylum simson Fabr. 218 CHARLES W. METZ CyYcLORRHAPHA Syrphidae Eristalis tenax Linne. Eristalis bastardi Macq. Eristalis aeneus Fabr. Eristalis meigent Wied. Volucella obesa Fabr. Mesogramma marginata Say. Toxmerus annulatus Loew. Acalypterae Micropezidae Calobata lasciva Fabr. Calobata nebulosa Loew. Sepsidae Piophila casei Linne Ortalidae Chaetopsis fulvifrons Macq. Camptoneura picta Fabr. Euzxesta stigmatius Loew. Euxesta anonae Fabr. Trypetidae Euaresta melanogaster Loew. Sapromyzidae Physegenua vittata Macq. Drosophilidae Drosophila.—27 species, many undescribed, see text. Cladochaeta nebulosa Coq. Scaptomyza adusta Loew. Scaptomyza graminum Fall. Sciomyzidae Neuroctena analis Fullen. Calypterae Anthomyidae Homalomya spp. Fucellia marina Macq. Ophyra leucostoma Wied. Muscidae Calliphora viridescens Desv. Calliphora erythrocephala Meig. Musca domestica Linne. Muscina stabulans Fall. Phormia regina Meig. Lucilia sericata Meig. Pseudopyrellia cornicina Fabr. Sarcophagidae Sarcophaga falculata Pand. Sarcophaga tuberosa serraceniae Riley. Sarcophaga dalmatina Schin. Sarcophaga bullata Park. Ravinia communis Park. Ravinia peniculata Park. ASSOCIATION OF CHROMOSOMES IN DIPTERA 219 Preparations have been made from gonads of both sexes, and somatic tissues of various kinds. Almost all of the latter represent embryonic stages, including eggs, larvae and pupae. The former have been taken from larvae, pupae or adults, or all three, depending upon the species. In some species all stages from early spermatogonia or oogonia to the formation of spermatozoa or eggs could be secured from adults, but in most cases it was necessary to use pupae or even larvae in order to obtain the desired stages. This is especially true of the family Drosophilidae. Jn all cases the gonads or small bits of tissue were dissected out of the specumens and then fixed; none of the specimens was fixed entire or partially intact. This fact is emphasized because it has been found that regardless of the fixative used, inferior results are obtained if tissues are fixed in situ. Dissections were usually made in Ringer’s solution except in the case of large specimens, when tissues were dissected out in the body fluid. Dissection in tap water was tried with fairly good results, but mitotic figures were less distinct after this treat- ment than after the use of Ringer’s solution.? For fixation Flemming’s strong solution was found most satisfactory and was most frequently employed. Objects were fixed from ten minutes to three hours depending upon their size. Longer treatment was tried, but with less satisfactory results due to frequent osmication and distortion. In addition to Flemming’s fluid various other fixatives were tried. Of these Hermann’s platino-aceto-osmic, and Gilson’s mercuric-nitric gave the best results (in many cases as favorable results as those obtained by the use of Flemming’s fluid), especially when it was desirable to differentiate the chromosomes without refer- ence to other nuclear structures. Sublimate acetic and Gilson- Carnoy’s acetic aleohol with sublimate were found fairly satis- factory for somatic tissues, but were inferior for the gonads. Bouin’s fluid (formol mixture) though frequently used, proved quite undesirable because of rts tendency to distort and produce 2 Dissection in tap water has been recommended by Doncaster (’14) for Abraxis. 220 j CHARLES W. METZ clumping of chromatic materials. Good fixation with this method was secured only in the case of eggs and occasional large pieces of somatic tissue where its penetrating power was advantageous. To supplement the permanent preparations, temporary ‘smears’ were frequently made with the use of Schneider’s Aceto-carmine (Stevens ’08 pp. 359-360) which proved to be a valuable agent for rapidly determining whether or not materials contained stages suitable for study. Frequently, one gonad would be prepared in this way and if found to be in the proper stage of development, its mate would be fixed in Flemming. The aceto- carmine preparations often gave very good figures of metaphase chromosome groups, but were found to be unreliable for detailed study because of the frequent distortion incident to swelling or mechanical disturbance. Consequently, most of the obser- vations included within this study are based upon fixed and sectioned material. Sections were made 5 uz thick, except in a very few cases where unusually large cells were found and a greater thickness was desirable. Nearly all slides were stained with Heidenhain’s Iron Haematoxylin, either alone or with a counter-stain of eosin or light green. Safranin was used fre- quently, but gave less distinct images, and failed to differentiate the finer chromatic elements as distinctly as did the haematoxylin. For the study of cleavage and early embryonic stages Droso- phila eggs were used. These were fixed at different periods, from a few minutes to a few hours, after being laid. It was found necessary in most cases to puncture the eggs, in order to facilitate the penetration of the fixative. When the eggs were punctured, successful fixation was secured with Flemming, Gilson’s mercuric-nitric, Bouin, sublimate acetic and Gilson- Carnoy, all of which were about equally favorable. A large proportion of the species included in this study have been reared in the laboratory for one or more generations, and the cytological material which they have furnished has largely been derived from pedigree cultures. In a few instances material was taken from jars of food which had been set out-of-doors, bat this was used only when the identification of larvae and ASSOCIATION OF CHROMOSOMES IN DIPTERA 221 pupae could be determined by the flies which subsequently hatched from the food. In no case is there any question as to the genus of the flies concerned and only in a few cases is the species doubtful. Such cases are mentioned in the text. Of tne families Asilidae, Bombyliidae, Syrphidae, Sapromyzi- dae, Ortalidae and Trypetidae, only adult flies were used. For the identification of the Sarcophagidae, the writer is indebted to Mr. R. R. Parker, for that of Culex pipiens to Mr. Fred. Knab, for that of the Drosophilidae to Dr. A. H. Sturte- vant,? and for all other identifications to Mr. C. W. Johnson who has very kindly examined a large series of specimens. REALITY OF CHROMOSOME PAIRING IN THE DIPTERA Since Stevens’ observations on chromosome pairing in the Diptera were more or less incidental to other features, and since her conclusions have been directly opposed by those of Taylor and of Lomen on Culex—material upon which part of Stevens’ work was based—it seems desirable first of all to ascertain definitely whether or not the so-called pairing phenomena in flies do in reality represent the association of independent chromo- somes. In the opinion of Taylor (’14) and of Lomen (’14) the duality of the chromatic elements in Culex (and hence by in- ference in the other Diptera), is due, not to a pairing of two chromosomes but to the precocious splitting of one. Hence they conclude that the haploid number is present in both germinal and somatic cells, and that the somatic divisions are essentially the same as the maturation divisions. According to their idea each chromosome divides in anaphase, giving rise to two daughter chromosomes which remain separated during the rest- ing stage and prophase (thus simulating a pair), and go to opposite poles in the succeeding division. Before considering the contentions of Taylor and of Lomen further, I will present some of the evidence that has led me to conclude that the double chromatic elements in flies are really 3 Several species of Drosophila included here are undescribed, and are given Sturtevant’s manuscript names. 222 CHARLES W. METZ pairs of chromosomes. This will make clearer the exact points at issue and facilitate subsequent discussion of the contrasting views. The evidence which I wish to present may be considered under three heads as follows: In the first place the number of chromosome pairs in diploid groups is the same as the number of single chromosomes in ma- ture germ-cells. Figures of the chromosomes in spermatocyte divisions, either first or second, or both, accompany those of diploid groups in most of the species included here, and speak for themselves in this regard. A comparison of figures 13 and 15, 27 and 33, 24 and 25, 44 and 48, 52 and 53, 74 and 77, 125 and 126, 137 and 139, ete., clearly shows the relation between haploid and diploid groups. In some species, the chromo- somes are evident even in the spermatids leaving absolutely no doubt as to the number contained in the spermatozoa. It must be concluded, therefore, that fertilization results in a diploid group in which the members of two haploid groups have associa- ted in pairs, unless we resort to the very improbable assumption that an eliminating process intervenes at some stage of fertiliza- tion to throw out half of the chromosomes or to fuse them to- gether two by two. Even this assumption, however, is over- thrown by the relations of the sex chromosomes described below. Secondly, if the diploid metaphase group were not made up of pairs, but were composed of double, univalent chromosomes, the two elements of these double chromosomes ought to lie one above the other, not side by side, in polar view, and in early anaphase a haploid group should be seen going to either pole. As a matter of fact neither of these conditions is realized outside of the maturation divisions. The two members of a chromosome pair lie side by side in metaphase, as shown by the figures, ex- cept for an occasional displacement, and frequently all of the chromosomes (the double number), may be seen dividing (figs. 7, 8, 9, 16, 28, 32, 40, 77). The side by side association and the method of division are clearly shown in figures 1-5, 7-9, 17, 19-24, 37, 39-46, 77, 98 and 99, etc. Figure 1, for instance, is composed of five symmetrical pairs, the members of which lie side by side. Figure 2 from the same species, shows similar ASSOCIATION OF CHROMOSOMES IN DIPTERA 223 features. Likewise in figure 3 the side by side arrangement is obvious. Figures 4 and 5, 17, 19 and 20, from species possess- ing another type of chromosome group, bring out the same relations. In each case the two members of a pair lie side by side, not one above the other—with the exception of one mis- placed chromosome in figure 19. Similarly in figures 21-24, representing another type of group, the side by side pairing is very distinct. Other examples are given in figures 27, 28, 37, 39-46, etc. These figures are not selected from among many in which pairing is less evident, but are perfectly typical and represent the normal condition in their respective species. The manner in which division takes place during late meta- phase or early anaphase is shown by figures 7, 8, 9, 16, 28, 32, 40, ete. Figures 7, 8, and 9 represent the same type of chromo- some group as do figures 4, 5, 17, 19, 20, namely, a group com- posed of two long U-shaped pairs, one straight pair, and one small spherical pair. In all of these figures each chromosome (save the smallest in 8 and 9) may be seen dividing equationally, in the ordinary manner. In figure 16 the mode of division in a similar group is seen at a somewhat later stage. The dark chromosomes are seen at a high focus, the light ones at a lower focus. It is evident that each member of the diploid group has divided and sent a daughter half toward either pole. The smallest pair cannot be seen in this figure. In figures 28, 32 and 40 the same process is indicated in the case of two other species. Earlier stages in the same species are represented in figures 27 and 39 respectively. The features indicated by fig- ure 28 are brought out even more clearly by figure 32 (a side view at the same stage). In figure 32 each of the short chromo- somes has divided, while the two long ones have split in prepara- tion for division. Passing now to the later anaphases it may be seen that during this period a diploid, not a haploid, group goes to each pole, and in many cases the two members of a pair of chromosomes are so clearly separated from one another that they cannot be considered the result of a precocious split as suggested by Tay- lor and by Lomen. This fact is demonstrated conclusively 224 CHARLES W. METZ in those cases in which the two members of a pair have become separated and do not lie side by side in metaphase. A few cases have been found in which the two members of a pair lie on op- posite sides of the spindle. In anaphase, each of these is seen to have divided and sent a daughter half to either pole. Figure 29 (same species as 27 and 28), for instance, shows a metaphase in which the two large members lie on opposite sides of the groups. In figure 30 a similarly arranged group is seen in ana- phase. Itis perfectly clear from the position of the large chromo- somes in figure 30 that the two large elements going to one pole are not sister halves of one chromosome, but are daughter halves of two separate chromosomes, else they could not lie on opposite sides of the spindle at this stage. A comparison with figures 27 and 31 shows how this differs from the normal condition in which the large as well as the small chromosomes are paired. The duality of the chromosomes in figure 31, if this figure were taken by itself, might be interpreted as indicating a precocious division of single chromosomes, rather than as indicating pairs of chromosomes, but other facts, as just described, preclude such an explanation. It is doubtless such appearances as those given by figure 31 that have led some authors to misinterpret entirely the nature of Diptera chromosomes. Fully as convincing evidence is furnished by other cases in which the two members of a pair have become only slightly displaced, instead of lying on opposite sides of the spindle. Such cases are shown in figures 7, 9, 12, 16, 28 and others. Fig- ures 7, 9, 12 and 16 are different stages in nuclei containing the same type of chromosome group. It is obvious that here one of the large pairs has been disturbed in such a manner that its two members resemble two horse-shoes placed side by side. Ac- cording to the ideas of Taylor and of Lomen these two members should go to opposite poles, but it is clear that they do not. On the contrary each divides and sends a daughter half to either pole. Figure 12 represents a particularly interesting case, for here the chromosomes have all divided and the daughter halves have separated. The figure on the left represents the upper group, that on the right the lower group (displaced in order to ASSOCIATION OF CHROMOSOMES IN DIPTERA ° 225 show the chromosomes clearly). Above them is a diagram show- ing the two groups in position as they appear in the section. Each chromosome in the one group is seen to be represented by a corresponding sister chromosome similarly oriented in the other. Such cases furnish unequivocal evidence that the two members of a pair are not daughter halves of a univalent prophase element, but are distinct chromosomes, and that they both divide equationally in metaphase. In the third place, diploid groups in the males of species hav- ing an unequal X-Y pair, demonstrate by the morphological difference between X and Y that the pair is composed of two distinct chromosomes. A striking example of this is seen in the three species of Drosophila shown in figures 41, 42, 44 and 45 (compare with figs. 49 and 50) in which species the X-chromo- some of the males is fully twice the size of its mate Y. It would be difficult indeed to imagine these being daughter halves of a univalent chromosome. The same features are also brought out by other species having unequal sex-chromosomes (figs. 85, 86, 88, 124, 135, 137, etc.), although the evidence is not always so striking as in the three species cited. These lines of evidence, I believe, leave no escape from the conclusion that pairing of chromosomes is a reality in the species here considered. That the mosquitoes are no exception to this rule will be shown below when the different groups of flies are treated independently. The essential difference between the above results and those of Taylor and of Lomen center around one particular feature— the behavior of the chromosomes in late metaphase and early anaphase. The other stages are not seriously disputed. The question, therefore, is whether the two metaphase elements separate from one another in anaphase, thus effecting a reduc- tion division, as described by Taylor and by Lomen, or whether each divides and sends a daughter half to either pole as Stevens maintained. I believe that I have demonstrated the correct- ness of the latter conclusion in the above paragraphs, and need not dwell further on it. The difficulty in the work of Taylor and Lomen is due, I believe, to faulty fixation of their material. 226 CHARLES W. METZ In my experience good preparations have been obtained only when the gonads or small bits of tissue were dissected out and fixed separately—never when the whole insect, or a consider- able part of it was fixed intact. The latter method, which is apparently the one used by Taylor and by Lomen, produces a clumping or running together of the chromosomes, ‘which is exactly the kind of behavior that would cause pairs to give the appearance of single chromosomes. Any tendency toward fusion is especially apt to exhibit itself in the anaphases, and hence it is to be expected that such figures as those obtained by Taylor and by Lomen would result whenever the fixation was defective. I have frequently obtained such a result when the fixation was poor, especially after Bouin’s, Gilson-Carnoy’s or alcohol-acetic fixatives. DETAILS OF CHROMOSOME BEHAVIOR DURING ONE CELL- GENERATION The mutual relationship of’ homologous chromosomes dur- ing the various stages of cell division has been carefully studied in both somatic and early germinal tissues of several species, and it is believed that the main facts regarding this relationship are now evident. In brief they are these: In metaphase, either in somatic cells, oogonia or spermatogonia, the chromosomes lie in a flat equatorial plate, the two members of each pair, with occasional exceptions, being arranged side by side as de- scribed above (figs. 1, 2,3, 17, 19, 20, ete.) Eachof these chromo- somes splits longitudinally, and during anaphase sends a daugh- ter half to either pole, still associated with its mate from the other member of the pair. Figures have already been given (7, 8, 9, 16, 28, 40) showing the chromosomes in the act of split- ting, or the daughter halves in the act of separating from one another, also figures (12, 30, 31, 95, etc.) showing later stages in which the halves have become well separated and are going toward their respective poles. Retention of the paired associa- tion during anaphase is evident in all, except those in which one or two pairs have* been disarranged. In the telophase, the chromosomes become closely massed and rapidly lose their ASSOCIATION OF CHROMOSOMES IN DIPTERA 227 staining capacity, so that very little can be determined about the behavior of individual chromosomes. It is significant, however, that these chromosomes normally enter the telophase in a closely paired condition (figs. 31, 95, 169, 171) and it seems highly probable that they retain this relationship during the transformations in the resting nucleus. Such a conclusion is rendered almost certain by their subsequent behavior in com- ing out of thé resting stage. The earliest prophase or spireme stages in which the chromatic threads may be distinguished with any degree of clearness show these threads to be intimately associated in pairs (figs. 11, 14, 34, 58, 65, 70, 71, 78, 80, 91, 92, 100, 123, 130, 131, 155, 165); and from this time on they may be seen to retain this association during their condensation and contraction from early prophase up to the time at which definite chromosomes are formed ready to go on the spindle. Some of the earliest prophases in which the chromatic threads were well defined are shown in figures 58 to 63 (Calliphora). Each of the double threads in these figures represents a pair of chromo- somes. In figure 62 all six pairs are shown (the smallest being very faint), but in the others only parts of the nucleus are repre- sented. Figure 65 is a later stage showing the chromosomes more condensed and contracted, but still closely apposed in pairs. Figure 66 is a still later stage, in which the chromosomes are assuming their definite shape preparatory to disjoining and going on the spindle. It is followed by the late prophase and metaphase stages represented in figures 53, 54, 55, 56 and 57. These are succeeded in turn by the late metaphase and anaphase in which each of the twelve chromosomes divides equationally as described above. Other early prophases are shown in figures 100 to 102 (Homalomya). The chromosome group here is indistinguishable from that of Calliphora (five large and one small pairs). In figure 100 the long, delicate but double threads are clearly distinguishable. It is impossible to determine precisely how many double threads are present, for some are broken, but the number is clearly about five or six, cer- tainly not ten or twelve. Part of a similar nucleus is shown in figure 101. One of the most interesting features about THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 21, NO. 2 228 CHARLES W. METZ these figures (100, 101) is the evident polarization of the chromatic threads. This appears to be characteristic of very early prophases, although such stages are seldom clear enough to draw. When this polarity is compared with that shown by telophases (figures 111 and 112), it is difficult to avoid the con- clusion that the two are correlated,—that is, that the chromo- somes reappear during prophase in the same relative position, and polarized in the same manner as in telophase. Prophases in other species similar to those cited above are represented by figures 70, 71, and 72 (Musca), 106-108 (Fucellia), 78-80 (Phor- mia), 91, 92 (Sarcophaga), 130, 181 (Anthrax lateralis) and 123 (Hristalis). These are all essentially alike and involve corre- sponding chromosome groups. Prophases, together with meta- phases for comparison, in species having fewer chromosomes, are shown in figures 14 and 15; 11 and 4, 5; 34, 35, 36, and 4, 5; and 165 and 166. As seen in the figures all stages subsequent to the condensation of the chromatic elements in early prophase are easily followed, although the behavior of the chromosomes differs slightly in different cases. Usually the association, of the two members of a pair becomes loose long before contraction is completed. At this time the two threads are loosely and irregularly coiled about one another (figs. 34, 59, 71, 92), and as contraction pro- ceeds they become more and more loosely associated (figs: 35, 72, 93, 94). Occasionally, however, a close association is re- tained up to a very late period of contraction (figs. 36, 66, 108, etc.), with the consequent production of figures which very closely simulate those of haploid groups. Such figures as these might readily create the impression of haploid groups in diploid nuclei. By the time spindle formation takes place the chromo- somes are usually distinctly disjoined from their mates, although the paired association is still conspicuous and may be very close (figs. 15, 41, 68, 117, 132, ete.) Occasionally the process of separation has been carried on so far that pairing is very in- definite (figs. 153, 161), but such cases are decidedly exceptional. Soon after the chromosomes become arranged on the spindle they begin to show evidences of splitting in preparation for ASSOCIATION OF CHROMOSOMES IN DIPTERA 229 division (figs. 8, 28, 40, 77, 99, etc.), and by the time the equatorial arrangement is completed they may all exhibit a longitudinal split. It is this stage that demonstrates unquestionably the presence of a diploid instead of a haploid group. As shown by the figures, especially numbers 7, 28, 29, 68, it occasionally happens, as mentioned above, that the members of a pair appear in metaphase on opposite sides of the spindle, or separated from one another by other chromosomes. ‘This disarrangement apparently takes place in late prophase while the chromosomes are becoming equatorially oriented. Several cases have been observed in which the members of a pair were partially separated by other chromosomes. and it seemed a question as to whether they would be forced completely apart, or would succeed in taking their places together. The frequent appearance of the condition in which the two members are on opposite sides of the plate appears to be due to their having ap- proached the equator of the spindle vertically instead of hori- zontally, i.e., from one pole instead of from the side—and thus having been pulled diametrically apart, with their points of attachment near together but their extremities pointing in opposite directions. At first sight it would appear that chromo- somes once separated in this manner would have difficulty in associating again, and that after many divisions all the pairs would be disarranged. An examination of chromosome ar- rangement in late anaphase indicates one reason at least why such a confusion does not occur. During this stage the chromo- somes are drawn out in a slender cone with their apices brought close together at the pole. As a result all of the chromosomes are rather closely approximated throughout their lengths, and an ample opportunity is afforded for the reunion of separated members of a pair, even if they previously lay on opposite sides of the spindle. In my paper on Drosophila chromosomes (Metz 714, p. 56), I mention the apparent occurrence of a ‘second conjugation’ of chromosomes in early metaphase, after the separation which normally occurs in prophase. The details of this phenomenon were obscure at the time, and were left for further study. It 230 CHARLES W. METZ appears now, after a careful study of these stages in a large number of flies, that the so-called ‘second conjugation’ is of only occasional occurrence, and is not a uniform stage in the chromo- somal activities. In some, if not all cases, it is simply a retention of the close approximation that existed in prophase. When considered step by step, as has just been done, it is remarkable what a resemblance the above processes bear to those of maturation. In the early prophase stages of either » somatic or gonial nuclei an almost exact simulacrum of diplo- tene nuclei is often found. This extends in some cases, even to definite polarization of threads within the nucleus, such as is shown in figures 100 and 101.4 PAIRING IN DIFFERENT TISSUES AND DURING DIFFERENT STAGES IN ONTOGENY No attempt has been made in this study to examine in detail all of the somatic tissues in any one species. Various tissues have been dissected out at different times, however, and fixed with the gonads. In this manner I have been able to study division figures in most of the tissues of the body and during most stages of ontogeny. Among the organs and tissues defi- nitely identified in these studies the following may be mentioned; embryonic brain, eyes, malpighian tubules and wing buds, and somatic as well as germinal parts of the testes and ovaries. I have also examined various bits of tissue taken at random from dissected larvae and pupae of various ages. In addition to studying isolated pieces, I have studied sec- tions of entire embryos (larvae) in which all of the tissues could be examined. Of course division figures were never visible in all the tissues of these total preparations, but they were fre- quently found in several parts of one object.® In regard to the ontogenetic development I may state that I have examined all stages from the newly hatched larvae up 4See concluding paragraph, page 257. 5 As mentioned under ‘Methods’ the figures in total preparations are poor, but they are sufficient to show whether the chromosomes are paired or single. ASSOCIATION OF CHROMOSOMES IN DIPTERA 231 to the sexually mature fly in several species of Drosophilidae, Muscidae and Anthomyidae. The results of all these studies on somatic tissues may be summed up in one sentence, namely, that in all tissues of the body and during all stages in development from the newly hatched larva to the adult fly the paired association of corre- sponding chromosomes is a universal characteristic. So far as I have been able to determine, the pairing phenomena are identical in all diploid cells, whether somatic, spermatogonial or oogonial, from the egg to the adult. DIFFERENT SPECIES AND FAMILIES COMPARED In order to determine whether the paired association of chromo- somes is characteristic of all Diptera or whether it is restricted to certain individuals or groups, an attempt has been made to study representatives of all the principal divisions in the order. As a result, sixteen families ranging from among the lowest to the highest have been included in the survey. Some of these families are represented by one or two species, others by several species. Since the principal aspects of the pairing phenomena are essentially the same in all of the flies studied no attempt will be made to treat each individual species. In- stead, a few characteristic members will be chosen as repre- sentatives of the respective families. Likewise, no attempt will be made to give a complete account of the chromosome behavior in each species treated. In many cases only enough figures are reproduced to show the nature of the chromosomes and their paired association. For convenience the order of treatment of the families is the reverse of that given in the synopsis (i.e., from the highest to lowest instead of vice versa), except that file Muscidae will be considered before the Sarcophagidae. Muscidae Calliphora erythrocephala (figs. 51-66). Figures 51 and 52 represent the haploid group of this species, taken from first spermatocyte divisions. The group consists of four similar, 232 CHARLES W. METZ long chromosomes, one shorter chromosome and one small, spherical chromosome. Figures 53 to 57 illustrate correspond- ing diploid groups of the same species. taken from ovarian (53-56) and somatic (57) cells. From these figures it is evident that for each single chromosome of the haploid group there is a pair of chromosomes in the diploid group, and that the members of this pair are in close proximity to one another. Earlier stages, show- ing the origin of the pairs in prophase, are given in figures 58-66. Some of the figures represent only sections of the nucleus, but others (58, 62, 63, 65, 66) are taken from uncut nuclei and in- elude all of the chromatic material. In early prophase stages the five pairs of long chromosomes are clearly represented by the five long, double threads as shown in figures 58, 62 and 65. Frequently the small pair is concealed and cannot be distin- guished, but in many cases it is as clearly evident as are the others (figs. 62, 65). The duality of the threads in early pro- phase is perfectly distinct in almost all cases. The figures given here are entirely typical of scores studied, and are taken from various tissues of the body allof which show the same phenomena in dividing cells. Very rarely a figure is found in which no duality can be seen in the threads (fig. 63), but it seems certain that this appearance is due merely to overstaining which con- ceals the true dual nature. Figure 65 is a good example of such a case. When first studied the members of this group appeared to be perfectly homogeneous elements and were drawn as such, but after the material had cleared in balsam a few months, the duality of the threads became very evident, as shown in the figure. I have no hesitancy, therefore, in considering figure 63 to be of the same nature, especially since it is almost the only clear case of its kind found. During later prophase stages such as shown in figures 53, 55 and 66, the chromosomes rapidly contract, and condense, and the members of a pair dissociate somewhat in preparation for division. When they go on the spindle they form a flat equatorial plate, with corresponding chromosomes arranged side by side in the same plane. Only in exceptional cases, such as are in- ASSOCIATION OF CHROMOSOMES IN DIPTERA Paps evitable under the circumstances, are the two members of a pair in any other relation than this during metaphase. Musca domestica (figs. 68-72). In Musca the chromosomes are very similar in form and behavior to those of Calliphora, except in respect to the sex-chromosome pair, which is almost as large as the autosomes. Haploid groups of Musca have already been published by Stevens (’08, fig. 3). The accompany ing figures are taking solely from diploid groups to illustrate the pairing phenomena. They are all from ovarian tissue far in advance of maturation stages and may be said to represent the characteristic features of prophase and metaphase in early ovarian and somatic cells. Figures 68 and 69 are metaphases showing the six pairs of chromosomes in the equatorial plate. It will be noticed that in each figure the members of one pair of chromosomes are displaced and are not closely associated. These are in all probability the sex-chromosomes (XX). Prophases showing the early appearance and the disjunction of the chromo- somes are represented in figures 70, 71 and 72. The former is from an entire, or nearly entire nucleus, the latter two are from cut nuclei, but each includes almost all of the chromatin. Phormia regina (figs. 73-80). Haploid groups of this spe- cies are shown in figures 73 to 75 (second spermatocytes) and figure 76 (first spermatocyte). As shown in figures 73 and 75 in contrast to 74, the sex-chromosomes (smallest in each case), are very unequal. In figure 76 they may be seen sepa- rating from one another in the reduction division. Figure 77 is taken from a spermatogonial cell in early anaphase (or late metaphase), and shows the six pairs of chromosomes, correspond- ing to the six single chromosomes of the haploid group; each of these is split lengthwise in the process of division. In the cen- ter may be seen the unequal X-Y pair splitting in the same manner as are the autosomes. A comparison of this figure (77) with that of a similar stage in the reduction division (76) clearly brings out the relation between the two groups (haploid and diploid). Prophases from early ovarian tissue showing the origin and behavior of the pairs in preparation for division in 234 CHARLES W. METZ _ diploid nuclei are given in figures 78 to 80. They differ in no essential respect from those in Calliphora and Musca. Likewise the other Muscidae studied (Muscina stabulans, Calliphora viridescens, Lucilia sericata, and Pseudopyrellia sp.) agree with those already described. Sarcophagidae Sarcophaga (figs. 81-97). Several species of Sarcophaga have been used in this study and have been found to agree so completely in respect to chromosome behavior that they will be treated as a whole. For specific references see explanation of figures on p. 270. Haploid groups from second spermato- cyte divisions are given in figures 81 and 82, and from first spermatocyte divisions in figures 83 and 84. The last named is a side view showing the inequality of the X Y chromosomes at the time when they separate during reduction. Corresponding diploid groups are represented by figures 85-88 (spermatogonial), figure 89 (ovarian follicle cell) and 90 (somatic, embryonic cell). In the male groups (figs. 85-88) the difference between X and Y (smallest chromosomes) is plainly evident. Prophases showing the early appearance of the pairs, and quite comparable with those in the Muscidae, are given in figures 91 (somatic, two sections of same nucleus), and 92 to 94 (somatic). An anaphase from a similar cell (embryonic glandular tissue) is given in figure 95. It clearly shows the persistence of the paired association and indicates the relative positions occupied by chromosomes when they enter the telophase and subsequent resting stage. In this figure the spindle fibers are schematized, but the chromosomes as in other figures are drawn in their exact position. Figures 96 and 97 are taken from multiple groups (somatic) showing respectively 24 and 48 chromosomes. The former is significant because it shows tetrad aggregates instead of pairs (compare with figs. 85-90 and see pp. 252 and 253). In the latter the chromosomes are so massed together as to obliterate the associations. ASSOCIATION OF CHROMOSOMES IN DIPTERA 235 Ravinia peniculata (figs. 98, 99). This species is indis- tinguishable from those of Sarcophaga in respect to pairing phenomena. Figures 98 and 99 are ovarian (early pupal) metaphases showing the six pairs of chromosomes essentially like those of Sarcophaga. The latter shows the metaphase splitting of the chromosomes very clearly (compare with figure aay Anthomyidae Homalomya sp. (figs. 100-105). Particularly clear prophase figures have been secured in this species, both with respect to somatic and to spermatocyte divisions. The chromosome group is practically indistinguishable from that of Calliphora (figs. 51-57). Figures 100 and 101 are very early prophases from somatic nuclei, illustrating the configuration of the chromat- ic threads at this time. The former is from an entire, or almost entire nucleus, in which the bivalent (double), long drawn out threads, each representing a pair of chromosomes, are discern- ible. Attention is particularly called to the polarization of these threads and the resulting similarity in appearance between this somatic prophase and the synaptic stages accompanying maturation in many other animals. Figure 101 represents a similar stage from the same tissue, but includes only a portion of the nucleus. A later stage in which these threads lose their polarity and contract before giving rise to the metaphase chromo- some pairs is shown in figure 102. In comparison with such somatic and spermatogonial prophases it is of interest to exam- ine corresponding stages in the maturation divisions. Figure 103 is a.portion of a second spermatocyte prophase and shows sister chromosomes closely intertwined preparatory to going on the spindle. In metaphase (fig. 104) they come to le one above the other in the equatorial plane. Figure 105 is a second spermatocyte anaphase. In figure 103 only three of the chromo- somes are represented, but in 104 and 105 the full (haploid) comple- ment is present. The double elements in these cases are split univalten chromosomes, the two members of which separate in anaphase as shown in figures 104 and 105. It is important 236 CHARLES W. METZ to note that at certain stages in prophase the figures of all three (somatic, first maturation and second maturation) divisions are superficially very similar, although the actual processes in the three cases are very different. Fucellia marina (figs. 106-110). As in the previous case, so in the present, the paired relationship of the chromosomes is essentially like that described for the Muscidae and Sarcopha- gidae, and requires no detailed description. A few somatic prophases have been reproduced to show the origin of the chro- mosomes in the former in the form of closely paired threads, and the subsequent disjunction of these into the less closely associated condensed chromosomes found in metaphase. Figure 106 is an early prophase showing the six bivalent threads. Fig- ures 107 and 108 are somewhat later stages illustrating the sepa- ration of the threads. All three are complete (diploid) figures. - The most interesting features observed in Fucellia are those shown by prophases containing multiple (probably tetraploid) groups (figs. 109, 110). Each chromatic aggregate in these, contains four (or eight) chromosomes instead of the usual pair, (compare with figs. 96 and 97 and see pp. 252 and 253). Ophyra leucostoma (figs. 111-114). In most of the Diptera studied so far great difficulty has been experienced in analysing telophase figures. Usually the chromatin is so massed at this point that no details whatever can be distinguished. In the present species, however, a few figures have been obtained, which although far from satisfactory, are nevertheless sufficient to show something of the chromosomal behavior during this stage. Two of these are shown in figures 111 and 112. They suffice to show the loop or U-shape of the chromosomes, and sug- gest the process of reticulation that is taking place as the chroma- tin becomes diffuse. The polarity of these U-shaped threads bears a significant relation to the similar polarity evident in early prophase when the chromosomes reappear (figs. 100 and 101). The chromosome group and the pairing phenomena of Ophyra are practically the same as those of Homalomya and Fucellia. Figure 113 shows a late diploid (spermatogonial) prophase with six pairs of chromosomes, some of which already ASSOCIATION OF CHROMOSOMES IN DIPTERA Dat indicate the metaphase split; and figure 114 shows a corre- sponding but somewhat later stage in the first maturation (re- duction) division. Sciomyzidae Neuroctena analis (figs. 115, 116). There is nothing peculiar about the chromosomal behavior in the Sciomyzidae, so far as I have been able to determine. Several specimens of N. analis have been studied, with results comparable in every way to those already described. The two accompanying figures are sufficient to show the paired association and the relation be- tween haploid (fig. 115, second spermatocyte) and diploid (fig. 116 spermatogonial) groups. Trypetidae Flies of this family, so far as my experience goes, are not favor- able for chromosome studies. Nevertheless they present suf- ficiently clear figures to show that the paired association is present here just as it is in other flies. Most of my studies were made up- on Euaresta melanogaster, material of which I secured in Cuba. The chromosome group of this species appears to be composed of six pairs similar to those in the Muscidae, although no fig- . ures have been found that are complete and at the same time clear enough to settle this point. Ortalidae No embryonic stages (larvae or pupae) have been secured from any members of this family, and consequently no somatic divisions have been studied. Spermatogonial and spermatocyte divisions have necessarily formed the basis of my observations on both of the following species, yet there can scarcely be any question that there is a definite correspondence between the phenomena exhibited by spermatogonia and somatic cells. Chaetopsis fulvifrons (figs. 117-119). Chromosomal behavior in spermatogonia of this species corresponds fully with that described for ovarian and somatic cells in species of Drosophili- 238 CHARLES W. METZ dae (figs. 4-20) having a similar chromosome group. In Chae- topsis no good figures of early spermatogonial prophases have been secured, owing to the small size of the nuclei, and to dif- ficulties in fixation. Metaphases, however, are distinct (figs. 117-118) and plainly show the paired arrangement of the chromo- somes. These, when compared with maturation divisions show- ing the haploid group (fig. 119, first division) leave no doubt of the relations in this species. Camptoneura picta (figs. 120, 121). Since C. picta shows pairing relations similar to those in the last named species it attracts attention only because it differs so markedly from Chaetopsis in respect to the number and size relations of its chromosomes. As a matter of fact Chaetopsis excites the greater interest, for Camptoneura has the chromosome group (fig. 120, diploid, and 121, haploid) found in several families (all those above mentioned, as well as the Sapromyzidae, Micrope- zidae, Sepsidae, Syrphidae, and one species of Bombyliidae), while the group found in Chaetopsis is found in no other species I have studied outside the Drosophilidae. Sapromyzidae Physegenua vittata (fig. 122). I have had difficulty in ob- taining suitable material from Sapromyzid flies, but as in the case of the Trypetidae enough has been secured to determine the essential point—that the chromosomes are associated in pairs. Figure 122 (spermatogonium) represents one of the few com- plete polar views found. It is seen somewhat diagonally, with the result that some of the pairs appear to lie beneath the others, but in reality they form an almost flat plate, entirely comparable with those seen in the Muscidae, etc. The two small chromo- somes are doubtless the sex-chromosomes (X be just as are the small ones in the Muscidae. Drosophilidae (See pp. 222-224, “Reality of chromosome pairing.’”’ For spe- cific references see explanation of plates; also. Metz ’14.) ASSOCIATION OF CHROMOSOMES IN DIPTERA 239 Syrphidae Eristalis tenax. My studies in this species have included pupae as well as adults, and in both I have found the chromo- some behavior to agree with that in the cases described above, and with Stevens’ (08) description. Eristalis bastardi (fig. 123); Volucella obesa (figs. 124-126); Mesogramma marginata (figs. 127, 128). These three species are very different from one another in appearance, but their chromosomes appear very similar (save for minor details. of size relations) and hence will be considered together. Figure 123 (Eristalis bastardi) represents part of a prophase figure showing the bivalent chromatic threads which are comparable in every way with those seen in Homalomya, Sarcophage, ete. Figures 124 and 125 are metaphases (spermatogonial) of Volu- cella, and clearly show the paired relationship. In the former one chromosome is missing, leaving a single member (in left margin of group), without a mate, but otherwise all are paired. This species is particularly interesting because of the differ- ent sizes apparent inits chromosomes. One pair is easily recog- nized by its large, and one (sex-chromosome) by its small size, and even the others show slight differences from one another. Figure 126 is a first spermatocyte division for comparison with the diploid groups; note the unequal X and Y chromosomes, which are paired in the diploid groups. Figures 127 and 128 (spermatogonial) of M. marginata are of significance only in showing the paired arrangement of the chromosomes. Sepsidae Piophila caset. There is no marked distinction between P. casei and the various species of Muscidae and Sarcophagidae, either in chromosome numbers and size relations or in the general chromosome behavior. Bombyliidae Anthrax lateralis (figs. 129-133). No more conspicuous cases of chromosome pairing have come tomy attention than those 240 CHARLES W. METZ exhibited by this and other species of Bombyliidae. Figures 129 to 133 are only a few from among scores of similar ones studied. In all cases the five large pairs and often the small pair stand out clearly and show a close approximation. The figures need little explanation beyond that given already for preceding species. Numbers 129 to 131 are spermatogonial prophases showing the five long and one short double threads, which later loosen up and contract to form the metaphase pairs shown in figures 132, 133. ; Anthrax sinuosa (figs. 1384-140). This species is very inter- esting from several standpoints. In the first place it possesses chromosome pairs of various sizes (figs. 134-137), which clearly illustrate the pairing of corresponding chromosomes. Secondly the evident dissimilarity between A. sinuosa and A. lateralis in number of chromosomes, the former having 18, the largest group in any fly within my knowledge, and the latter possess- ing but twelve, presents the greatest divergence of this nature that I have observed between two species in one genus. ‘Thirdly, the sex-chromosome pair is apparently one of the largest in the group, instead of the smallest, as has been the case in all of the above species exhibiting a conspicuous inequality between X and Y. Unfortunately I have been unable to identify the sex-chromosome pair in A. lateralis. If the small pair in A. lateralis (figs. 132, 133) is the sex-chromosome pair, as it is in many flies, then a remarkable difference exists between the sex-chromosomes of the two species, such a difference as I have found in no other closely related flies. Similar differences have been observed between related species of Hemiptera and Coleop- _tera, but seem to occur very rarely among the Diptera. In maturation divisions of this species (figs. 138-140) the short chromosomes show a tendency to become rounded, but the rela- tive sizes are readily seen to correspond with those of the dip- loid groups. Figures 139 and 140 (second spermatocytes) appear to be respectively X- and Y- containing groups. As the spermatogonial figures (134-137) show, X is the largest chromosome present, while Y is smaller than the two largest autosome sizes. Comparing figures 139 and 140 it may be seen ASSOCIATION OF CHROMOSOMES IN DIPTERA 241 that the latter contains three large chromosomes (X and the two largest sized autosomes), while the former (139) has only two large chromosomes but has an extra small member which must be Y. No sufficiently clear first maturation divisions have been found to show the X-Y relations of that stage, unless the apparently single element projecting from the largest chromo- some in figure 138 is the unmatched end of X. If so, one of the smaller pairs is concealed. The figure is drawn just as it appears, but I am not sure of its significance. Spogostylum simson (figs. 141, 142). No males of this species were secured, but very clear figures were observed in ovarian follicle cells. Two of these are given to indicate the similarity between the pairing here and in the other species. Figure 141 is a metaphase plate showing the diploid group and the associa- tion in pairs. Figure 142 illustrates a similar cell in prophase with corresponding chromosomes forming closely united double threads in the characteristic manner. As the figures indicate, this group differs markedly from both species of Anthrax in the size and form relations of its members. Apparently there is no dominating type of chromosome group in the Bombyliidae such as is seen in the majority of other families. Asilidae Twelve species of this family have been studied as indicated in the synopsis (p. 217), but only a few of them need be considered. ‘Those chosen are selected particularly to illus- trate the various numbers and sizes of the autosomes, and the varying degrees of inequality of the sex-chromosomes. Pair- ing is constant in all of them. Asilus sericeus (figs. 143-145). This species has perhaps the most simple group found in the family, containing as it does only five pairs of chromosomes, and lacking any conspicuous inequality between the sex-chromosomes. Yet it is one of the most inter- esting groups I have found, for each pair appears to differ from all the rest in respect to size. The two large pairs are admittedly 242 CHARLES W. METZ very nearly the same size, but even they may be distinguished in some figures (note especially figures 143 and 144). Asilus lecythus (figs. 146-148). Scarcely less striking in the matter of size differences is the evidence presented by this species. Upon close examination its seven pairs (or its seven single chromosomes in haploid groups) are seen to be definitely graduated in size from the smallest to the largest. The grada- tions are somewhat confused in the diploid groups by the uneven- ness and the flexures of some of the chromosomes, but in hap- loid groups (fig. 148, second division) the gradation is much more conspicuous. The sex-chromosomes, apparently, are not unequal. | Astlus notatus (figs. 149, 150). What has been said of the last species (A. lecythus) applies equally to the present one, except that the size differences between the larger pairs are searcely distinguishable. Figures 149, 150 show spermatogonial and second spermatocyte groups of this species. Leptogaster badius (figs. 151, 152). The diploid group of this species is shown in figure 151. As may be seen it consists of five pairs, only two of which may be differentiated by size. The largest of these is the sex-chromosome pair, whose mem- bers, as in previous cases are frequently not associated during metaphase. The haploid group is indicated by figure 152 (second division). : Erax rufibarbis (figs. 153, 154). In this species, also, five rather similar pairs of chromosomes are found. As in the pre- vious case only the smallest and largest (sex chromosome pair) may be differentiated. Figure 153 shows the chromosomes in a flat plate and indicates their size relations. In spermatocyte divisions the chromosomes of E. rufibarbis show a decided ten- dency to condense and become rounded, but the size relations are nevertheless conspicuous (fig. 154). This tendency toward condensation extends even into the spermatids, thus enabling one to count the chromosomes with ease, and to determine with- out doubt the number of chromosomes carried by the spermato- zoan into the egg. ASSOCIATION OF CHROMOSOMES IN DIPTERA 243 Dasyllis thoracica (figs. 155-158). D. thoracica furnishes evi- dence very similar to that presented by Asilus sericeus. No two of its five pairs of chromosomes (fig. 156) appear to be the same size. The smallest and next smallest pairs are very dis- tinct, as is also the largest. Possible confusion arises then, only in connection with the two intermediate pairs, but since one of these appears to be the X-Y pair its dimorphism, if the apparent dimorphism is real, serves to differentiate it from the other intermediate pair. I have been unable to obtain sufficient spermatogonial figures to determine definitely the - gex-chromosome relations, but evidence from the first sperma- tocyte divisions makes it probable that the relations shown in figure 156 are correct. In the first spermatocytes (fig. 157), one of the intermediate pairs (corresponding to XY in figure 156) appears to have a univalent attachment (X in the figures) at one end, which strongly suggests the unpaired end of an X- chromosome. Analysis of the first spermatocyte group (fig. 157) then, reveals one small spherical chromosome (1), one small, elongate chromosome (2), one larger, symmetrical chromosome (3), one similar, but asymmetrical chromosome (4), and one largest chromosome (5), each distinct from all of the others. In the diploid group each of these is represented by a pair of chromosomes. A diploid group showing the intimately paired association in prophase, similar to that in the Muscidae, etc. is given in figure 155. Fate as viv Gave Pe ye , A ris R ras ‘ 4) Sie ‘ = q : a] - , 7 ne ; : a i Hy / ’ ‘ai ¥ Vere Wy I if > A ~ * ~ 4 Y a - . ‘ . “ ’ ; ial ies J } \ > - * 4 ‘ : i ‘ wt ; S . 5 . ' . FS 3 j r ‘ i . ve ' * ny j 1 le : at i veka ges ? . ‘ 4 : - | r 4 ’ ro = be Ag | ; 4 7 4s Pud ’ 3 a : + * j ee * \ y if . wie tix a ¥ 1 i ‘ . j rar, si i = Fr 2 us ‘ @. 4 4 A 4 ‘ ‘ t jue - 1 UJ i THE GROWTH OF THE BODY AND ORGANS OF THE ALBINO RAT AS AFFECTED BY FEEDING VARIOUS DUCTLESS GLANDS (THYROID, THYMUS, HYPOPH- YSIS, AND PINEAL). EK. R. HOSKINS Institute of Anatomy, University of Minnesota FOUR CHARTS CONTENTS LS TET DYSON nee Gc Bkd 5 Shy: ER” ON, 29 yO ee a ac 296 RMN ceri Cray ee eee hte, Me es ht a Be 297 Rie Meier andere thodss.44-e5 .. a a 336 eM SINTATTATIEUES 7 Aart Sit SON SET RO RRR Oe ie seek | 337 A eC ONPEOIsEAM A MOLMIS..) oc) sR EES bee 2c ac dee ch oud 337 Bedkittectsvotptinyroid feeding. aan. ste eek occ de ds cd ode. 338 295 THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 21, No. 3 OCTOBER, 1916 296 E. R. HOSKINS ec: ‘Efiects: of thymiis feedinpsameameiere a. 2's 2 ws. a eee eer 340 de shiffects of Mypophiysissee din camer). .). fae oe see eee 340 e. "Effects of pineal feeding serene... 6... 0... - eee eee 341 Wo MOON CltiistOng ss he< oie ke Oe ici oa OEE are cis oa stios eon ee 341 Vila Genat ure vCUbe Cs ssc iecz.c csv essten ae oe ake a oi oiis6 die. sve) lane eee 343 I. INTRODUCTION The present investigation was undertaken in the hope of throwing further light upon the relations of some of the ductless glands to the growth process in albino rats. During the experi- ment it became evident that the growth rate in the control rats was in many cases somewhat different from that which has been generally described as normal, so it became necessary to include incidentally the question of the normal growth rate. The investigation was carried on in the Anatomical Institute of the University of Minnesota, under the direction of Profes- sor C. M. Jackson, to whom my grateful thanks are due for his constant interest in the work and his many very helpful sug- gestions. Since the ductless or endocrinous glands were first recognized anatomically, various methods have been applied in investi- gating their functional significance. The four most commonly used, are extirpation of the glands, their transplantation, in- jection of their extracts and the feeding of the glandular sub- stances. . Feeding ductless glands in order to study the effects of the hyperactivity thus possibly produced has certain objections. ' The absorption is slower than when injections are made, and the danger of infection is lessened; but the substances fed may un- dergo digestive changes in the alimentary tract. That the active principles of the ductless glands are not necessarily de- stroyed by digestion, however, is proven by abundant experi- mental and clinical results (Gudernatsch 712, and Abderhalden 715 especially in thyroid feeding). The feeding method was selected for the present investigation. — GROWTH AFFECTED BY FEEDING DUCTLESS GLANDS 297 II. LITERATURE A general discussion of the literature of all of the ductless glands is given by Vincent (712), “Internal Secretion and the Ductless Glands,’’ and in the more extensive work of Biedl (713), ‘‘Innere Sekretion.”’ The hypophysis literature has been reviewed by Cushing (712). ‘‘The Pituitary Body and _ its Disorders;’’ and the work on the thymus has been considered recently by Basch (713), ‘“‘Beitraége zur Physiologie und Path- ologie der Thymus.’ A complete list of references to work done upon the albino rat is given by Donaldson ’(15). A preliminary report of the principal results of the present investigation has already been published (Hoskins ’16). Thyroid experiments Iscovesco (’13) found that daily injections of thyroid extract stimulated growth slightly in young animals but decreased the weight of old animals. He found nearly 100 per cent hyper- trophy (measured in grams per kilogram body weight) in vari- ous viscera. The extreme and uniform hypertrophy of the organs and especially that of the uterus is difficult to under- stand. The fact that the liver and the female kidney show no overgrowth is remarkable. ‘Magnus-Levy (’95) found that feeding thyroid may cause loss in weight in an animal. Cunningham (’98) fed ‘considerable amounts’ of thyroid to various animals without noting any toxic effects. Moussu (’99) reported that small doses of thyroid stimulate the rate of growth in young dogs, but that large doses are toxic. Rudinger, Falta and Eppinger (’08) and Kostlivy (10) found that feeding thyroid stimulates the suprarenal glands. Bircher (710 a) (10 b) fed thyroid to young rats and found a retardation in growth and body weight, but an acceleration in the process of ossification. Utterstrém (’10) reported an enlargement of the thymus in thyroid-fed rabbits. Hoskins (’10 a) fed daily, for 15 days varying amounts (5-15 mg.) of thyroid to 18 young guinea pigs. Their suprarenal glands 298 E. R. HOSKINS were 25 per cent heavier than those of 18 controls. The same author (’10b) fed thyroid to pregnant guinea pigs. Many abortions and several still-born occurred. The newborn (ap- parently normal) young of the treated mothers weighed on the average 12 grams less than the controls. The hypophysis showed an average decrease in weight of 10 per cent, the supra- renals 26 per cent and 2 per cent for the females and males respectively; the ovaries 26 per cent; and the thyroid gland 18 per cent. The thymus was increased 58 per cent. Carlson, Rooks and McKie (’12) and Ferrant (’13) fed thy- roid to birds and mammals, including man. They concluded that large doses of thyroid are toxic. In the thyroid fed rabbits of Ferrant the heart, liver and pancreas showed degenerative changes. Schafer (’12) fed thyroid to young white rats, noting an in- creased food consumption, increased metabolism and accelera- tion of growth. Gudernatsch (712, 714) found that thyroid administered to a large number of tadpoles retarded growth (i.e., a toxic effect) but hastened metamorphosis of the limbs and tail. It is prob- able that this acceleration is due to an increase in the rate of the circulation and general metabolism. Lenhart (15) reaches the same conclusion. The present writer, in a similar (unpublished) experiment (with larval frogs and Ambystomae) fed large doses and noted only the toxic effect. Coutronei (14), West (14), Morse (715) Abderhalden (’15) and Romeis (715) have recently confirmed in general the results of Gudernatsch. Hewitt (14) fed thyroid gland to white rats, noting a loss in weight. Livingston (’14) found that feeding thyroid inhibits the hyper- trophy of the hypophysis which follows thyroidectomy in male rabbits. Gudernatsch (’15) fed thyroid to albino rats. This treatment retarded growth and interfered with pregnancy. The effect produced is probably due merely to the toxicity of thyroid, as Stockard (’13) obtained quite similar results in animals treated with alcohol. GROWTH AFFECTED BY FEEDING DUCTLESS GLANDS 299 Thymus feeding For complete literature see Vincent (’12), Biedl (713), Basch (13) and Paton (713). Gudernatsch (’12) (14) found that a thymus diet stimulated body growth in tadpoles, but retarded metamorphosis of the limbs and tail. Similar results in some cases were obtained by Romeis (715) and Abderhalden (715). Gebele (11) and Miss Hewer (14) reported negative results with thymus feeding. In the latter’s work, thymus, when fed (1 to 4 g. daily) to males, retarded the development of the testes in young rats and caused degeneration of the testes in adults. Salkind (15) reported stimulation of growth from feeding large amounts of thymus. Hypophysis experiments Cushing (712) concludes that total loss of the anterior lobe of the hypophysis is followed by death and that partial loss is fol- lowed by obesity, sexual infantilism, and retardation of skeletal growth. Frequent injections of hypophysis extract interfere with growth, (Cerletti ’09, Fodera and Pittau ’09). A loss in weight with no skeletal changes is reported by Franchini (’10) and by Crowe, Cushing and Homans (710). A gain in weight after continued injections of hypophysis extract is reported by Delille (09), and cardiac hypertrophy by Etienne and Parisot (’08); but Caselli (00) obtained negative results. Retardation in growth (especially skeletal) as a result of hypophysis feeding was reported by Thompson and Johnston (05), Etienne and Parisot (’08), Sandri (’07), (’09) (posterior lobe), and Cushing and Goetsch (cited by Cushing (’12), Wulzen (14) and by Pearl (16) (anterior lobe) Abderhalden (’15), and Robertson (’16). Negative results from feeding hypophysis are reported by Caselli 00), Sandri (’09) (anterior lobe), Hoskins (’11), Aldrich (12a) (12b), Schafer (’12), Lewis and Miller (13), and Guder- natsch (14). 300 E. R. HOSKINS Schafer (09) had thought that feeding small amounts of an- terior lobe of the hypophysis favors growth. Goetsch (’16) from very few data reports that excessive dosage retards growth in young rats, whereas smaller dosage accelerates growth in body weight and especially the development of the sexual system; that this acceleration is due to the anterior lobe; and that feed- ing posterior lobe does not accelerate growth, and even retards sexual development. Robertson (’16a) reports that hypophysis extract accelerates growth after inanition. | So far as the literature shows, in none of the above experiments with hypophysis substance was a complete autopsy performed or complete histological examination made. Hallion and Alquier (08) report hypertrophy of the suprarenals in rabbits, and no changes in the gonads. Wulzen (14) noted in chickens fed fresh hypophysis that the long bones were shorter than in her control animals, and that the involution of the thymus was hastened by the treatment. Pineal experiments In experiments on growth, Dana and Berkeley (713) reported that after injecting and feeding a few young animals with pineal substance, they noted growth acceleration. These results may be due to the pineal medication or merely to normal variation. Berkeley (’14) reported mental and physical improvement in backward children to whom pineal substance was administered. Priore (’15) reports that young rabbits into which pineal ex- tract was injected frequently, grew slightly more than the con- trols, but his results are probably within normal variability. In an older group the controls outgrew the experimental group. In none could any skeletal alterations be seen. McCord (’14) fed pineal substance to a few chicks, young guinea pigs and pups. He reported an acceleration in their rate of growth, increased mentality in the dogs, and sexual precocity in some of the guinea pigs. The same writer (’15) later reported that the growth of young guinea pigs which received daily 10 mgms. of pineal substance from adult cattle was delayed. Adult guinea pigs were not affected by calf pineal substance (dried), but the growth rate — GROWTH AFFECTED BY FEEDING DUCTLESS GLANDS 301 of young guinea pigs was accelerated by injections of extract of calf pineals. Pineal substance hastened sexual development in guinea pigs. Males were more susceptible than females to pineal treatment. The treated animals did not grow larger than normal adults and their growth was apparently propor- . tional in all parts of the body. The ‘sexual precocity’ noted by McCord perhaps can be explained by the fact that in grow- ing animals the sexual development normally tends to keep pace with somatic growth. In stunted animals sexual maturity is retarded as seen in inanition experiments (Jackson 715); con- versely it is to be expected that a hastened somatic growth is accompanied by a corresponding development of the reproduc- tive organs. This condition can hardly be called sexual pre- cocity, especially if the sex organs are not relatively over- developed. Dandy (’15) was able to remove the pineal body from young dogs with no apparent after-effects. Many experimental data that have been published are worth very little to us because of incomplete records and also because animals of one strain kept under certain conditions have been used as control animals and checked against experimental ani- mals of a possibly different strain kept perhaps under different conditions. It is also true that in many experiments, although all the animals are of the same strain and are kept under similar conditions, only a small number are used and these represent different litters. Since it has been shown by Jackson (’13) and King (15) that in albino rats variability in body weight within a litter is only about half as great as general racial variability, the advantage of taking controls and experimental animals from the same litter is obvious. The apparent ‘results’ obtained in many experiments may very well be due merely to the above- mentioned factors and not to the experiment itself. III. MATERIAL AND METHODS The albino rat (Mus norvegicus albinus) was selected for the experiment because it is a convenient form for use and more is known about the growth of this animal than of any other, ow- ing to the work of Donaldson, Hatai, Jackson, Slonaker, Lowrey, 302 E. R. HOSKINS King, and others. Owing perhaps to their diet and very favor- able environment, most of the animals (especially the younger male groups) are somewhat larger than the average ordinary albino rats at corresponding ages. Twenty-nine litters were used but-where litters contained less than four rats of a sex these individuals were usually rejected. A few rats were killed by the mother before weaning time. In all, 59 females and 73 males were fed and of these all of the females and 59 of the males were carefully autopsied at the termination of the experi- ment. While the number of observations is not large from the statistical point of view, and more would be necessary for final conclusions, it is believed that the present data are suf- ficient to establish certain points with a considerable degree of probability, and to furnish valuable evidence upon other points requiring further data. Most of the older animals which were born during the summer and early fall were smaller at the be- ginning of the experiment than were those born in the winter and spring. This is in accordance with previous observations. These initial differences generally persisted throughout the period of experiment (as likewise found by King (715). Of the older litters, two of which were of purely local stock were quite large rats, whereas three litters which were of pure ‘Wistar’ stock were considerably smaller. Other litters used were of a mixture of these two strains. ; The rats with few exceptions were weaned at 3 weeks of age, and kept in a well ventilated room in fairly large wire net cages, with wire net bottoms which allowed waste matter to drop through. ‘The males and females were of course separated. All were fed (ad libitum) upon whole wheat (Graham) bread soaked in whole milk, a diet which seems to provide abundant nourish- ment, as shown by the rapid growth of the animals. During the first part of the investigation the rats were fed once a day and their water jars washed once a week. Later all of the ani- mals were fed 3 times a day, the water jars washed daily and at all times the cages were kept clean. Each animal was given a mark of identification with picric acid and a separate growth record kept for it. In general, each rat was weighed at wean- GROWTH AFFECTED BY FEEDING DUCTLESS GLANDS 303 ing time, and each day (before feeding) for about a week, but thereafter the interval between weighings was gradually in- creased. The autopsy technique employed by Jackson (138), with a few modifications, was used. ‘The various organs were placed in a moist chamber when taken from the animal and weighed when all had been removed. ‘The thyroid and thymus were freed from their capsules. In the younger groups of animals the mesentery and pancreas were removed from the stomach and intestines. These cases are marked (c) in the tables. All the organs were weighed in closed containers to 0.1 mgm. The skeleton was prepared by heating the body (eviscerated and skinned) at 90 degrees centigrade for 2 or 3 hours in 2 per cent aqueous ‘Gold Dust’ (a proprietary soap powder) solution. The skeleton (including cartilages) was cleaned, drained care- fully, and weighed, and then dried for 2 weeks (in an oven heated at 88 to 90 degrees centigrade) to constant weight and weighed again. This technique gives fairly constant results. In the final averages shown in the various tables, extreme data are in a few cases excluded. (These extreme cases were probably due either to experimental error or to abnormal variation.) The tables show only the averages of individual data. A copy of the original observations have been filed at The Wistar Insti- tute of Anatomy and Biology, Philadelphia, where they may be consulted by those interested. The material for feeding was obtained every 2 weeks by the author in person, from newly-killed calves 6 to 10 weeks of age, and ground fine in a kitchen meat-grinder. Some of the sub- stance of each kind was spread out thin and dried before an electric fan at room temperature. The material was quite dry within 5 to 10 hours of the death of the calves. It was diluted with known amounts of milk sugar for measuring. A portion of each kind was kept fresh (at from zero to 5 degrees centigrade). No constant difference in the effects produced by the fresh and dry glands was noticeable, so in grouping the 304 E. R. HOSKINS data the difference in the two conditions of the material adminis- tered was disregarded. The rats used, were grouped to exclude so far as possible the error introduced by the racial variability. One rat of each litter (or one of each sex if both males and females of the litter were used) was kept for control, the remaining being distributed among the groups treated, in such manner that (so far as possible) each group contained individuals from every litter. There are five groups of each sex. To eliminate the ‘age’ factor, the data for each sex were subdivided into ‘older’ and ‘younger’ groups, as shown in the tables. In calculating the results, two methods were used. In one the average of the percentage of the net body weight of each organ of each group of experimental animals was compared directly with the average value of that organ in the control group while in the second method a comparison was made with the Wistar reference tables as suggested by Donaldson (715). In the latter comparison the gross body weight was used instead of the length because the weight of the organs of the controls corresponded somewhat more closely to Donaldson’s Wistar norm for rats of the same gross body weight than to those of the same length. The values obtained, however, would have been practically the same in either case. The tables published by Donaldson (’15) used here are desig- nated as ‘Donaldson’s Wistar tables’ or the ‘Wistar norms.” For the dosage employed see table 3. The experiments began when the rats were weaned, at the age of about 3 weeks, excepting 15 rats which were 8 to 11 weeks old. Of these 15 rats, only 4 were autopsied. Each rat was treated on alternate days throughout the entire period of the experiment. Nine litters were given fresh glandular substance, and all others received dried glands. The material to be fed was mixed with a small amount of bread and milk and the ani- mal kept in an individual cage until all of this had been eaten. GROWTH AFFECTED BY FEEDING DUCTLESS GLANDS 305 IV. ORIGINAL OBSERVATIONS 1. Body as a whole (weight and length) a. Comparison of controls with norms. The growth of the younger groups of rats, especially the males, in this investiga- tion varied considerably from that of the rats described by most of the previous workers for rats collected at random from a colony. It is necessary therefore to make a comparison of the animals with the norms of other investigators before discussing the effects of the ductless gland feeding. In this comparison, only the control rats will be considered directly. With the exception of certain data of the thyroid-fed groups which will be discussed later, however, there is a close agreement among the corresponding data of all the groups so that the data of the control animals represent indirectly most of the data of the entire series. A great difference between my rats (especially the ‘younger’ or ‘winter-born’ males) and those previously described by Don- aldson and Jackson is to be noticed in the rate of growth both of body length and of body weight. The albino rats described by Donaldson (’06) and by Jackson (’13 and ’15) of different strain and different diet are considerably lighter in weight than are mine at corresponding ages, up to the fourth month. (See table 1 and charts 1 and 2). The selected ‘strong and vigorous’ litters described by King (’15), however, correspond rather closely with mine in body weight. The rate of growth varies thus in albino rats from different sources, depending partly upon the ‘strain,’ but more upon diet and general environment. King’s results indicate also that a more rapid growth may be expected from those litters in which at birth the individuals are especially large and strong. The vigorous average growth of my rats appears chiefly in those designated as the ‘younger’ (‘winter-born’) litters. The growth of the ‘older’ (‘summer-born’) animals throughout is more nearly like that found by Donaldson and Jackson (charts 1 and 2; table 1). As the ‘younger’ group had been eliminated largely at 110 (and partly at 70) days, the final averages are relatively 306 E. R. HOSKINS TABLE 1 Average gross body-weight of normal albino rats at various ages (in comparison with data from Donaldson, Jackson and King), showing variability probably due to various causes. AGE DAYS (gonencue KING (715)! JACKSON (713) DONALDSON (’06) grams grams grams grams 20-21 (1m)? 32.6] (50m) 32.0 | (53m) 24.0 | (19m) 21.2 (10f) 29.6)" (50f) » 2820 (59f) 21.5 (17i), 22:6 30-31 (18m) 46.4] (50m) 48.5 (19m) 31.8 (16f) 44.4) (50f) 45.7 (17f) 32.9 40 (18m) 69.5 (16f) 70.1 42-43 | @8m) 80.9] (50m) 78.0 | (45m) 63.7 | (19m) 46.3 (16f) 76.4] (50f) 70.0 | (50f) 64.3 | (11f) 47.9 70 (18m) 164.2} (50m) 143.0 (23m) 130.4 (19m) 106.6 (16f) 127.3] (50f) 123.0 | (25f) 108.9 | (11f) 99.8 90 (15m) 184.6] (50m) 184.8 (13f) 143.3| (39f) 148.0 110-112 (lim) 205.9] (50m) 214.0 (19m) 183.8 (10) 151.6| (42f) 166.0 (11f) 160.2 150 (5m) 221.53) (50m) 243.0 (20m) 167.5 (19m) 225.4 (8f) 164.6 | (45f) 185.0 (21f) 142.1 (11f) 184.6 1 King’s data partly from her table 3, partly estimated from her chart 3. 2 The 21 day group includes only the larger ‘winter-born’ rats. 3 The 150 day group includes only the smaller ‘summer-born’ group. lower. At 150 days (table 1) the averages are slightly less than Donaldson’s and King’s, but somewhat greater than Jackson’s. i As shown in table 1, the females at six weeks (40 days) aver- age heavier than the males in the control group. This agrees with the results of Donaldson (’06) and Jackson (’13), but not with King (’15). The difference in the rate of growth of the body length and tail length of my rats, as compared with the data of Jackson (15) and of Donaldson’s Wistar tables is shown in table 2. It may be seen that my rats averaging 13 weeks old are longer than Jackson’s rats at 5 to 13 months. Had the table included my entire autopsied series of 59 males and 59 females the differ- ence would have been slightly greater, as seen in tables 6 and 7. The ratio between the tail length and the body length is dif- GROWTH AFFECTED BY FEEDING DUCTLESS GLANDS 307 2 Vis 3 4 a % o- Chart 1 A graphic comparison of the data of various investigators with those of Hoskins for the growth of normal female albino rats. The weight in grams is plotted against the ageindays. D., Donaldson ’06, J., Jackson 713; K., King 715; H., Hoskins’ older (low dosage) group of control female rats and H', Hoskins’ younger (high dosage) group of control female rats. The sudden flattening of the graphs H! and H after 70 and 90 days respectively is due largely to the fact that at these points autopsy of rats was begun. The curves H! and H are not closely comparable with the others after these two points, because in most cases the largest rats were killed first. Chart 2 The same as in Chart 1, except that it is for normal males. further explanation. see that oiven for chart 1. For GROWTH AFFECTED BY FEEDING DUCTLESS GLANDS 309 TABLE 2 Comparison of data for younger and older groups of control (muscle-fed) albino rats with Jackson’s data and Donaldson’s tables, showing growth of body length and tail length. HOSKINS 4 2 ieee asso Vinal aa CAPES Younger group Older group 5-13 Mo. BODY LENGTHS) (av. 13 weeks) (av. 28 weeks) Number and sex.....| 8F*| 10M*| 8F.| 6M. | 34F. | 16M. F. M. Body length, em.....} 18.7.| 20.3 | 19.4 | 21.2 | 18.3 | 19.0 |18.7-19.4)20.3-21.2 Tail length, cm......| 16.0 | 16.5 | 16.5 | 16.6 | 16.8 | 16.2 |16.4-17.0|17.2-18.0 Tail Body- Ratio....| 0.86) 0.81} 0.85) 0.78} 0.90) 0.86 0.88 0.85 * Tf all the thirteen weeks old rats on experiment are included, the number of females and males changes to 28 and 29 respectively, and the lengths of the body and of the tail are but very slightly changed. (See tables 7 and 8.) ferent in the different series. In all of my groups of rats the tails are relatively shorter than those of Jackson’s rats, and than those at corresponding body lengths in Donaldson’s tables. If, instead of the age, the body weight is taken for the basis of comparison, a similar difference in the ratio of the tail length to the body length of the two series is evident. The growth of the albino rat in weight and length under dif- ferent circumstances thus varies considerably. A norm must therefore be established not only for each strain but also for each litter under a given set of environmental conditions. If, however, similar litters from the same strain are kept under simi- lar conditions variability will be at a minimum. For a comparison of the various organs and parts of the rats with those described by previous writers, data are shown in tables 8 and 9 for relative (percentage) weights, and in tables 4, 5, 6 and 7 for absolute weights. Data from Donaldson’s Wistar tables for rats of corresponding body weights and lengths are included in tables 4 and 5. The body weight in general is slightly greater in the Wistar norms than in my rats of corre- sponding body length, excepting the younger males. As to the individual organs, it is evident that in some cases the weights in my series are nearer to those of Donaldson’s tables at correspond- ing body lengths, while in others they are nearer those of cor- responding body weight. Some differences are due probably - 310 E. R. HOSKINS to age. On the whole, the correspondence with the Wistar tables is as close as could be expected. The individual organs will be considered later. b. Effects of thyroid, thymus, hypophysis and pineal feeding. The weight and length of the growing albino rats fed various ductless glands are shown in tables 6 and 7, and (for the ‘higher dosage’ groups) in charts 3 and 4. For the sake of elimination of any variation that might be due to the age of the rats, each sex group was subdivided into 2 smaller (‘old’ and ‘young’) groups, depending upon the age of the individuals. Some of the younger animals received fresh and some dried glands (see ‘Material and Methods’), but no difference was noticed in the effects produced by the two forms. The effect of a ductless gland diet upon the growth of the females is seen in table 6 and chart 3. The various experimental groups may be compared with each other or with the controls, and it is found that the difference in weight at every age is remarkably slight. At the beginning of the experiment when the rats were 3 weeks old the different groups averaged nearly the same in weight, excepting the male thyroid group and the pineal groups. Into these groups were purposely placed slightly more than their share of smaller animals because it has recently been claimed that thyroid and pineal substances accelerate growth in various species (Schafer 712) (Dana and Berkeley "13, McCord 714). At 70 days of age, when the period of most rapid growth had ended, it is seen that among the younger (‘high- er dosage’) animals there had been a remarkably small difference in the growth rate of the various groups. The same is true of the older rats, if the thyroid group (which contains 2 rats that were not healthy) is left out of consideration. At 90 days of age the weights of the different groups still remain fairly close together. After 70 days the groups are no longer directly comparable because many of the rats had been killed. Upon the comparison of individuals within each litter, no constant difference appears, although considerable variation is shown. The only probable conclusion to be drawn is that the glandular substances (in 200 GROWTH AFFECTED BY FEEDING DUCTLESS GLANDS dll Chart 3 Graphic representation of the growth of the ‘younger’ or ‘high dosage’ female albino rats to which ductless glands were fed. The weight in grams is plotted against the age in days. 1, Thyroid-fed; 2, Thymus-fed; 3, Muscle-fed (controls; see also chart 1); 4, Hypophysis-fed; 5, Pineal-fed. Note how closely the various groups agree in weight. At 70 days, autopsy of the larg- est rats was begun, hence the greater variation after this_point. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 21, NO. 3 the graphs between 21 and 25 days (as compared with those of the females, chart 3) possibly indicates that males are more susceptible to environmental changes than are females. The rats were weaned at 21 days. 312 ’ GROWTH AFFECTED BY FEEDING DUCTLESS GLANDS als the amounts fed) had no effect upon the weight of the female albino rats in this experiment. A graphic representation of the growth of the ‘higher dosage’ group of females is shown in chart 3. The body weights in all cases remain close to those of the controls (muscle-fed). In the ‘higher dosage’ groups, the body weights (chart 1) are seen to be much higher than the normals of Jackson and also of Don- aldson (excepting near the end of the experiment). In the ‘lower dosage’ groups the body weights average lower, more nearly comparable to the normals of Jackson and Donaldson. This difference in body weight between the ‘higher dosage’ groups and the ‘lower dosage’ groups is not due to the different amounts of ductless glands fed, however, as a similar difference is shown by the controls in each group. A careful study of the growth of individuals within each litter shows that in nearly every in- stance those rats which at three weeks were larger (or smaller) than the controls retained the same relative position as regards body weight throughout the experiment. The male albino rats to which ductless glands were fed also seemed not to be affected in body weight by the treatment. A careful study of table 7 and chart 4 shows nearly the same facts for the male rats as have just been stated for the females. The weights of the different ‘higher dosage’ groups are unusually close together at 70 days of age, except in the case of the pineal- fed. The ‘higher dosage’ pineal-fed animals at this age are 16.5 grams lighter in weight than the controls; but this difference is not great, and as this group averaged less in weight than the controls at the beginning of the experiment, the difference in weight between the two is probably due to normal variation. In the ‘lower dosage’ group, the pineal-fed are slightly above the controls in weight. After 70 days of age many rats were autopsied and hence the groups are no longer directly com- parable, but individuals of the same litter were compared with each other and showed the same results as in the case of the females. On the whole, there appears to be no evidence indi- cating that the ductless gland feeding has naterially affected the body weight in any case. The differences are inconstant, and well within the limits of the variability to be expected. 314 E. R. HOSKINS It may be noted (see table 3) that thyroid was fed in varying amounts from the negligible quantity of 10 mgms. of dried gland on alternate days to a nearly maximum non-toxic dose of 200 mgms. of dried substance (or an equivalent amount of fresh gland) on alternate days. In all cases no appreciable TABLE 3 Amount and range of dosage employed for each rat in feeding the various groups. The growth in body weight for the ‘high dosage’ groups is represented in charts 8 and 4. DOSAGE EMPLOYED FOR SIZE OF SINGLE DOSE Females Males re Dried | Fresh age age sub- sub- |number| Range |number} Range stance | stance of of doses doses mgms. | mgms iy POld=teGie waa, eee as cece 10 40 SOWA OSACCr ce N94: hee eee A sare ac seme 4.3] 45] 1.6 | 0.52 pHiichtd Sage eerie... e cee: ce 11.7 | 10-20; 16.3 | 10-20 Syms ledser.. cy. oc Pathe Na adr) ed ee 15 70 BOWAGOSALC Renae eee to emer a2) | 2-0) 1.0 | 0f52 PE hi dOSageasen ane ees oh reer 16.7 | 10-40} 16.4 | 10-20 Miusele-feds (controls) ees nee coe 8 25 PIGOW: GOSADE crochet election Ae 3.8 | 2-5 | 2.4 | 015-5 “] 6 lifed ote KayspeqVin een 5 Ae AP in cee 15.8 | 10-40} 17.0 | 10-30 Fey pophysis-fedi:) 28). Sans Mes oe ah 5 25 OW COSA LCi tc mine eho ota Solano ene olor 3.9 | 2-5 | 2.3 | 0.5-5 Pro ht OSA@e’ smears. fee eae Lee es 12.5 | 10-20} 18.6 | 10-30 Rime alafedics., ceca peice coke AL ee eee 1 250\| 740 pO WIC OSAP CVA pier oie ae ad fae POE oe 3.84, 2-9'| 2.0) | 075=3 Brie OSAP ore acon Cae Saas eee 17.5 | 10-40} 138.3 | 10-20 effect upon the growth of the body as a whole was evident. Thy- mus was fed in variable amounts up to 300 mgms. of dried substance, but two individuals receiving twice this amount showed no effects different from the others. Hypophysis was administered-in fairly small doses (5 to 100 mgm. of dried sub- GROWTH AFFECTED BY FEEDING DUCTLESS GLANDS abe stance). Pineal substance was given in amounts larger than those fed to guinea pigs (in most cases) by McCord (’14), but in one of McCord’s experiments where doses of 100 mgms. of dried substance was used daily, the experimental animals in a given time grew in weight 40 per cent more than the controls. As stated above, there seemed to be no difference in the result with albino rats in the present experiment, whether small amounts of 20 mgms. or larger amounts up to 150 mgms. of dried pineal substance was fed on alternate days. With the larger doses there was no evidence indicating that toxic effects of gastro- intestinal disturbances were produced by the medication. _ As shown in tables 2, 6 and 7, the various rats of the same age and sex are also of nearly the same body-length regardless of treatment. The agreement here is even closer than in re- gard to comparative weights and emphasizes still more strongly the fact that the experimental rats as compared with the con- trols suffered no marked gross body changes on account of the administration of ductless glands. The growth records of the individual rats illustrate, as has been pointed out above, that very serious errors might easily creep into the conclusions from an investigation of this kind, which includes animals from several different litters. A preliminary comparison of the individuals in each litter showed negative results, so all were finally grouped as shown in the various tables. There is always, of course, some danger in drawing conclusions from averages, but this danger is slight if the individual data are also carefully studied, and the grouping judiciously made. Care must be taken in work of this nature to select experimental and control animals from the same litter and as nearly alike as possible. In many investigations on growth this has not been done. Experimental groups of animals of one sex have even been compared with groups of another sex; or, more com- monly, an experimental group has contained both sexes in a ratio different from that in the control group. Finally, the depressing effects upon growth and body weight obtained by some investigators by the administration of ductless glands especially the thyroid (Magnus-Levy, Bircher, Carlson, Far- 316 E. R. HOSKINS rant, Gudernatsch, Cotroni, Hewitt, Romeis, Lenhart), is in many cases possibly a general toxic effect, which is produced whenever the dosage is too high. Even a high protein diet, e.g., an ex- cessive meat diet, may likewise be detrimental to growth, as has been shown for the albino rat by Chalmers Watson (06). 2. Head Data for percentage weights of head (and other organs and parts) are found for females in table 8, and for males in table 9. Absolute weights are given in tables 6 and 7. a. Controls. The head forms an average of 9.5 per cent of the body weight in the females and 8.3 per cent in the males among my younger control rats. In the older groups, the head averages 10.3 per cent in the females, and 8.4 per cent in the males. These results are in general somewhat lower than those obtained by Jackson (’13, 715). b. Thyroid group. The head averages very slightly heavier in both females and males to which thyroid was fed. The dif- ference is probably insignificant. c. Thymus, hypophysis and pineal groups. The head in these groups shows no constant variation from the controls in either direction. The few small differences are probably not significant. 8. Eviscerated Body (Tables 6 to 9) a. Controls. The eviscerated body forms an average of 80 to 84 per cent of the net body weight in both males and females. This part of the body contains the muscles, skin, skeleton, body- fat, great vessels, lymph nodes, and spinal cord. b. Thyroid group. The eviscerated body in all rats of both sexes (excepting 3 old males, in which the dosage was slight) is about 4 per cent less in relative (percentage) weight than that of the controls. This loss is due probably to loss of fat, which is a well-known effect of thyroid-feeding, especially with high dosage. A comparison of the body weights and body lengths in the control (muscle-fed) and thyroid groups shows a GROWTH AFFECTED BY FEEDING DUCTLESS GLANDS 317 slight relative decrease in weight in the thyroid males of higher dosage, but not in the females. c. Thymus, hypophysis and pineal groups. In all these groups, the weight of the eviscerated body is very close to that of the controls (muscle-fed). The variations are slight and probably insignificant. 4. Integument (tables 6 to 9) a. Controls. The average relative (percentage) weight of the integument is relatively fairly close in the various groups of both sexes. In general, the percentages range between 21 and 24 per cent of the net body weight, the value for the male rats being slightly greater than for the females, due possibly to the presence of a greater amount of fat in the latter. This is some- what higher than that observed by Jackson and Lowrey (12), by more than 4 per cent of the entire net body weight, and is 2 per cent of the entire body weight higher than that observed by Jackson (15). These differences are probably due largely to the varying amount of fat (or muscle) present im the integument. b. Thyroid group. In the thyroid rats of each sex the integu- ment appears usually very slightly lighter in weight than in the controls, owing probably to loss of fat. Jackson (’15) has shown that during inanition the skin loses greatly in weight (probably due chiefly to loss of fat) but that the relative (percentage) weight remains unchanged in adults. c. Thymus, hypophysis and pineal groups. In these groups, the differences in weight of the integument, as compared within each group and with the controls (muscle-fed), are well within the limits of normal variation. 5. Cartilaginous skeleton (tables 6 to 9) a. Controls. The relative weight of the ‘wet’ cartilaginous skeleton averages about 7.6 per cent of the body weight in the older females, and about 6.6 per cent in the older males. The skeleton in the younger females averages about 6.8 per cent of 318 E. R. HOSKINS the net body weight and in the younger males about 6.1 per cent. These correspond fairly well with the estimate of Jackson (’15) which was 7 per cent of the adult net body weight, and with Con- row’s observations (cited by Donaldson 715, table 53). The dif- ference between the sexes is accounted for by the heavier body weight (with correspondingly lighter skeleton) in the male groups, rather than by any true sexual difference. The older groups of each sex on the average appear to have a relatively heavier skeleton than the younger. ‘This is con- trary to the general tendency of the skeleton during growth to lag behind in relative weight. In these groups, however, the differences in body weight are much less than usual for the cor- responding age differences, and the increased weight in the older skeletons is possibly due to more advanced stages of ossi- fication and calcification. Thus in two animals of the same body weight, the older apparently has a heavier skeleton. This tendency is not evident in Conrow’s data (cited by Donald- son 715), however. In this connection may be cited the observations of Jackson (15) who found that in young rats held at constant body weight by underfeeding the skeleton continues its development (dif- ferentiation and increase in wet and dry weights). It is there- fore probable that the relative weight of the skeleton depends somewhat upon the age factor, as skeletal growth is to some extent independent of the general growth of the body. The dry cartilaginous skeleton is likewise relatively slightly heavier in the female (3.7 to 4.6 per cent) than in the male (3.2 to 3.8 per cent), and the older rats of both sexes have relatively heavier dry skeletons than the younger. ‘The explanation for this is doubtless the same as that above given for similar rela- tions in the weights of the wet skeleton in different groups. There is considerable variation shown. by individuals, so that the average values for both wet and dry skeleton can be con- ‘sidered as only approximate. Differences in the technique may also modify the skeletal weight considerably. For example, Conrow’s data for the dry skeleton (cited in Donaldson’s table 53) are too high, because her method of drying at room tempera- GROWTH AFFECTED BY FEEDING DUCTLESS GLANDS 319 ture is inadequate to remove all the moisture. More data are needed to establish a satisfactory norm. b. Thyroid groups. The skeleton appears slightly heavier for the thyroid treatment in the average of the younger groups. A comparison of the individual data shows that both wet and dry skeletons average somewhat heavier in most of the rats receiving higher dosage of thyroid than in those of low dosage and controls. This suggests that the thyroid treatment may perhaps tend to stimulate skeletal development, as claimed by Bircher ('10b). However, the possibility of errors from ac- cidental variations must be kept in mind, as the differences found are not very great, and were not constant in every litter. c. Thymus, hypophysis, and pineal groups. The various groups fed thymus, hypophysis and pineal glands show no im- portant or constant variation from the controls in regard to the skeleton. If any effect was produced, it is so slight as to be masked by normal variation. 6. Brain (tables 4 to 10) » a. Controls. As shown by tables 4 and 5, the absolute weight of the brain does not vary greatly from that shown by Donald. son’s tables for rats of similar sex and body weight or body length. In percentage weights, as might be expected, the males have relatively lighter brains than the females, and the older brains average lighter than the younger, on account of the fact that the brain lags behind in the growth of the body. b. Thyroid groups. On comparing directly the average rel- ative (percentage) weights (tables 8 and 9) of the controls (mus- cle-fed) and thyroid-fed animals, the brain in the latter appears slightly heavier in all except the younger group of males; but on comparing the absolute weights with the Wistar norms for animals of corresponding body weight, it is found that the dif- ference in the various litters is not constant and the brain of the thyroid-fed rats averages even smaller than the normal (table 10). 320 E. R. HOSKINS c. Thymus, hypophysis, and pineal groups. There is no constant variation of the brain in these groups. to be a slight tendency to increase in the brains of the hypo- physis-fed group (less marked than in the case of the thyroid), but the difference is not constant. 7. Eyeballs (tables 4 to 10) There appears a. Controls. The absolute weight of the eyeballs in my ani- mals corresponds fairly closely to that of Donaldson’s norms (ex- cepting the younger males), as shown in tables 4 and 5. In comparing the relative weights of groups of each sex (tables 8 and 9) the eyeballs appear relatively heavier in the older groups, although in these, with heavier body weight, the eyeballs would be expected to be relatively lighter. As suggested by Jackson TABLE 4 Comparison of principal data for control female albino rats with the Wistar tables of Donaldson (’15). Weight is the average expressed in grams. Comparison is made with Wistar norms both of the same body length and of the same gross body weight. HOSKINS (OLDER) WISTAR (SAME LENGTH) Body length, em..... Gross body weight, g. al Rabon. seer LNG GERD ceobes bcce ¢ Organs: Biveloallllsaene eee Hearts.) 5c eee Alimentary Tract. . WOvamlesea) Laas Hypophysis....... | Suprarenals....... | AU one cose oot grams 1.781 276 737 379 530 751 900 966 053 O11 .049 .134) wea 19.4 188.5 0.88 grams 1.823 0.263 0.765 1.622 9.600 0.505 1EUL9 9.680 0.049 0.015 0.050 (0.153) WISTAR (SAME BODY WT.) 18.9 172.6 grams 1.801 0.252 0.715 1.503 9.000 0.465 1.040 9.120 0.048 0.012 0.046 HOSKINS (YOUNGER) 18.7 159.8 0.86 92 grams 1.782 0.223 0.694 1.589 7.405 0.611 0.982 0.064 0.010 0.049 (0.264) WISTAR (SAME LENGTH) 0.045 (0.278) WISTAR (SAME BODY WEIGHT) 18.5 160.8 grams 1.782 0.244 0.677 1.414 8.540 0.435 0.981 0.047 0.011 0.044 * Thymus compared with regard to age (213 days average for the older group, and 92 days for the younger group) instead of regard to body length or weight. GROWTH AFFECTED BY FEEDING DUCTLESS GLANDS 321 TABLE 5 Comparison of principal data for control male albino rats with the Wistar Tables of Donaldson (15). Weight is the average expressed in grams. Comparison is made with the Wistar norms of the same body length and of the same gross body weight. WISTAR WISTAR WISTAR WISTAR Ce eee ee, | OO all aetene |e Body length, cm..... 21.2 21.2 21.1 20.3 20.3 20.6 Gross body weight, g. | 2382.5 238.1 234.1 214.9 204.3 215.0 Parl Raiios: line se oe: 0.783 0.85 0.85 0.813 0.85 0.85 Ig ex (anys) seme aos: 193 90 Organs a grams grams grams grams grams grams [3 Se Ais ea ea 1.852 1.911 1.907 1.909 1.872 1.885 Biyeballs:. 22: 552. 0.286 0.294 0.291 0.220 0.273 0.280 PACE be see epee res 0.893 0.918 0.905 0.988 0.814 0.847 KVne ys eae ok. 1.867 1.992 1.962 2.050 1.740 1.820 iver...) oan! 10.701 | 11.440! 11.290 | 10.915 | 10.200 |) 10.590 Spleen iets! 55): 0.959 0.630 0.620 0.895 0.545 0.572 LSPS aA eae eae B 1.813 1.365 1.346 1.199 1.198 1251 Alimentary Tract. . OFS pie ie s60) ate 230 REESE oe a: 22222 2.397 Prey (d 2.248 2.203 2.267 Etypophysis: |...) 02: : 0.009 0.009 0.008 0.008 0.008 0.008 Suprarenals........ 0.031 0.038 0.038 0.031 0.035 0.036 MMV MUS i... os (0.155) | (0.170) (0.298) | (0.283) * Thymus compared with regard to age (193 days average for older group, and 90 days average for younger group) instead of regard to body length or weight. (13), however, the growth of the eyeballs may tend to be cor- related with age, rather than with body weight (as is known to occur in the thymus). These differences may therefore be due to age changes, to normal variability, or perhaps merely to differences in the technique of removal of the eyeballs. b. Thyroid groups. The eyeballs average heavier in relative weight than those of the controls in both sexes, but the differences obtained are slight and probably of no significance. Comparison according to the method of Donaldson (’15) shows them slightly lighter in weight in the males than those of the corresponding controls, but in the females slightly heavier. c. Thymus, hypophysis and pineal groups. In most cases, the eyeballs in these groups average relatively slightly heavier than those of the controls. No particular significance other than variability is attached to this fact. S22, E. R. HOSKINS 8. Thyroid Gland (tables 6 to 9) a. Controls. The thyroid of my rats cannot be compared directly with that of other investigators owing tothe different technique with which it was removed from the body, as de- scribed in ‘Material and Methods.’ The weight is about one- third less than the Wistar norm, probably on account of removal. of the capsule. There is also extreme variation in the weight of this gland even in different members of the same litter. Usually it appears relatively heavier in the older than in the younger rats, and also slightly heavier in females. Jackson ’(13) and Hatai (’13) also have found the thyroid gland to be exceedingly variable, so no final conclusions can be drawn as to the normal weight of this gland. b. Thyroid groups. The thyroid gland (tables 6 to 9) shows no constant changes as a result of the thyroid feeding. Any effect if produced is hidden by the great normal variability. In view of the great variability in the weight of this organ, final conclusions in regard to the effect of thyroid feeding upon the weight of the thyroid gland are not justified from the available observations. c. Thymus, hypophysis, and pineal groups. In these groups likewise the thyroid appears variable when compared with the controls, and the results are not sufficiently marked or con- stant to warrant any conclusion regarding the effects upon the weight of the thyroid gland of feeding these substances. 9. Thymus (tables 4 to 9) a. Controls. The thymus also is not to be compared very closely with that described by previous workers. The usual method of comparison considers the weight of the gland at dif- ferent ages. Theoretically, the organ in the albino rat in- creases in size gradually until at about 85 days it reaches its maximum weight of 0.29 grams (Hatai’14), but as is well known, many conditions influence the involution of this organ. Its weight varies considerably in rats of the same size and age even when the animals are apparently normal and kept under nearly ‘sporouvd pue Sol10j}UaSeUI JO Posed] 1 2 STS OCA ET SOUTJSOJUI-TOVUIO}G etette te. fekelinteiemelie Tare) s uaa{dg AS au leinee nar 10} 0]0¥s Snoulse[Iyavo AIG Sin Sonees - m040a4s SNOUISETZIBO JOAA iio camo G88 jyueuInso4uT wae “ a Gen” a & Py oF 01 | SO'ZI DH fe) Z 66 ecg a 6 LIL | 6°SPT FA 0'FT € 91 iS suvs6 suoio ee 0°02 6°11 < 9ST 6 ST i 81 61 2s eset | 2° P9T O Z-Ori se Cul i C6 FIZ G l sunox PIO qdad-1IVaNid F600 0 TOTO 0 1100'0 ~=|z100'0 cT9 0 OES 0 19 T 686 1 8Fo0'0 =|€6F0'0 IPG €& T61 2 vEs 0 €€9 0 VSP 2 1¥6 9 1160 L¥0°G 919 0 £99 °0 06a 0 CSI 0 TS¢10°0 = |2610°0 866 0 9160 908 1 VIL T v9 ¢ GO 2 6G OT v6 IT 0°CE G IE F 9CT P 8cl v FI 9ST swmnib sunub 0GL 8 GI 0°9T € ST 9ST 8 81 Pv Chl 8 IST 9 6FT jE CE 66 961 ¢ 9 sunox PIO dad-sisxAHdOdAH “UTD “YASUG] [TB wo ‘Yysue] Apog "++ 8 Qystam Apo 48N “* “3 “Qystem Apoq ssoiy 0010 0 I1T0°0 ~—-|8800°0 10100 11000 T100'0 - |é@100°0 T1000 F900 6900 F190 9h 0 6891 62E 1 PLE T 029 T 98700 |Sst0'0 |FSF0'0 |\§6E0'°0 1696 G 9969 Meal ts 8699 119 0 Tg2 0 LES 0 F6L 0 COP 2 ogg 2 199° 8 €F6 9 686 0 006 'T cr6 0 88¢ T 769 0 LEZ 0 9S9°0 619 0 F9G 0 Tél 0 Ice 0 FL0°0 S100 L10°0 810°0 FLO 0 €66 0 916 0 L110 Glo 0 682 T 182 1 GrL I €89 T 1g¢ LVL 66 G 699 6€ OT ST IT F6 2 88 OL Me Ws i he Onis FE L861 ¢ Srl 8 SEI ¢ &cl Wt (sl 6 &1 $8 GT sunib sunib - sunib sulouB rs} (IE io) UI Boyt Lae 0 91 ON 9° ST P OL L’8T v 61 G81 6 81 6 GST Vv SOL I 6F1 § ISI 6 6ST 6 TZT § 9ST 0 8ST 66 §16 88 GIG 8 8 9 g uno x PIO SUnOK PIO (STOULNOD) dad-aINsAW ddd-SOWAHL 8800°0 |&600' 0 G100 0 T1000 6S9 0 967 0 668 T 916 T Gcg0'0 =|2790°0 1GLT € GW 2 6&9 0 O16 0 996 6 €8E 6 1S8 0 L16 T 192 °0 OS8 0 GEG O OGL 0 FL 0 L100 I1z 0 916° 0 (Os, |I WAVE | 67 ¢ 69 °9 [66 VG II GCE Ge [lhe L SCT 9 ZCT i? tal! 9ST suMnsb sUnib CGI 98 0 OL 6 ST Pv St 6 61 6 OFT L691 § STI G O9T 08 L0G v ¢ suno x PIO dadad-dqilowxA HL (SABP) 05% ISBIOAY mentite s}wa Jo taqunN paisdojnn uaym abo pup jarp 07 burp.ovan padnoib syou ourgiy ur sztod fo (swpib U2) qybram aynjosqv abowaay :sayowag 9 ATAVL HOSKINS E. R. 200°0 F100 0 6890 T80°% 92951 £9600 116 € ZTZ°0 802 6 GOP T ¥28'0 G2 0 Tc0°0 802 0 661° T L0°9 LEVI v &P ¢ 6S1 Gil sunub 0ST 6 ST Fy 61 8 PSI 2 G6L 8200 °0 c100°0 9€Z°0 OSh'G COTS c080 0 6176 602 °0 666° TT 800 'T L980 6E1 0 €c0 0 LGZ0 608 'T 99°8 &I tI 6 TS & S06 6°81 swnib BL 0 LT GIG € L6G 9 €&% qdd-TVaNId €200°0 é100°0 L¥S°0 606 'T 626 'T 6080°0 180P & 898°0 €92 OT 160'T TS8°0 &6& 0 0z0'0 01z'0 008 'T c0'9 6? IT Me tes) Z 9ST 0 '9T sunib 1°06 6 91 ZL 61 Gg L6I L° 806 é800°0 F100 '0 6L9°0 L¥0°S 168 T 18200 G68 8 6880 166 OT v96'T 968 °0 €ST 0 &c0°0 L160 9F8'T Ww 8 L9°&T G OV 9 LLT L8t Swn.b € 01 8°91 0°16 v Sic add-SistHdOdAH 6200°0 =|¢800°0 F100'0 =F T1000 &19°0 €S2°0 ShG'G GGG G 090 'G L98°T ITg0'0 =|90€0°0 1460 F T9€°6 c68 0 896 0 166 OT TOL OT 661 T G99 T 8860 €68'0 86 0 SST 0 910°0 8600 06z 0 18z 0 606° T GS8 1 0L°9 698 LL GL 69 FI G LY eh 9 TZ1 6 PSI Ge 8 81 sun. suinub 0 ZT 0 OT GOL 9 OT & 0G GIG €& 906 €& G63 6 FIG 0 SES 06 661 OL 9 BunoX PIO (SIOULNOD) aaXd-a1NsoOW €800 0 €100'0 ¥29 0 PLE % 980°G 9¢80°0 10PL '€ T8000 c100 0 r6L'0 ballte F68 TI 0F80'0 6896 962°0 c8I OT 966 'T 8860 0Gz 0 € Not published, 362 E. I. WERBER to imagine that some free lentoids found in such embryos may owe their origin to heteromorphosis from dispersed optic cup fragments. However, it seems doubtful whether fully differ- entiated free lenses of large size originate from this source. While, of course, this possibility cannot be entirely disregarded, the evidence points to the conclusion that the fully differentiated free lenses found in some teratophthalmic embryos are due to a chemical stimulus of blastolyzed potential optic cup substance on any part of the ectoderm with which it may chance to come into contact. It is quite possible that this stimulus (the lento- genic reaction) is in the nature of a catalytic reaction such as was assumed by Herbst (’01) to underlie the differentiation of the secondary sexual characters owing to products of internal secretion of the sex glands. Reactions of this nature (autocatalysis) have in recent years been assumed by J. Loeb (’02, ’09), Robertson (’08) and Hage- . doorn (711) and others to underlie the mechanisms of develop- ment, growth and inheritance. According to Hagedoorn the hereditary (genetic) factors are to be regarded as autocatalytic substances. 150. 366 ORIGIN OF INDEPENDENT LENSES PLATE 2 E. I. WERBER THE PHYSIOLOGY. OF CELL-DIVISION VI. RHYTHMICAL CHANGES IN THE RESISTANCE OF THE DIVIDING SEA-URCHIN EGG TO HYPOTONIC SEA WATER AND THEIR PHYSIOLOGICAL SIGNIFICANCE RALPH S. LILLIE Clark University, Worcester, Mass. INTRODUCTION In experiments performed at Woods Hole during the sum- mer of 1901 Lyon! found that dividing Arbacia eggs varied greatly in their susceptibility to cyanide poisoning at different periods of the cell-division cycle. Eggs placed in cyanide-containing sea-water (m/100 to m/200 KCN) some time previously to the first cleavage, but not too soon (later than 15 or 20 minutes) after fertilization, withstood exposures of several hours without losing the power of development; while eggs exposed to the same solution at the time of cytoplasmic division were promptly killed. After the completion of cleavage a return of resistance was observed; this was followed by a second decline at the time of the second cleavage. It has been shown by Loeb? that the resistance to cyanide poisoning is much greater in the unfer- tilized than in the fertilized egg; Lyon found that fertilization is immediately succeeded by a period of high susceptibility, lasting some ten or fifteen minutes; then follows a resistant period which is terminated by the first cleavage. At first it was uncertain whether the resistance reached its minimum during or immediately after cleavage; Lyon inferred the latter, and supposed that the cyanide acted by preventing the oxida- tions necessary for the nuclear resynthesis following cytoplas- mic division. |SI-Ol |Sé-se es Saal! c> |OT 8°0F-0€ \S9-GS 0 “89; 0 “BO eS I> | SS. |St-01 Gi-O1 | -S8-8c | 08 “SO |G0-GS |S8-GZ |S6 “8 ¢°3o | CI-OL | G&-Ge [Sh-GE 08-02 (02-09 €-Z “B9| OT °89) GI-OL | OZ@-ST \0&-02 \0Z-09 08-02 oe |S Wee g"eo (0T-S (0r-0e | O¢< GI-OT | $2-02 | GE-G% (08-0Z | 96 “89 +06 CI-OT | $2-02 08-S2'89 | 0O6< 9-09 os< “Ur 09 | “W 8¢ “UL 9G “UL pg | “Ur gg | “Ur os 0¢-0F OF-0 |ST-O0T | ¢ "89 | 0 0 | 0 | ‘wogs °:yuV80 tad oun Geen | aos Ts. | ¢ 0 0 | oO | ‘WE | -104¢°%¢ “Ge “Bny 06 ‘89 |06-08 |0¢ *Bo| ¢-vo | ¢g-~ | ZT] O | ‘Woss **yue0 sod own 0€-02 | OT 89} T°B2} 0 0 0 | 0 | ue \ Toa gue “pe “BY 02-09 j0S-OF |¢e-4z | ¢°B0 | 0 0 | 0 | ‘wogs **yU00 xed ovun OI *B9] F-Z 89] €-Z 0 0 0-4) Gore. exc ec say, 06< 06-08 [02-09 | st-Se | e-2z 0z-01] 0 | “MOss ***}u00 red oun og ‘volog-oz jor-¢ | 0°%> | 0'w | O | O | “WE \ -104¢°%9 “IZ “Bny 02-09 |09-0¢ |0F-08 | OI-G | ¢ “80 Q | ‘wogs *°*yue0 sed oum OF-08 |SI-0T | ge | ZT | 0 0 | "mg | -10409 02 “Sny GS-08 |GF-GE | OT °B2] GS—P Bd] Z-T “BO 0 “Ul og f ‘*‘qu90 ted oun 0 | we \ -10409 “6I “Sny CE-—GS 0 | ‘woes **'gueo zed oun 0 | we \ -10409 ‘ST ‘Sny “UU Sh | “UOP | “Ul FP “UL GP “Ul Op | “USE | UOI-S UALVM-VaS JO NOIWOATNIG AUNSOdXT SULANIW 0F GNV ¢ UALAV GAZATOLAO SODA AO ADVINGOUAd ALVWIXOUddVY GNV (NOILVZITILURA UALAV SULANIWN) UALVM-VaS ALOATIdG NI PNIOVId JO AWIL ¥ ATAVL PHYSIOLOGY OF CELL-DIVISION 383 In the experiments cited in table 4 the period of maximum susceptibility varies between 48 and 54 minutes after fertiliza- tion; in each series there is a period corresponding to the change of form and lasting about four minutes, during which the sus- ceptibility remains about the same; the resistance then returns rapidly; its return evidently signifies the completion of the cleavage-process. With the most dilute sea-water used (62.5 per cent) a well marked increase of susceptibility is apparent twelve minutes before the period of maximum susceptibility. In most series a few eggs remain susceptible ten minutes after cleavage is complete in the great majority; these probably repre- resent chiefly the minority of slowly cleaving eggs; possibly, _ however, as already suggested, the recovered resistance is not quite equal to that of the uncleaved egg. Another and independent method of demonstrating these variations of resistance is by returning eggs to normal sea-water after exposure to the dilute medium for a definite period, and later determining the proportion which continue development to the blastula stage. In several series of such experiments the surviving eggs were always found much more numerous in those lots which had been exposed before cleavage or in the intervals between cleavages; while few if any eggs exposed dur- ing the formation of the furrow continued development. The five per cent or fewer eggs which formed blastulz in Experiments 6 to 9 B (table 5) undoubtedly represent the relatively resist- ant minority which had not yet begun to cleave, or had com- pleted cleavage, at the time of placing in the dilute sea-water. The degree to which the resistance changes at the time of cytoplasmic cleavage is perhaps best shown by comparing the effects of a graded series of dilutions upon the same lot of eggs, part of these being placed in the dilute sea-water well before the beginning of cleavage (about midway between fertilization and the first cleavage), and part at the time when the furrow is form- ing in most eggs. Such an experiment shows that at the height of susceptibility many eggs are destroyed by dilutions so low as 40 volumes per cent; while in order to cause an equally rapid cytolysis in the uncleaved eggs, dilutions of 65 to 70 per cent are LILLIE RALPH S§. 384 “BlNYSVl[q UIIO;J qu90 tod QZ 0} GI “BO *B[NISV[G ULLOJ 4ueo stad Qe 09 GZ ‘BO ‘BlNASV]q uloj yueo tod GE 04 Gz ‘Avp }xou pap 4SOj ‘pvop 4ysor oy} fBlnysvyq ur1ofy (ued tod OF 07 0g) Aplouru Vy ‘enyselq ur1o0j Ayrisolepy 10}8M -vas OyN[IpP 0} soynurur cy Jo einsodxe J9qje yuourdopsAeq *pozs]| -0j49 4u90 19d (0G 03 OF “yore -UI 4ysol oY} fpozAT -0j340 yue0 10d GF 04 GE “poR4y -UI 4Sod oY} {pazAo4 -£0 yu90 red YE 03 GZ "BO “qyouqutr Japureuiet !pezA]o4 -A0 Ju909 19d GZ 04 0% °BO ‘pozs] -0j49 Jueo aad GT 04 OT "BO £40B{UT 918 SOT “BlNYSBIq uo0y Aywofevur 4vary “eln4 -SB[q wWdoj |B ApIvON se[ny -Ssv[q wo} [eB ApTIwON 10}8M-BOS 9N[Ip UI S9yNurUL 0% J0qjye Ss509 jo UOlpZIpUuo_, royeM -89S O4N[Ip 0} SeyNUIU GT Jo ainsodxe Joye yusurdojeAeqd GOVAVEGIN ONIVAG UALVM-VaOS ALNTIG NI GHOVId SOOa “A “POR FUT [TV q084UT [TV “qoRqUT [[B INQ Wa[jomg 10}BM-B9S OJN{Ip UL SoyNUTUL 0% 10}J8 sdso JO Uor}IpuoD G@OVAVAIN GTUOIAA UALVM-VAS ALONG WI GHOVId SPOH “V 0g g Gly i cP & NG G OF I (MaLVM ae oy dvi NGO Uda “S10 A) MOLY M-Vas ALOTIG ao NOILISOdWOO ainsodxa saynurwm uaarfif azfo 1ayOM Das 07 ULNJaL UO H)NISD]g pamsof yoy? shba fo uwoy -wodoud ajounxosddn ay) osyp fainsodxa saynunu hyuan) 1ajfp paasasqo sp sbba ay, fo woYyrpuos ay} Uuoynjup Yana Lof saarb 279?) 9U.D, La {D LAJDN-DAS JOUWALOU 07 PAUINjas alan Sbba awos Yysip yoDA WOLY ‘inp yrau paurmusajap som yuaudojaaap burinurjuos sbhba asay) fo uoysodoud ay) ‘sajnuru uaaify fo ainsodxa up *S/DALAIUL 1D paasasgo SoM LajDM Das aynzip ay? UL SbBa ayy {0 uonrpuos ayy, “passafsuv.y fizupjpuurs asian (gq) .4od puovas v ‘sia ysow ur Burwurbaq som abnanaja wayn amy dD yw ‘(1Q°S]) 427M) saynunu aay-Ayuamy Suunjoo ysuy ay) ur uaarb suoyisodwuos ay} fo sajnm-vas aynjyp Bururwyuoo saysyp fo Salsas D 07 patwafsun.ly alan (PW) quod ‘1ajvM-vas JOULLOU U2 Sanu UaazxIS LafY “OG TD pazyysaf alan sbby “gg ysniny $ HIdVL 385 PHYSIOLOGY OF CELL-DIVISION ‘dojaAop ou0N ‘dojoAop ouo0N ‘elnyse[q queso red ¢ 0} Z °BO {peop IB A[AvON ERSE qsol oy} fepnyselq (que 10d G °B0) Moy V “BN 884 w10j jued sod Gg “Bd ‘pBop 4sod oy {quedo sod G “vO ‘ABp yxou BlNysviq MoT “pozh}oq ho IlV “pozf[oq49 [TV “pozf]o} -£90 a10ul 10 yud0 1ed GG “‘pozhy| -0j49 yuod aod (6 “BO “*pozh]| -0j40 quod red Q) 04 09 *pozhoyAo quod 19d Og UvY} O10] ‘(que0 aod [> ) enyse[q Moy B {ABp qxou pvop [[e A[IvoN quao Jed ¢ ‘vo [x] -N4sB[q WO] ssd0 MoT “el -N4sv]q WIOJ Usd Jd OF 09 0% faIp S830 YSO] “BlNASV]q UOJ J[VY uByy sse'yT ‘BlNYSBI ulIoJ quedo 190d QG “Bd ‘B[NYSB[ WIOF ISO] *JOVYUL JN Wa[[OMsS qsol oy} ‘pazh[o4 -Ad9 yuod 19d QG 04 OP *pozfoyAo yuso0 Jad QT 03 G “BO fgowzur Ay1ofeur ywadry “qoRqUl [[B JSOWLW “UOT[OMS “40B4UT [[V “UdT[OMS “9B4UT [TV "Ud|[OMS “JORIUT [TV uD IT Or 386 RALPH 8S. LILLIE required. Table 5 gives a detailed description of an experi- ment of this kind (see table 5). The contrast between the resistance of the eggs at the two stages is sufficiently evident from the table. The decline in resistance during cleavage is such that many eggs undergo cytoly- sis in sea-water of 40 per cent dilution; these are probably the eggs that were introduced at the period of maximum suscepti- bility. As the dilution increases the proportion of eggs under- going cytolysis ‘also increases progressively. The progressive character of the variation in susceptibility is well indicated by these results. A dilution of 60 per cent, in which uncleaved eggs remain intact, destroys all of the cleaving eggs in twenty minutes or less. One remarkable difference between the cycle of susceptibility to hypotonic sea-water and that to cytolytic substances and cyanide, is that there is no period of increased susceptibility to osmotic disruption immediately following fertilization. It is known that there occurs at this time, together with the in- crease of susceptibility to poisons, an increase of electrical con- ductivity; and other evidence indicates that the plasma- membrane then undergoes an increase in its general permeability to water-soluble substances and to water.2% Yet the membrane appears to be no more readily broken down by osmotic disten- tion than at any other time previously to cleavage. This differ- ence is surprising, and not readily to be accounted for. The lack of any change in the extensibility of the membrane is prob- ably to be correlated with the absence of any change in the form of the egg; apparently the loss of resistance to extension occurs only at the time when the cell is undergoing, or is about to undergo, active change of form. This would imply that the membrane-change accompanying cleavage is of a different kind from that immediately following fertilization, and involves a loss of coherence or extensibility. The basis of this difference 28 McClendon, Amer. Journ. Physiol., 1910, vol. 27, p. 240; Gray, Journ. Mar. Biol. Assoc., 1913, vol. 10, p. 50. 2° For a review of this evidence, and an account of experiments showing in- creased permeability to water, cf., my recent paper in Amer. Journ. Physiol., 1916, vol. 40, p. 249. PHYSIOLOGY OF CELL-DIVISION 387 cannot be assigned at present with any certainty; there is, how- ever, some evidence, partly cited above (p. 375), that resistance to hypotony does not necessarily depend upon the same condi- tions as resistance to cytolytic substances; thus in the case of blood-corpuscles, there appears to be an inverse relation between the two kinds of resistance—i.e., those corpuscles (e.g., of the rabbit) which are most readily cytolyzed by saponin, resist best the exposure to hypotonic media. As already stated, such corpuscles appear to be relatively deficient in cholesterol. It may be that those membrane-constituents which prevent the entrance of water-soluble substances and water (possibly chole- sterol?) are diminished immediately after fertilization, while those which impart coherence to the membrane (e.g., lecithin?) remain unchanged; on this view the latter membrane-components would diminish at the time of cleavage;the question, however, cannot be settled without further investigation. The fact remains that the membrane preserves unaltered its resistance to osmotic disruption throughout the entire period succeeding fertilization, until shortly before the egg begins to cleave. Properties of the fertilization-membrane Certain incidental observations on the formation of extra- ovates are cited in table 3, since they appear to throw some light on the vexed question of the nature of the fertilization- membrane. When fertilized eggs are placed in dilute sea-water, the fertilization-membrane is often ruptured by the pressure of the distended egg, and the protoplasm partly flows out, the egg assuming the shape of an hour-glass.*® It was always found that these extra-ovates were much more readily formed in eggs brought into dilute sea-water immediately after fertilization (within two minutes) than later. Table 6 gives the results of seven experiments in which the proportion of extra-ovates was estimated in eggs placed in dilute sea-water at 2, 4 and 6 minutes after fertilization (see table 6). These results show that the membrane is more easily ruptured soon after fertilization than later, i.e., its consistency and ten- 3° J. Loeb, Arch. f. Entwicklungsmech., 1895, vol. 1, p. 453. 388 RALPH Ss. LILLIE TABLE 6 PROPORTION OF EXTRA-OVATES IN EGGS PLACED IN DILUTE SEA-WATER AT DIFFERENT INTERVALS DATE AND COMPOSITION OF DILUTE AFTER FERTILIZATION SEA-WATER 2m 4m . 6m per cent per cent 1 | August 17. 65 volume per cent Many _ Few 2 | August 18. 60 volume per cent 10-15 1-2. Almost none 3 | August 19. 60 volume per cent 30-40. Caml: Almost none 4 | August 20. 60 volume per cent 20-30 ca. 1 Almost none 5 | August 21. 60 volume per cent 20-30 Ob Olean o 410 Direct absorption by cells of outer body walls.....................,.... 41l IOUS ous @ ong t cha Uo iced CRORE Eke oc or a Rt RSE Cat ar SEE REE gM eck a 411 Ree era cuneate DIVO tates aks, . SV chee dente hs Sc Sand Se 411 Comparisonvoreranulessm the celllspeemeres:4..6-aaos «2 wae. cs eae - 413 Tests for specific stains......... WRT seth nr aES « « Gehan das ERI: 414 AMUSsOTECNCMMEHMOME: eeecha. ... . a ecy mena cts ae ck, Wis Ltahne 5 aCe an a eee 416 Direct absorption by cells of outer body walls..................-.-+6+: 417 Sac liga LD API LS, ,.. BaMEEs 7 rn PE il sic 8 Jes ear 418 NtchiMte RUGS imeem renee a: «cca eke ee as ee a ees 418 Lb eatial a TEC Wil GTO .. 0. mcreahis Gey Meh IRM sade shad SMa Re eG 419 Pest viOlvOts LHETEGEPHSCICS 6)... .°s'. sae, SRA eA 2 F< Siv}o, Said aaechaiocn OE 419 Summary and discussion.............. Settee Bor Ut ety a to OTE Sek Soe 420 2, LESHIETCE 8 ole Ren aot ao aS Sagat Se oa ee ee 424 INTRODUCTION In 1914 the author began a series of investigations upon fresh-water mussels designed to ascertain whether or not there could be found any direct proof, especially of a histological nature, of the correctness of Piitter’s theory that animals living 1 Published by permission of the Commissioner of Fisheries. 403 404 E; P. CHURCHIEL, JR: in the water use, in addition to ‘formed’ food, nutriment which is in solution in the surrounding medium. He wished also to put to the test Piitter’s assumption that some of such nutri- ment is absorbed directly by the cells of the outer body walls, especially by those of the gills. Piitter (’06) was the first seriously to advance the theory that food could be so taken. He based his conclusions in part on a comparison of the amount of carbon necessary for the main- tenance of the organism with the amount furnished by the plankton. Pitter considered the latter too small for the needs - of the animal and urged that it must use some carbon which is in solution in the water, resulting from the decay and disin- tegration of organic life. He further stated that the amount of material found in the alimentary canal is never large enough to supply the requisite quantity of carbon. Besides the ali- mentary canal, the uncutinized epithelium of the outer surface of the body, especially that of the gills, was thought to function in absorbing dissolved food. This process was considered to go on in addition of course to the digestion of ‘formed’ food by the alimentary canal. Pitter concluded that the above con- ceptions applied to Protozoa, Porifera, Echinoderms, Crusta- ceans, Mollusks and Fishes. He tested the matter experiment- ally in two ways: first by noting that goldfish and perch lived longer in solutions of asparagin, somatose and glycerin thanin tap water: secondly by comparing the amount of oxygen needed to oxidize the lost weight of tissues of actinians, tunicates and fish while kept in their natural medium, with the amount of oxygen actually used. As the latter was found to be greater than the estimated quantity needed he concluded that the extra oxygen was used in oxidizing some food that had been taken from the water where it had been present in the form of a solute. Lohman (09) estimated the amount of plankton found in sea-water from various parts of the ocean and concluded that plankton comprised the main source of the food of aquatic forms. He quoted Henze’s researches as showing that there is present in sea-water no appreciable amount of carbon com- pounds in solution. ABSORPTION OF NUTRIMENT BY MUSSELS “<4 Knorrich (10), working with Daphnia, which lived 14 days in sterilized hay solution, concluded that nutriment was absorbed from the solution. Kerb (710) kept eels in sugar solutions and noted no diminu- tion in the amount of sugar from day to day. He obtained similar results while working with Corethra larvae in sugar solutions. Also he found that Daphnia lost in dry body weight as rapidly in solutions of peptones as in tap water. Wolff (10), working with Simocephalus, found that it lived twice as long in bacteria-free water, which contained some dis- solved carbon compounds, as it did in tap water. He made no observations as to body weight lost or gained. Several American investigators of the question of the nutri- tion of the oyster have felt that the amount of material found in the alimentary canal was not large enough to account for the growth of the oyster to the relatively large size which it attains in the first two years of its existence. Grave (12) in particular suggested that in the oyster food may be taken in by means other than that of the alimentary canal. Moore, Whiteley, Edie and Dakin (12) investigated the rate of oxidation and the output of carbon dioxide by aquatic inverte- brates in relation to the available food supply in sea-water. They also made chemical analyses of samples of sea-water from various regions. Their general conclusion was that sea-water does not contain any appreciable amount of dissolved organic matter capable of acting as a nutrient medium for animals living in it. Lipschitz (13) reviewed almost the entire subject including his own previously published experiments along that line and offered criticism of Piitter’s work. Lipschiitz noted that fish and eels when kept in nutrient solutions lost as much body weight as in tap water. He also thought Piitter had overestimated the amount of material in solution in the water and underesti- mated the carbon content of the plankton. His general con- clusions are the opposite of those of Piitter. 406 E. P. CHURCHILL, JR. Blegvad (’15) found that the invertebrates on the sea-bottom in Danish waters live on detritus and to a lesser extent on plank- ton. He does not corroborate Pitter’s theory but on the con- trary thinks that the latter did not allow enough for the aun of detritus. Esterly (16) found that in pelagic copepods the amount of food that is indicated by the intestinal contents is ‘surpris- ingly small.’ He suggests however that it is probable that much of the food of these forms consists of organisms without shells and which do not leave recognizable remains in the intestine. Further careful investigation he feels is necessary before Piit- ter’s theory is accepted. Lund? found that if Bursaria are kept in a weak soap solution they will absorb fat from such solution through their body walls. The present author, in 1915, published a paper in which were set forth the results of the earlier steps of his investigations. Fat was selected for the initial experiments as it? is most easily traced histologically. Following the method of Lund, olive oil was saponified and the mussels kept in weak solutions of the resulting soap. In some cases the solutions were stained with Sudan III. Briefly stated, it was found that the mussels ab- sorbed fat from the solutions to a marked extent. This was transported over the body by the plasma of the blood and by the corpuscles. Several of the experiments went far toward proving that some of the fat was absorbed directly by the cells of the gills, mantles and foot. The present paper discusses the results of the continuation of the series of investigations. Further work was done on the absorption of fat using a commercially prepared soap in- stead of the saponified oil. The absorption of protein was studied somewhat extensively and some efforts were made to ascertain the facts in regard to the absorption of starch. The author wishes to express his obligations to Dr. Caswell Grave, at whose suggestion the work was undertaken, for ad- vice and aid given throughout the course of the investigations: 2 Dr. Lund’s results are not yet published. ABSORPTION OF NUTRIMENT BY MUSSELS 407 to Dr. H. 8. Jennings and Dr. R. E. Coker for their suggestions and interest in the work: to Dr. G. L. Houser for library facili- ties at the University of Iowa; and to Mr. A. F. Shira, Director of the Biological Station at Fairport, Iowa. MATERIALS AND METHODS The mussels upon which the investigations were carried out were individuals selected from the more common species found in the Mississippi River near Fairport, Iowa. Adult specimens were employed in the following experiments. Care was exer- cised to choose non-gravid mussels which were in a seemingly healthy condition, ‘shoulder-raked’ or hand collected individ- uals being used in preference to those dragged out of the water by ‘crow-foot’ hooks. At Fairport soft water was used for the work with fat, for the reason that the soap was more soluble in it than in river water. For the experiments with protein and starch filtered river water was used. At the Johns Hopkins Laboratory, where the work concerned only protein and starch, either city tap water or Chattolanee spring water was used. The mussels experimented with were kept in glass aquaria of about 5000 cc. capacity. Control individuals in filtered water were kept side by side with those in the solutions. The solutions were changed daily in most cases. In some of the experiments with protein the changes were made twice daily. By this procedure the protein was prevented from undergoing marked decomposition. As the mussels are accustomed to a current of water this method of manipulating the aquaria seemed to be the closest approximation to natural conditions which it was possible to obtain in the laboratory. For the experiments with fat a ‘non-alkaline’ soap commer- cially prepared from olive oil was used. This was done to avoid the free alkali and oil present in the saponified olive oil employed in the earlier investigations. The strength of the solution used was 0.002 of one per cent. The tissues of various portions of the mussels experimented with were sectioned by the freezing THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 21, NO. 3 408 E. P. CHURCHILL, JR. method, stained with Sudan III and mounted in glycerin. In some cases the soap solution was stained by dissolving Sudan III in it to the point of saturation. Sections of mussels which had been kept in such solutions were mounted directly after cutting. A method of attacking the question of the absorption of pro- tein was much more difficult to devise. Egg albumin was chosen as the most convenient protein with which to work. The white of one egg was stirred into 300 cc. of water and the mixture filtered.. The soluble albumin passed through the filter leaving the insoluble globulin behind. (The presence of albumin in the filtrate could be demonstrated by the formation of a white precipitate on the addition of alcohol.) 50 to 100 cc. of the filtrate were used in 5000 cc. of water to make the solutions in which the mussels were kept. Some experiments were continued for considerable periods of time, thirty to sixty days, and his- tological evidence sought of differences in the condition of the cells of those individuals which had been in the solutions and of the control mussels. As the latter, being in filtered water, received little food it was thought that evidence of starvation might be detected by the appearance of the cells. In other cases the experiments were continued only a few days, the tissues of the mussels were fixed in absolute alcohol or other fixing fluids and sectioned in paraffin. These were studied both stained and unstained. Efforts were made to find a specific stain for the albumin. Cowdry (14) has found that Janus green, be- sides staining mitochondria, stains albumin in the living cell. This fact was made use of and mussels were kept in solutions of albumin stained with Janus green. Sections of such indivi- duals were made by the freezing method and mounted in gly- cerin. The paraffin method failed with such sections as the stain is washed out very readily by the alcohols. It was neces- sary to study the sections mounted in glycerin within the course of a day or two as the stain is also removed by the glycerin in a short time. Soluble starch was made by boiling and weak solutions were used in which to keep the mussels. Sections of such mussels ABSORPTION OF NUTRIMENT BY MUSSELS 409 were prepared both by the paraffin and by the freezing methods and efforts were made to detect starch in the cells by staining with iodine. Mussels were also kept in solutions of starch which had been colored blue with iodine. Spring water was used in this case as the tap water contained so much free chlorin that the blue color was removed from the solution. Further details of the methods employed will be stated in connection with the observations recorded below. OBSERVATIONS Fat The solutions made from the commercial soap were found to be more satisfactory than those used in the previous work with fat where olive oil was saponified with sodium hydroxide. In the solutions from the latter the mussels were found to throw off much mucus, no doubt owing to the free alkali present due to a slight excess of sodium hydroxide employed in the process of saponification. The mussels threw off no appreciable amount of mucus while in the solutions prepared from the non-alkaline commercial soap. None died while in the solutions. There was no evidence of any toxic effect due to the soap solution. The results were the same as those obtained in the previous work with fat. The mussels which had been kept in the solu- tions were found to contain much more fat than those not so kept. This fat was distributed in the form of droplets? in greater or less abundance in nearly all parts of the body. Figures 1, 2 and 3 represent portions of a specimen of an adult Quadrula ebenus which had been allowed to remain in a 0.002 of one per cent soap solution for five days. The fat, represented in the drawing by the heavy black dots, may be seen in the epithelial cells of the gill filaments, the ostia, the water tubes and in the corpuscles. 8 The term ‘dropiets’ will in this paper be applied to the spherules of ab- sorbed fat found in the tissues of the mussels experimented upon. These drop- lets usually were of diameters varying from less than one micron to 3 or 4 microns, though instances were found in which the diameter was as great as 7 or 8 microns. 410 E. P. CHURCHILL, JR. Mussels were kept in soap solutions stained with Sudan III. A gill filament of such a mussel after it had remained in the solution for five days is shown in figure 4. Pink or orange colored fat droplets were found in great abundance in the epithe- lial cells of the filaments, ostia, water tubes and in the corpuscles. This fact furnishes direct proof that the heavy loading of fat found in the mussels which had been in the soap solutions was not due to the chance use of an extraordinarily fat individual. The colored fat could have come only from the stained soap solution. Mussels with the valves wedged open were suspended over the solutions in such a manner that only the lower edges of the mantles were immersed in the solutions. The animal soon re- laxed the foot so that the end of that organ also was covered. Control individuals were suspended in like manner over filtered water. The experiment was continued for forty-eight hours. Numerous fat droplets were found in the cells of:the parts of the mantles and foot which had been immersed in the solution and none were discovered in similar tissues of the controls. Epithelial cells of the mantles and foot that had been covered by the solution are represented by figures 5 and 7, while figures 6 and 8 show cells of corresponding parts of the control muscles. It was thought possible that in this case the presence of fat was due to fatty degeneration incident to the weakened or mori- bund condition in which the mussel might be supposed to be after having been suspended thus with the larger portion of its body in the air for forty-eight hours. Such a supposition was rendered negligible by the facts that the mussels in which the fat was found were apparently in a healthy condition when killed; that no fat was found in the controls and none in mussels which had been allowed to die while so suspended over filtered water. Finally, a large mussel with the valves wedged open was placed over the edge of the aquarium in such a manner that the lower edge of one mantle only was immersed in the soap solution, the entire remainder of the body being in the air. After a period of forty-eight hours sections of the part of the mantle ABSORPTION OF NUTRIMENT BY MUSSELS 411 which had been in the solution revealed many fat droplets in the cells on the side next to the body. No fat droplets were found in the cells of the corresponding part of the other mantle which had not been in the solution. Some of the cells of the mantle which had been in the solution are represented in figure 9, while figure 10 shows cells of the other mantle which had re- mained in the air. In all these cases in which the mussels were suspended in such a manner as to allow only a portion of their body to come into contact with the solution, the oral and anal openings were above the solution so that none of the solution could have entered the alimentary canal directly. In the case last mentioned, in which one mantle of the mussel was immersed in the solution and compared with the other mantle of the same mussel, this mantle not having been immersed, direct proof of the absorption of fat from the solution by the epithelial cells of the mantle is given. In all cases in which one mussel is used as a control for another the objection can be raised that by chance a naturally fatter individual was chosen to be kept in the solu- tion. But that objection can not hold in this case as it is hardly supposable that corresponding parts of the two mantles of the same mussel would differ in fat content as markedly as indicated in figures 9 and 10. Protein The problem of the absorption of protein was first attacked by means of the ‘starvation’ method. Five mussels were kept in solutions of egg albumin and five in filtered water. Both experiments were started on the same day and at intervals one individual was removed from the solution and one from the water at the same time and their tissues fixed and sectioned. Bouin’s fixing fluid was used for this work and the sections were _ lightly stained with haematoxylin and eosin. One pair of mus- sels was killed at the expiration of ten days, other pairs at the end of thirty, forty-five and fifty-eight days. One of the in- dividuals in filtered water died on the twentieth day. The tissues of the mussels were compared by the study of the sections for the purpose of discovering any possible difference in the degree 412 E. P. CHURCHILL, JR. of emaciation between those that had been in the solution and those that had not. The criteria taken into account in judging the degree of emaciation were shrinkage of the cells and loss of capacity of the cytoplasm of taking up the eosin stain. Many workers on starved tissues have described great shrinkage and loss of staining capacity on the part of fasting cells. Morgulis, - Howe and Hawk (’15) found that the cytoplasm of the cells of a dog which had suffered extreme starvation took a much less pronounced eosin stain than the cells of normal tissues. The cells of the mussel which had been in fltered water for ten days showed no appreciable signs of emaciation. Various portions of its tissues were compared with corresponding parts of those of the individual from the solution, special attention being given to the epithelial cells of the gills in all the starva- tion experiments. In making the tests with the eosin stain a slide holding the sections from the mussel which had been in the solution was placed back to back with a slide of the sections of the control and both were passed through the alcohols and eosin together. In this way it was certain that both were ex- posed to the action of the reagents for exactly the same length of time and that any difference found in depth of stain was due to a difference in the cells and not to a variation in the time the tissues had remained in the stain. In some cases sections from mussels to be compared were mounted on the same slide. In the comparison of the mussels killed at the expiration of thirty days, marked differences were found between the cells of those that had been in the solution and of those which had been in filtered water only. These differences were especially noticeable in the cells of the gills. The cells of the mussels which had been in the solution were plump, unshrunken and took the eosin stain in a normal manner. The gill filaments of those from the filtered water were small and wrinkled. The cells were shrunken and took a much lighter eosin stain than did those of the mussels from the solutions. Figures 11 and 12 represent respectively filaments of the mussel which had been in filtered water and of the one which had remained in the solu- tion of albumin. These figures show most clearly the shrinkage ABSORPTION OF NUTRIMENT BY MUSSELS 413 and distortion of the cells, the differences in the depth of, stain taken being less exactly represented by the stippling. In figures 13 and 14 are shown respectively filaments of a mussel which had been subjected to filtered water and of one which had been in the solution for fifty-eight days. The same differences were found in this case as in the former. The pair killed at the end of forty-five days also manifested these differences. The above facts leave little doubt that the mussels in the solutions make use of the albumin as food. The mussel that died during the cowrse of the experiment was one of the lot in filtered water and its death was probably due to starvation. Efforts were made, however, to devise methods whereby the albumin might actually be seen in the cells. Experiments were carried out in which some mussels were kept in solutions of albumin and others in tap or spring water. Sections of those in the albumin solutions were mounted on s'ides with corre- sponding sections of individuals which had been in water only. These were stained with eosin or Bordeaux red and a study of the minute structure of the cytoplasm was made. In general the cytoplasm of the mussels which had been in the solutions was of a more granular nature than that of the other mussels. The granules were not only more numerous but many of them were larger. For example, in one case where 100 cc. of albumin in 5000 ec. of water had been used in an experiment of twenty-one days’ duration the difference was quite marked. The effect of the fixing agents is to precipitate the proteins of the cell. Apparently there was a greater quantity of protein present in the case of the mussels which had been in the solutions of ablumin. In an effort to discover whether or not the same result could be obtained if the alimentary canal were eliminated, the mouths of some of the mussels were plugged with dumbbell-shaped pieces of paraffin. In many cases the mussels died due to in- jury incident to the operation. If no injury was inflicted the mussels often lived for a week or two. If the mussel did not die in the course of one or two days it might be assumed that no injury had been inflicted and that the mussel would live for at least eight days, which was the length of time it was desired 414 E. P. CHURCHILL, JR. that the experiments should continue. The cytoplasm was studied after staining with eosin or Bordeaux red with the same results as in the previous experiments. The cytoplasm of those mussels which had been in the albumin solutions was more granular, the granules being more numerous and larger, than in the mussels which had remained in water only. This was especially apparent in the cases of two pairs of mussels from an experiment extending over six days. The solution in which the mussels had been kept contained 100 cc. of albumin in 5000 ec. of water. The cytoplasm was stained, in the case of one pair of mussels, with eosin and in the other with Bordeaux red. In both cases there were more granules in the cytoplasm of the mussels which had remained in the solutions than in that of the others. If the presence of a greater quantity of granules in the cells of the mussels which had been in the albumin solu- tions may be taken as evidence of the presence of albumin, in these cases in which the mouths of the mussels were plugged the albumin could have been taken up only by direct absorption from the solution. Considerable effort was expended in attempting to discover a specific stain for the albumin, so that its presence might be detected in a manner similar to that by which fat is revealed by means of Sudan III. Albumin was precipitated with absolute alcohol, caught on filter paper and mounts made of it, both while wet from the alcohol and also after dehydrating, clearing and embedding in paraffin. While unstained, the appearance of the albumin under the microscope was that of a quantity of granules of various sizes. Most of them were irregular in shape but now and then one of a spherical form was discernible. Some of the mounts were carried through one stain, some through another, the object being to find a dye that would stain albumin and yet not stain the general cytoplasm of a cell. No great degree of success was attained. The albumin took the various stains tried, including eosin, Bordeaux red, haematoxylin, Janus green and others. All of these dyes stain the cytoplasm of fixed cells except haematoxylin, which is not considered a cyto- plasmic stain. Since, however, the only color this stain im- ABSORPTION OF NUTRIMENT BY MUSSELS 415 parted to the albumin was that of a weak bluish-brown it proved to be of no great value. When fixed tissues of the mussels which had been in the albumin solutions were treated with haematoxylin, it was found impossible to distinguish any gran- ules in the general mass of the cytoplasm that might have been stained with the haematoxylin. The cytoplasm in unstained fixed tissues is of a brownish color and it was difficult accurately to determine the color of the small opaque granules found in these sections. Albumin solutions were stained by dissolving haematoxylin in them. Mussels were kept in such solutions and control in- dividuals were allowed to remain in tap water in which haema- toxylin had been dissolved. The mouths of the mussels were plugged as described above. The experiment was continued for five days. At the close of that time portions of the gills _ were fixed in absolute alcohol and embedded in paraffin. As abso- lute alcohol precipitates albumin and does not dissolve haema- toxylin from tissues, it was considered a favorable fixative for this purpose. ‘The sections were mounted unstained and search was made for blue granules which might have been taken up from the solution by the cells. The results of such a study were of a negative nature. The same difficulty was experienced as described above in regard to the fixed tissues. While some of the granules in the cells which had been in the solutions seemed to have a bluish tint, in general it was impossible to be certain of the matter. In other words, haematoxylin does not impart to albumin a stain sufficiently characteristic to permit of its being distinguished from the unstained granules about it. ’ However, another result was obtained which confirmed the results of an earlier experiment. The cytoplasm of the cells of mussels which had been in the solutions was of a much more granular nature than that of the cells of the individuals which had not. The granules were both more numerous and larger. Also, many could be seen adhering to the outer ends of the cells or entangled in the cilia. Appearances indicated that these granules were particles of albumin and that some had been taken into the cells. The gill filaments of mussels which had 416 E. P. CHURCHILL, JR. been kept in albumin solutions thus stained are represented in figures 17 and 18. Figure 19 is a 1epresentation of the gill filaments of a mussel which had remained in water in which haematoxylin had been dissolved. The granules in the latter will be seen to be much less numerous than in the former and none can be observed clinging to the outer ends of the cells or to the cilia. Finally a fact was discovered that led to more definite results. Cowdry, in a paper on mitochondria, makes mention of the fact that he has found that Janus green will, in the living cell, stain lecithin and albumin, the latter more heavily. As mentioned above the author had found that in fixed preparations Janus green stained the entire cytoplasm. An attempt was now made to determine whether or not Janus green might be employed as a means of detecting albumin that might have been absorbed by the cells of the gills or other parts of the outer body walls . of the mussels. Mussels were put into solutions made up 5000 cc. of water and 100 cc. of albumin stained with Janus green. Control in- dividuals were placed in water in which Janus green had been dissolved. The mouths of these mussels were left open. The experiments were continued for eight days. At the expiration of that period the mussels were killed and their t’ssues were hardened in 10 per cent formalin for a few hours. As albumin is precipitated by formalin, this procedure would cause the pre- cipitation of the albumin in any of the solution which had adhered to the outer surfaces of the cells in addition to the coagulation of what albumin might have been absorbed. In this way ag- gregates of albumin sufficiently large to be visible under the microscope would be formed and thus render it possible to trace this protein histologically. Sections were then cut by the freezing method and mounted in glycerin. The gill filaments of a mussel that had been in the stained solution are represented in figure 20. Numerous green granules of various sizes were found within the cells and within corpuscles in the blood vessels in ~ the center of the filaments. The granules which had taken the stain are represented in the drawing by heavy black dots. In ABSORPTION OF NUTRIMENT BY MUSSELS 417 some cases these green bodies were almost spherical. This was especially true of the larger granules. In general the diame- ter of the granules varied from less than one micron to that of three or four microns. Green colored granules were also found clinging to the outer ends of the cells or adhering to the cilia. Figure 21 shows the filaments of a mussel kept in the water and Janus green. In this case no green granules were found either within the cells or on their outer surfaces. In some eases, both in the mussels that had been in the solutions and in the - controls, the entire cytoplasm of the cells of the outer ends of the filaments showed a pale green tint. The granules in ques- tion, however, took a much deeper stain and were readily dis- tinguishable from the surrounding cytoplasm. Similar experiments were carried out in which the mouths of the mussels were closed with paraffin plugs. Exactly the same results were obtained. In figures 22, 23 and 24 are shown gill filaments of a mussel which had been in the stained solution. The black dots represent green granules and spherules which were found within the cells and corpuscles and clinging to the outer ends of the cells. The mussels used as controls in these experiments had no definite green granules within the cells or on the outer ends, though the cytoplasm often exhibited a faint green tint, as did that of those which had been in the stained solutions. A drawing of the gill filaments of one of the control mussels would present the appearance of the filament shown in figure 21. The presence of the green tint found in some cases over the entire cytoplasm does not militate against the efficacy of the Janus green as a specific stain for the albumin. Sudan III, while it is a specific stan for fat, often imparts a reddish tint to the entire cytoplasm, but stains the fat droplets a much deeper hue. Figures 25, 26 and 27 represent respectively portions of the palp, foot and mantle of a mussel which had been in the stained solution. The mouth of the mussel had been closed with a paraffin plug. Green granules were found in the epithelial cells of all three tissues. 418 E. P. CHURCHILL, JR. The above results demonstrate that the cells of the outer body walls of the mussels have the power to absorb at least one protein, albumin, from the surrounding solution. Starch Soluble starch was made by boiling commercially prepared cornstarch in water. One gram of starch was used in about 400 cc. of water. The solution, after boiling, was allowed to settle and 50 or 100 cc. of the supernatant liquid used in making the solutions in which the mussels were to be kept. This method affords only a rough estimate of the amount of starch present, as the amount of starch that went into solution was not deter- mined. However, qualitative results only were sought in these initial experiments. The supernatant liquid was proved to contain starch by testing with iodine. After the mussels had remained in the solutions for various lengths of time they were killed and sections of their tissues were made both by the freez- ing and by the paraffin methods. The sections were then placed in alcoholic solutions of iodine for different periods of time and an effort was made to find blue granules within the cells. The — results were negative. The tissues as a whole often assumed a slight yellow tint from the influence of the iodine, but no gran- ules were found that could be considered to have a blue color. As iodine stains dextrin or causes it to assume a red color, search was made for red particles in the cells. It was thought possible that the starch might have been more or less completely con- verted to sugar by the cells and that evidence of some of the incidental steps could be found. Very small red granules were found both in the tissues of the mussels which had been in the solutions and in those of the controls. These red particles were no doubt the pigment granules which impart to the tissues the pinkish hue observable in many specimens of Quadrula ebenus. They presented no difficulties in the work with fat which had been stained red with Sudan III, as the pigment was in the form of small, irregular, highly refractive granules readily distinguishable from fat droplets. Their presence offered some ABSORPTION OF NUTRIMENT BY MUSSELS 419 complications, however, in the search for any red granules that may have been present owing to the conversion of starch to dextrin and to its subsequent staining with iodine. No satis- factory evidence of the absorption of starch was disclosed by the above methods. The experiment of keeping the mussels in a starch solution which had been stained blue with iodine was then tried. The mouths of the mussels were plugged with pieces of paraffin as described above. After the expiration of four days the mus- sels were killed and sections were made of their tissues by the freezing method. The sections were mounted at once in gly- cerin without staining. In the cells of the gill filaments some granules were found that were blue in color, though the obser- vation was made with difficulty owing to the opacity of the granules. Similar granules were found within the corpuscles and also clinging to the outer ends of the cells and to the cilia. -They were quite irregular in shape and varied from one to three microns in diameter. The gill filaments of a mussel which had remained in the stained starch solution for four days are represented in figures 28 and 29. Figure 30 is a representation of the gill filament of a mussel kept in tap water. These results present some evidence that starch may be taken up directly from a starch solution by the epithelial cells of the gills of the mussel. Since fewer experiments were carried out with starch than with fat and protein, the results can not be considered to be as conclusive as were those attained in the work with the latter. Behavior of the corpuscles An interesting point in connection with the behavior of the corpuscles of the blood was noted. In the author’s previous paper dealing with the absorption of fat, mention was made of the fact that the corpuscles were often found pressed closely against the bases of the epithelial cells of the gills and that the corpuscles carried some of the fat droplets from the cells to the other parts of the body. Some of the fat droplets were appar- 420 . E. P. CHURCHILL, JR. ently transported while floating in the plasma of the blood. A description has been given above of the discovery of fat drop- lets, granules of albumin or of starch, as the case might be, in the corpuscles of the mussels which had been in the solu- tions. It was also found that the corpuscles, in addition to pressing closely against the bases of the cells of the gills, actually wandered out between the cells in such a manner that their en- tire surfaces were in touch with the cells, the corpuscle lying in a sort of cup between the contiguous cells. This phenomenon is illustrated in figures 15, 16 and 17. In great numbers of cases the corpuscles were found pushed out between the cells in such a way as to be entirely surrounded. In other instances a corpuscle might be observed partially surrounded as in figure 15, the process of fixation having killed it at the moment of entrance to or exit from between the two cells. Undoubtedly a position in which the corpuscle is entirely surrounded by the epithelial cells affords more surface for the exchange between the cells and the corpuscle not only of materials concerned in respiration, but also for the reception by the corpuscle of food material absorbed by the cells of the gill. Similar behavior of the corpuscles was not noted in parts of the body other than the gills. SUMMARY AND DISCUSSION The above results leave no doubt of the fact that mussels may make use of some kinds of food whlch are in solution in the water. A part, probably a small one, of such nutriment can be taken up directly by the outer epithelial cells of the body. As these animals are provided with a well developed digestive apparatus we may suppose that the absorbing power of these outer epithelial cells is a property that has been retained from amore primitive state in which the cell was less highly specialized and that this property is not a special adaptation correlated with the lack of a functional digestive system. The nutriment taken up from the solutions by the mussels by means of the alimentary canal was no doubt absorbed by the cells lining the intestine m the usual manner in which ‘formed’ ABSORPTION OF NUTRIMENT BY MUSSELS 421 food is absorbed after being rendered soluble by the digestive juices. In regard to the mechanism of the absorption of the foods by the cells of the outer body walls little can be offered. In the case of the fats, numerous droplets which took the Sudan III stain were found closely attached to the outer ends of the epithelial cells of the gills or mantle. These droplets were probably the fatty acids, a small part of which were pres- ent in the soap solution owing to hydrolysis, by which process sodium hydroxide and the fatty acids would be formed; the remaining droplets were no doubt due to a slight acidity of the surface of the living cells resulting from the union of carbon dioxide from the cells with the water, forming carbonic acid. In the experiments in which olive oil was saponified no attempt was made to remove the glycerin from the mixture. There- fore free fatty acids, from the droplets just mentioned, and glycerin may have been absorbed separately and resynthesized to fat within the cells. This absorption may have been effected by phagocytic or amoeboid action of the cells or by solution in the plasma membrane and reprecipitation within the cell. In the case of the commercially prepared soap the objection might be raised that there was no glycerin present to be absorbed and reunite with the fatty acids after their entrance into the cells. However it is safe to assume that a moderate amount of glycer- in was present in the soap from the fact that in the process of its manufacture the soap is ‘salted out’ of the glycerin, allowed to rise to the surface and removed while wet with the glycerin. Not all the glycerin is removed by the subsequent drying. Again it is not certain that the drops taking the osmic acid ‘stain as described in the author’s previous paper or the Sudan III stain in the sections prepared by the freezing method were fats or only the fatty acids. It is known that osmic acid blackens or browns free fatty acids. It is possible that Sudan III also stains them. At any rate solutions made from the commercially prepared soap dissolve Sudan III readily and assume the char- acteristic red color which that stain imparts to fat. It is prob- able that both these stains for fat really affect the fatty acid radical only and that it is that radical which carries the Sudan 422 E. P. CHURCHILL, JR. III into the cell. Therefore there may exist the possibility that the sodium oleate, stearate and palmitate from the soap may have entered the cell as such, the radical carrying the stain. The sodium may then have been separated leaving the radical free to unite with any glycerin which may have been absorbed. In regard to the absorption of albumin it is necessary to as- sume either a power on the part of the cell to split the protein into its amino acids and the absorption of these as in the ali- mentary canal, or the direct taking in by the cells of the col- loidal particles of albumin by means of something analogous to phagocytic action. The fact that the Janus green stain was carried into the cell offers some evidence that the albumin entered the cell as such without being previously split into the amino acids. In the case of starch it seems probable also that the granules entered the cell by amoeboid or phagocytic action. The presence of the definite blue granules within the cells would somewhat oppose the theory of any conversion of the starch previous to absorption. The present investigations demonstrate only the ability of the mussels to make use of nutriment which is in solution in the water by the twofold means cited above. They do not deal with the possible amount of nutriment present in various bodies of water in which aquatic animals are found. ‘The investiga- tions show that if dissolved material is present the mussels can make use of it. After the ability of the animal to make use of food in such form is proved, the question of whether or not, in any particular case, it does do so, depends upon the presence or absence of such food in the water. It is too sweeping a statement to assert that aquatic animals in general do not or can not make use of nutriment which is in solution in the water merely because no dissolved compounds are found in certain analyses of water taken from a more or less limited region. Nor is the fact of the presence of an amount. of detritus, apparently adequate for the nourishment of a bottom-living animal, suff- cient proof that all its nutriment is made up of such detritus. ABSORPTION OF NUTRIMENT BY MUSSELS 423 In the case of mussels in particular it is very probable that considerable nutriment is in solution in the water in which they live. While no analysis of the water has been made thus far in the investigations, the Mississippi River, as everyone knows, drains a vast area of land from which refuse from decaying ani- mal and vegetable material is collected in very great abundance. The dead bodies of aquatic forms of many sorts add to this sup- ply. The mussels are bottom-living organisms and come into contact with decaying and dissolving organic matter which is lying on or being slowly moved along the substratum. Solu- tions or colloidal suspensions of the proteins must certainly be present in some abundance. As the water is slightly alkaline some of the fat from decaying organisms is probably saponified and thus distributed throughout the water instead of rising to the surface where it would be inaccessible to many forms. In general the question of whether or not a particular aquatic animal absorbs nutriment from solution by means of the ali- mentary canal and the outer body walls probably depends on the presence or absence of dissolved substances in the water in which the animal in question lives. Fresh-water mussels can absorb nutriment which is in solution in the water and it seems very probable that other forms likewise possess this ability. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 21, NO. 3 424 E, P. CHURCHILL, JR. LITERATURE CITED Bieavap 1915 Nutrition of marine invertebrates. (Translation of title.) Rep. Danish Biol. Station. 22. (From review in Jour. Royal Mic. Soc. Aug. 1915.) CuurcHILL 1915 The absorption of fat by fresh-water mussels. Biol. Bull. vol. 29. Cowpry 1914 The comparative distribution. of mitochondria in spinal gang- lion cells of vertebrates. Am. Jour. Anat. vol. 17. EsterLy 1916 The feeding habits and food of pelagic copepods and the ques- tion of nutrition by organic substances in solution in the water. Univ. Calif. Pub. in Zodélogy, vol.1 6. Grave 1912 Fourth report of the shell fish commission of Maryland, p. 333. Kerp 1910 Uber den Nihrwert der im Wasser gelosten Stoffe. Revue d. ges. Hydrob. u. Hydrog. Bd. 3. Knorricu ©1910 Studien iiber die Ernihrungsbedingungen der Daphnien. Ploner Berichte Lipscuutz 1910 Zur Frage iiber die Ernihrung der Fische. Z. f. allg. Phys. Bd.12. 1910 Uber den Hungerstoffwechsel der Fische. Ebenda. 1913 Die Ernaihrung der Wasserthiere durch die gelésten organischen Verbindungen der Gewasser. (Eine Kritik.) Ergebnisse der Physio- logie. Bd. 13. Louman 1909 Uber die Quellen der Nahrung der Meerestiere und Piitters Untersuchungen hiertiber. Intern. Rev. d. ges. Hydrob. u. Hydrog. Bd.2. Moorr, Wuittey, Epir anp Dakin 1912 The nutrition, metabolism and respiration of aquatic animals. Rep. of British Assoc. Advan. Science. Dundee, p. 654. 1912 The nutrition and metabolism of marine animals in relation- ship to (a) dissolved organic matter and (b) particulate organic mat- ter in seawater. Bio-chem. Jour. vol. 6, pp. 255-296. Moors, Epig anp WuHiITLEY 1914 Nutrition of marine animals. Rep. Lan- cashire Sea-fisheries Lab. 22. (From review in Jour. Royal Mic. Soc. June, 1915.) Morcutis, Howr anp Hawk 1915 Studies on tissues of fasting animals. Biol. Bull., vol. 28. Pitter 1908 Die Ernihrung der Wassertiere. Z. f. allg. Phys. Bd. 7. 1909 Die Ernaihrung derFi sche. Ebenda, Bd.9. 1911 Die Ernihrung der Wassertiere durch geléste organische Ver- bindungen. Pfliigers Archiv. Bd. 137. 1911 Der Stoffwechsel der Aktinien. Z. f. allg. Phys. Bd. 12. Wo.rr 1910 Ein einfacher Versuch zur Piitterschen Theorie von der Ernéihrung der Wasserbewohner. Intern. Rev. d. ges. Hybrob. u. Hydrog. Bad. 2. rPave4d i) ? ! Aid A sian Asthinit si MU ON ote tc m Inte Ty ns ae ee ur . ae a Se em ea 120 ‘ aA Ls w “ea ae. SPOTS dno ern san a “i ys J an f bie Fuk Vv ‘ Mantel vn pai dn a ate Oe es ‘ riba Mee Vie ee i Lisinial, d ayaa SAY fh %)) ip ea tr} i ‘ a i ‘ hy ae nal Pit. 3 j Ler ye hy Al) try ,~ 33 CE. RP iin "ee eA toaeea ¥ nyt dyadc Stirelaiaeh 4) Yo Sis MAY lop iy Tai Mie p nda d Eerie (aba: OL hs hate oh eee ith ty Weni* Ty faut ana, taf ¥ ay, \ ane Ky dul i i }, i id vy"? ‘ ‘ a ‘ . ar n ot Aged “ts abv » | sens i ani 1. a J hd “Te Pa i po eae 4 Na "ee yy +, pry | el ST Le ates 5 i” pd Wee) Beh : * * am of ‘ ‘ aoc SA 78 Ale tt ere ‘wel Hip eet ys Ta eels in \ = OY! non We We ; ee : eae - * / al 5 2 ; > - ’ a Sag ee ; } . p : ij a hil tah 4 elt oh Pints (14 etal'y Yi Bis 4 OLD n= | 2 Aad ' . j ; fy , sve P IAS Sb a ea 4 j te ‘ ' “ : i on , Ai’ ‘ ‘ iy a REG a aa ie moe eee SSS = —ec drm. = 2 @ —— Oe SS 6 ms’enr ms’enr. == — ec drm. —_ - > eR ec’drm.- — se Be ae ~ Or \ } f ST TEs So. / "| | Wea art = = ! i} arr DURE HARARE RY on) ) > 459 THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 21, NO. 4 cs ee e ne Mere Pemaee oF ,<* " . “y ' ; CONTRIBUTIONS FROM THE ZOOLOGICAL LABORATORY OF THE MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD COLLEGE. NO. 282 THE EFFECTOR SYSTEMS OF ACTINIANS G. H. PARKER Every animal possesses means, more or less well defined, for responding to environmental changes and in many groups of animals these means are differentiated into definite systems of organs, the effector systems. In any given species the ef- fectors may be independent of, or more or less under the control of, the nervous system. They are obviously often very diverse. In the actinians the most conspicuous of these systems are the mucous system, the nematocyst system, the ciliary system, and the muscular system. With each one of these the animal is capable of responding to particular elements in the environ- ment and all have been suspected of being in one way or another under nervous influence. The nature of the systems and their relations to the nervous mechanism in the actinians will be taken up in the following pages. THE MUCOUS SYSTEM Every part of Metridium seems to be able to produce mucus, and, under normal conditions, all the surfaces of this animal are covered with a thin layer of slimy secretion. This is pro- duced in all probability by the so-called mucous cells and albu- men cells which have been described by Schneider (’02) in Anemonia as occurring throughout the ectoderm especially in the oesophagus and the mesenteric filaments, and generally, though less abundantly, throughout the entoderm. If an expanded Metridium is submerged in fresh water for about fifteen seconds and then returned to sea-water, its whole outer surface and especially that of its tentacles becomes cov- ered with a profuse secretion of mucus thus showing the extreme ease with which this material is produced. When a Metridium has been kept in an aquarium whose water supply is not of the 461 462 G. H. PARKER best, the animal commonly envelops itself in a mucous coat, which is often sloughed and forms a loose tube about its base. This secretion of mucus, which occurs in response to a great variety of stimuli, mostly noxious, is undoubtedly a means of protection, as has been suggested by Duerden (’06) and by Gee (13), and recalls the more permanent secretions of such actinians as the cereanthids, in which protective tubes are regularly pro- duced. But mucus is not only discharged in response to noxious stimuli; it is also produced through the action of favorable influences. Meat juices call forth, especially from the tentacles, a profuse discharge of mucus which renders these organs extremely sticky. In this way the capture of food is greatly facilitated. After Metridium has been submerged in fresh water for a brief period the mucus discharged on its oral disc is not usually simply sloughed off but much of it may be swallowed as it is in such corals as Favia and Fungia where, according to Duerden (06), the chief means of feeding is by swallowing the mucus of the oral disc in which small organic fragments, etc., have been entangled. This method of feeding is not known among the actinians, but the response of Metridium just described shows how easily the actinian method of capturing food could be converted into that employed by Favia and Fungia. If a given spot on the outer surface of the column of Metri- dium is touched three or four times with the end of a small, blunt glass-rod, a slight clot of milky mucus will soon appear. This clot is to be seen in the exact region of stimulation and though this region often shows a considerable depression due to muscular contraction which may spread horizontally half way round the column, the secretion of mucus always remains strictly localized and gives no evidence of being extended through muscle contraction or nerve transmission. Dilute acids and dilute alkalies also bring about a free dis- charge of mucus. Juice from the meat of the killifish, Fundu- lus, when applied to the isolated tentacles or acontia likewise produces a free discharge of mucus. The discharge of mucus on the acontia forms a tube through which this organ gradually makes its way by ciliary action. EFFECTOR SYSTEMS OF ACTINIANS 463 Wolff (04, p. 250) has described nerve fibers in connection with the gland cells in actinians and von Uexkiill (09a) has expressed the opinion that the secretion of mucus in these ani- mals is under the control of their nervous mechanism. ‘The fact that on stimulation mucus is discharged only over the surface stimulated and never beyond this area gives no support to this view but suggests rather that this activity is due to local stimu- lation unassisted by nervous action. This opinion is rendered probable by experiments made on anaesthetized actinians. If a Metridium that has been shown to secrete mucus on mechanical or on stimulation by weak acid is put in seawater to which mag- nesium sulphate has been added, the animal will cease to respond through its neuromuscular mechanism in about ten minutes. Nevertheless it will discharge mucus freely but locally to ap- propriate mechanical and chemical stimulation and this power will remain in full vigor even after the animal has been in the magnesium solution seven hours. Hence it would appear that the secretion of mucus is not dependent upon nerves. Similar results were obtained from chloretone. After a submersion of ten minutes in seawater containing chloretone, Metridium ceased to respond through its neuromuscular mechan- ism to mechanical stimulation. Notwithstanding, it secreted mu- cus to appropriate local stimuli, and this recurred even after 6 hours of immersion in the chloretone seawater. From the obser- vations already recorded and from the results of the experiments on anaesthetization, I believe we are warranted in concluding that the nervous system of actinians is not involved in calling forth the secretion of mucus but that this activity is accomplished by the direct stimulation of a body of independent effectors, the specific gland cells of the ectoderm and entoderm. THE NEMATOCYST SYSTEM No cellular constituents are more characteristic of any large group of animals than the nematocysts are of the ccelenterates, for the occurrence of these organoids in certain mollusks seems to be due to a process of digestive appropriation from ccelenterate sources. The offensive and defensive character of nematocysts 464 G. H. PARKER has long been recognized and much has been written about their method of development and the means by which they are ex- ploded. It is not my intention to take up a discussion of these problems but to limit myself to the single question of the rela- tion of the nematocyst to the nervous mechanism of sea-anemones as seen in normal activity. In 1871 Schulze showed that when the nematocysts in Cordy- lophora were discharged by impact with a foreign body this discharge took place only at the spot where the foreign body came in contact with the animal. Schulze showed further that the nematocyst cells were provided with a small bristle-like structure, the cnidocil, which projected beyond the general sur- face of the animal and served as a tuigger for the explosion of the nematocyst itself. Some eight years later the Hertwigs (79-80) discovered branched basal processes on the cells which produce the nematocysts and believed these to be nervous in function. Thus it came to be assumed that under appropriate nervous stimulation large numbers of nematocysts could be discharged in moments of need. This theory of the nervous discharge of the nematocysts ‘was supported from one stand- point or another by von Lendenfeld (’83), Chun (’91), Schneider (02), Wolff (04), GroSelj (09) and Baglioni (’13), while other investigators were rather inclined to accept some of the various modifications of the theory of non-nervous discharge as originally advanced by Schulze. Metridium offers some excellent oppor- tunities for testing these two general views. Several types of nematocysts occur quite commonly in the ectoderm of Metridium. Small cysts with short slender fila- ments occur sparingly on the mesenteric convolutions and in the ectoderm of the column wall. Large ones are found scattered over the oesophagus. But the regions in which the nematocysts are especially developed are the tentacles and the acontia. On stimulating the tentacles with dilute hydrochloric acid, great numbers of rod-like bodies about 20u in length together with many fine filaments at least 140u long are discharged. On similar treatment of the acontia, a perfect forest of nettle fila- ments are discharged. These come from capsules about 45u EFFECTOR SYSTEMS OF ACTINIANS 465 in length and consist of a spirally marked, basal stalk some 70u long and a terminal filament of over 700 length, the total extent of these elements being at least twenty-five times that of the shortest nettling organs in this animal. To ascertain the relation of the nematocysts to the nervous system in actinians the following experiments were carried out on the tentacles and the acontia of Metridium. A normal acontium can be cut from an animal and placed under a micro- scope without bringing about the discharge of its nematocysts. If, now, it is flooded with carmine in seawater or with a solution of methylene blue in seawater, its cilia can be seen to strike toward what was its distal end, but no nematocysts will be ex- ploded. If, next, it is flooded with HCl z a profuse discharge of nematocysts occurs. Judging from the observations of Glaser and Sparrow (’09) probably most acids would cause this reaction. Distilled water will also bring about the explosion of the nema- tocysts, but as a rule only a few are thus discharged, for, if the treatment with water is followed with dilute acid, a renewed discharge immediately takes place. Some samples of methyl green have been found to produce a very complete explosion of the nematocysts but others have not, a fact which indicates that the discharge was probably produced by some associated impurity rather than by the methyl green itself. A few nema- tocysts are always discharged near the cut end of the acontium, and if an acontium is shot in and out a pipette in seawater, many of the nematocysts will be found discharged. These organoids, however, are not exploded when the acontium is flooded with juice from the flesh of a fish (Fundulus) though they do discharge in fair numbers when a small piece of this flesh is brought into contact with them. This reaction is probably dependent upon a mechanical rather than a chemical stimulus from the flesh. The tentacles of Metridium discharge their nematocysts to dilute acids, certain samples of methyl green, mechanical insult, and more or less to distilled water but not to carmine nor methylen blue nor to meat juice in seawater. In fact the nematocysts of the tentacles respond to the various stimuli I have tried in precisely the same way as those of the acontia do, and I have 466 G. H. PARKER been unable to find even a slight difference between these two sets of nematocysts such as might be inferred to exist from the table given by Glaser and Sparrow (’09, p. 367). When the portion of the tentacle or acontium that receives an effective stimulus is compared with that from which the nematocysts are discharged a striking condition is found. With mechanical stimulation the nematocysts are discharged only in the immediate region of the application of the stimulus, as observed long ago in Cordylophora by Schulze (71); with chemical stimulation apparently the same is true. If the dis- tal end of a fragment of an acontium or a tentacle is treated with dilute acid, the nematocysts are discharged at that end and nowhere else. If the proximal end is similarly treated, they explode only in that region. There is thus no evidence in these two organs of nervous transmission in either direction so far as the nematocysts are concerned. But the relation of the stimulated area to the area of discharge is best seen when the stimulus used is a colored fluid such as methyl green. Ifa small erystal of that kind of methyl green which will cause the discharge of the nematocysts is brought near a living acontium, the nematocysts can be seen to explode as they become covered by the diffusing green solution and as they are never discharged in advance of the cloud of colored fluid, the evidence for the local action of the stimulus uninfluenced by transmission, is very conclusive. These observations are in exact agreement with those of Wag- ner (’05, p. 618) on Hydra and I therefore conclude that nema- tocysts are discharged by a local stimulus and not by an impulse that has been transmitted from a distance. Is this local action, however, of a direct kind or does it involve a minutely circum- scribed nervous mechanism? Such a question is not to be answered by localized stimulation but must be approached by other means. Two drugs are known which completely abolish nervous activity in many lower animals, including the actinians, and which therefore may be used in testing this matter. They are chloretone and magnesium sulphate. EFFECTOR SYSTEMS OF ACTINIANS 467 If a large Metridium is put in seawater containing magnesium sulphate, in a very short time it becomes insensitive to stimu- lation and will remain so for a long time, recovering only when it is transferred to pure seawater. If from such an anesthetized animal a few tentacles are cut and these are flooded with a dilute solution of hydrochloric acid in seawater, they will discharge their nematocysts precisely as normal tentacles do. Exactly similar results can be obtained from the acontia. So far as the explosion of the nematocysts is concerned, I have been unable to distinguish between an acontium that has been anes- thetized with magnesium sulphate and a normal one. The same is true of acontia that have been treated with chloretone. After a prolonged immersion in a solution of chloretone in seawater, acontia and tentacles in which there is every reason to believe that the nervous activity is completely abolished, will still discharge their nematocysts on appropriate stimulation pre- cisely as the unanesthetized parts do. It therefore seems clear that even the circumscribed local response which has been shown to be characteristic for nematocysts is not dependent upon any influence that may be rightly called nervous, but is determined by a direct stimulation of these elements as was implied long ago by Schultze (71). If the nematocysts contain myofibrils, as was first maintained by Chun (’81) for Physalia and has since been claimed by Jickeli (’83), Schneider (790) Will (09) and others, the activity of these elements must depend, as Will has pointed out, rather upon some form of direct stimu- lation, than any thing that can reasonably be called nervous. From the preceding discussion it seems fair to conclude that the basal processes of nematocyst cells assumed by the Hert- wigs to be nervous in function are in reality not so and that these elements are without nervous connections, as, in fact, Hadzi (09) has recently claimed for them in Hydra. There seem to be, therefore, no grounds, either anatomical or physiological, for the assumption that the nematocysts are effector endorgans of the nervous system and, of course, no grounds for the view held by von Lendenfeld (’83, p. 369) that they are under the control of the animal’s will. My opinion is in entire accord THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 21, No. 4 468 G. H. PARKER with that of Wagner (’05) that nematocysts are effector organs exploded by direct stimulation and not under the control of a nervous mechanism. This view makes clear how these organoids may be appropriated by another animal, such as a nudibranch mollusk, and still retain their effectiveness, a condition which would be difficult to explain if this effectiveness was in any sense dependent upon nervous action. THE CILIARY SYSTEM There seems to be ample ground for assuming that large ciliary organs, such as the swimming plates of the ctenophores, are under the influence of the nervous system of the animals in which they occur (Parker, ’05 c; Bauer, ’10), and it is therefore natural to raise the question whether any of the cilia in Metrid- ium are thus controlled. The cilia in this and other actinians have been more fully studied than any other of the effector systems possessed by these animals and the results of such in- vestigation show in general a striking agreement. The pedal dise of Metridium is without cilia. The same is true of the column wall in ‘this form (Parker 796, p. 109; Carlgren, ’05, p. 311) as well as in Haleampa and Sagartia (Carlgren, ’05), though in Protanthea and Gonactinia it is said to be more or less ciliated (Carlgren, ’05). In Metridium (Parker, ’96, p. 112; Carlgren 05, p: 311), Sagartia (Vignon, ’01, p. 475) and Halcampa (Carl- gren, 705, p. 311) the cilia of the oral dise beat toward the peri- phery of that part, while in Protanthea and Gonactinia (Carl- gren, ’05, p. 310) their effective stroke is toward the mouth. In all actinians with ciliated tentacles thus far observed the ciliary currents are from the base to the tip of these organs (Metridium, Parker ’96, p. 110; Allabach, ’05, p. 37; Carlgren, ’05, p. 311; Sagartia, Vignon, ’01, p. 475; Carlgren, ’05, p. 311; Aiptasia, Jennings, 05, p. 453; Protanthea, Gonactinia and Hal- campa, Carlgren, ’05, pp. 310, 311). The siphonoglyphs in Metridium, whether one, two, or three, possess cilia that in- variably sweep inward (Parker, ’96, p. 113; Carlgren, ’05, p. 312) as do also those in Sagartia (Vignon, ’01, p. 475; Carlgren, 05, p. 312). This fact was long ago pointed out for the aleyo- EFFECTOR SYSTEMS OF ACTINIANS 469 narians by Hickson (’83, p. 694) who observed that in these animals the current in the single siphonoglyph was inward whereas that in the rest of the mouth was outward. The cilia on the non-siphonoglyphic portion of the lips and cesophagus in the great majority of actinians and even in some corals, ordinarily beat outward, and they may be brought by appropriate stimulation to reverse temporarily and beat inward (Metridium, Parker, 96, p. 109, 1905 a, 1905 b; Carlgren, ’05, p. 312; Allabach, ’05, p. 37; Sagartia, Vignon, ’01, p. 475, Torrey, ’04, p. 212; Tealia and Actinostola, Carlgren, ’05, p. 318; Cribrina, Gee, 713, p. 324; and in such corals as Fungia and Favia, Duer- den, ’06, pp. 596, 604, and Isophyllia, Carpenter, ’10, p.153). This reversal is accomplished by the chemical and mechanical - stimulation of food materials. In my first study of this sub- ject I was led to conclude that in Metridium marginatum ciliary reversal could not be accomplished by mechanical stimulation, as has been demonstrated by Torrey (’04) in Sagartia davisi, but subsequent work has convinced me that in specimens of Metridium that have been a week or so without food, this re- versal, as suggested by Jennings (’05), can be accomplished by mechanical as well as by chemical means, thus confirming Allabach’s statement for this species (’05, p. 37). The con- tinuous inward current mentioned by Carlgren (’05, p. 314) as occurring in Sagartia viduata was seen in animals that had been kept a week without food and was probably in reality a tem- porary reversal of an unfed animal to mechanical stimulation, for ciliary reversal has already been observed in other species of Sagartia (S. parasitica, Vignon, ’01, p. 475; S. davisi, Torrey, 04, p. 212). Whether the inward current in Gonactinia and Protandra (Carlgren, ’05, p. 310) is of this nature or not, cannot be stated with certainty, but in view of the facts just presented, it seems possible. The cilia on the mesenteric filaments in Metridium margina- tum have an effective stroke from the pedal toward the oral end of these organs and, as Torrey (04, p. 214) states, are not known to be subject to reversal. The cilia of the mesenteric convul- utions are very active and irreversible, but because of the great 470 G. H. PARKER irregularity of the courses of the ccnvolutions I was unable to determine the sense of the direction in their sweep. On the acontia the cilia beat persistently toward the free end of these organs, as already described by Torrey (’04, p. 214). Excepting for the matter of reversal, there is nothing in the action of the cilia of Metridium, or in fact of those of other acti- nians, that would lead to the suspicion that these organoids are under nervous influence; and in those that reverse their effective stroke the reversal is so strictly local in reference to the stimulus that no nervous interpretation of this special activity is suggested (Parker, 796, p. 114). The same strict agreement in the localization of stimulus and area of reversal has been pointed out by Duerden (’06, p. 596) for the coral Fungia, indi- cating that in this group also the ciliary reversal is probably - non-nervous. Not only does agreement in the distribution of stimulus and of response favor a non-nervous interpretation of ciliary action in actinians, but experiments with anesthetics also support this view. If a Metridium whose labial cilia have been shown to reverse to meat, is placed quickly in a large volume of sea- water containing some chloretone and allowed to remain there for five minutes, all traces of neuromuscular activity will dis- appear, though ciliary reversal will still occur with great pre- cision. The reversal also takes place in a perfectly clear and indisputable manner in animals that have been anesthetized with magnesium sulphate. Since these two substances complete- ly abolish neuromuscular activity and yet interfere in no essen- tial way with the ciliary reversal and other like activities in Metridium, it is safe to conclude that the cilia of this and other actinians, unlike the swimming plates in ctenophores, are quite independent of nervous control and in this respect are like the cilia in higher animals. In this particular, then, the ciliary system in Metridium is like the mucous system, and the nema- tocyst system in this animal, an effector mechanism independent of nervous control. EFFECTOR SYSTEMS OF ACTINIANS 471 THE MUSCULAR SYSTEM The muscular system in Metridium consists of at least thir- teen muscles or groups of muscles (Parker and Titus, ’16) which show a rather unusual degree of differentiation when it is re- called that this animal is one of the lower metazoans. Not- withstanding this specialization, Metridium, like most other actinians, goes into a state of such complete contraction when vigorously stimulated and remains for so long a time entirely closed that the general impression made by it is that it is a sim- ple muscular sac capable of only one general form of response. This condition is both misleading and instructive. The prolonged contraction of Metridium shows that one of the most characteristic features of actinian muscle, like that of many other invertebrates, is its very high degree of tonicity as compared with the corresponding tissue of vertebrates. This contrast has been emphasized of late by von Uexkill (09 b) and especially by Jordan (07, ’08, 712). This exhibition of excessive tonicity, however, has led inves- tigators away from a finer scrutinizing of the muscular res- ponses in actinians, responses which will be found to justify, I believe, the high degree of anatomical differentiation already pointed out. It is from this standpoint that I wish to give an account of some of the muscular reactions of Metridium. The longitudinal and parietal muscles of the mesenteries. These muscles extend lengthwise of the mesenteries and for the most part from the pedal disc to the oral disc. By their action the oral disc with its crown of tentacles is drawn down toward the attached pedal dise (Hertwig, ’79, p. 527). In the non-directive mesenteries, the longitudinal muscles are on the endocoel faces of these organs, on which they show as a thickened median band which spreads out as one approaches the oral or the pedal disc. The arrangement and concentration of the fibers in these mus- cles is especially favorable for a depression of the oral disc. On the directive mesenteries the longitudinal muscles are on the exocoel faces and each muscle is thickened as a longitudinal ridge close to the accompanying siphonoglyph. As the wall AT G. H. PARKER of the siphonoglyph in Metridium is firm, almost as stiff as a thin layer of cartilage, it is quite clear that these muscles are so concentrated as to exert a tension favorable to folding the siphonoglyph walls when the animal contracts. The parietal muscles extend up and down the length of both sides of the mesenteries next the column wall. This, like the wall of the siphonoglyph, is somewhat resistant to folding and the parietals are undoubtedly especially concerned with drawing it together in general contraction. Thus the longitudinal muscles and the parietals show a certain amount of specialization adapted to their particular tasks and yet constitute a single physiological system for the depression of the oral disc. The means by which these muscles can be brought into action are extremely diverse. Ifa Metridium is kept inrunning seawater in the dark, it soonattains to itsfullestexpansion. Under such cir- cumstances its height may be six times the diameter of its column. If now it is suddenly illuminated by diffuse daylight or a strong electric light, it will gradually shorten its column to about one third or one fourth its former length, but without contracting its oral disc. This operation is accomplished by the simul- taneous and united action of the longitudinal and parietal mus- cles and, so far as can be seen, necessitates the codperation of no other muscles. In this change of form some of the water in the gastrovascular cavity is discharged through the mouth; in consequence of the retention of the rest the diameter of the animal is increased, thereby putting the circular muscle of the column under a certain tension. Recovery from this state is more gradual than its assumption and is dependent partly on ciliary action whereby water is returned from the outside through the siphonoglyphs to the gastrovascular cavity, and partly by the action of the circular muscle of the column, which by pressing on the fluid contents of the gastrovascular cavity thus pushes the oral disc away from the pedal disc. ‘The return to the ex- panded state is doubtless also in part dependent upon the re- laxation of the longitudinal and parietal muscles, a condition which comes about when the animal is in the dark in running seawater. Whether there is any reciprocal relation of a more EFFECTOR SYSTEMS OF ACTINIANS 473 intimate kind than has been suggested (compare Sherrington, 06) between the longitudinal mesenteric muscles and _ their opponent, the circular muscle of the column, I do not know; nor can it be stated whether or not the longitudinals work against other muscles such as the transverse muscles of the mesenteries. There are many other ways beside changes in illumination by which the contraction of the longitudinal muscles of the mesenteries may be brought about. Thus a mechanical stimu- lation of the tentacles or.of the column wall or a chemical iri- tation of these parts is almost always followed by a sudden and vigorous contraction of the longitudinals, but these forms of stimulation also bring into a¢tion the sphincter and the muscles of the oral disc and the column, and thus carry the action far beyond that of a simple muscle response. So far as I am aware, a change in illumination is the only means whereby the longi- tudinal and parietal muscles of the mesenteries can be brought into action unassociated, as far as can be seen, with other muscles. Changes in illumination, moreover, are the only means which I have found to eall forth a partial activity of the longitudinal and parietal muscles. In a general illumination these muscles contract uniformly and the oral disc, retaining its horizontal position, slowly descends. If, instead of illuminating a sea- anemone generally, it is strongly illuminated from one side, it will contract much more rapidly on that side and come to rest with its oral disc turned toward the light, as already pointed out by Bohn (’06). This condition, which is the state of posi- tive phototropism of a sessile animal or of a plant, has already been observed and photographed by Hess (13, p. 436) in Cerean- thus and Bunodes, and demonstrates the partial independence of the longitudinals of one side from those of the other. The sphincter. Many methods by which the longitudinal and parietal muscles of the mesenteries are excited to action also induce an ultimate activity of the sphincter whereby the oral dise after its withdrawal becomes covered by the upper part of the column wall. This is the usual final step in complete contraction, and it is natural to inquire whether the sphincter 474 G. H. PARKER can be excited to activity without being involved in a sequences of changes making up the total act of contraction. If a number of specimens of Metridium attached normally to stones, pieces of shell, and so forth, are allowed to stand for a day or so in quiet seawater in which there is more or less de- composing material such as dead Mytilus edulis, many of the specimens will contract their sphincters even though their columns remain elongated. Though the longitudinal and parietal muscles of the mesenteries in these specimens may not be fully relaxed, they are nearly so. Certainly the only muscle in these animals which is in vigorous contraction is the sphincter, and this remains firmly and tightly closed until the animals are transferred to pure seawater. What it is in the foul water that stimulates the sphincter to independent action and what part of the body of the Metridium serves as a receptor for this stimu- lus, I have not been able to find out, but of the essentially isolated response of the sphincter under the circumstances mentioned there can be not the least question. When a Metridium is placed in the dark in a strong flow of seawater, it usually expands to its fullest extent both in the spread of its oral disc and in the lengthening of its column. If now the flow of seawater is cut off, the animal is very likely to cover the oral disc by a contraction of the sphincter without shortening its column, at least to any great extent. Thus quiet seawater following current action induces an independent con- traction of the sphincter much as foul seawater does. The sphincter is opposed chiefly to the pressure of the fluids within the actinian’s body and it is this probably that restores the sphincter on relaxation to its most expanded form. The internal pressure that is effective in this respect is due in part to the intake of water by ciliary means but particularly to the action of certain muscles such as the circular muscle of the column. The longitudinal muscle of thé acontiwum. When a Metridium has drawn down its oral dise and covered this region by the contraction of the sphincter, further stimulation is followed usually by the discharge of numerous acontia through the mouth and the cinclides. As many as seven of these thread-like bodies EFFECTOR SYSTEMS OF ACTINIANS 475 may issue through a single cinclis. If the stimulus is unilateral, the acontia are discharged chiefly on the side stimulated, as Torrey (04, p. 208) has already noticed in Sagartia. The acontia do not emerge in consequence of their own activity . but are carried outward by the streams of water that are escap- ing under pressure through the mouth and the cinclides. After the acontia have emerged they are in no sense directed toward external objects, harmful or useful, but rest in long straightish lines on the surface of the actinian or they are wafted about slight- ly by the currents of water. They gradually disappear by being drawn back into the animal. They are ciliated and the effective stroke of their cilia is vigorous enough to move them bodily and is always toward their free ends, hence they themselves are moved by this stroke back toward their attachments. The discharge and the return of the acontia, therefore, are processes in which their contained muscle plays no part. If an extended acontium is stimulated by having seawater vigorously squirted on it, in the course of one or two minutes it gradually draws itself up into a close snarl. If seawater con- taining a little meat juice is used instead of pure seawater, the snarl is more pronounced and remains for a longer time. Sooner or later the acontium untwists and straightens out preparatory to its withdrawal into the body of the actinian. The twisting and contorting of the acontium is brought about through the contractions of its longitudinal muscle the fibers of which, as already mentioned, are closely applied to the mesogleal axis of this organ. Evidence has already been advanced (Parker and Titus, 716, p. 451) to show that the acontial muscle is an independent effector and not under the influence of nerves and that no nervous transmission occurs through the acontia. The straightening out of the coiled acontium is due in part to its ciliary activity and in part probably to the elasticity of its mesogleal axis against which its muscle probably acts. The longitudinal and circular muscles of the tentacles. In a resting, expanded Metridium the tentacles are usually quiescent, radially disposed, and directed in the main away from the mouth. If a tentacle in such an animal is touched or otherwise 476 G. H. PARKER mechanically stimulated, this organ responds by turning eventu- ally toward the mouth of the animal, after which it gradually assumes its initial position. The movements thus induced are in the beginning without doubt due entirely to the activity of the longitudinal muscle but before long both longitudinal and circular muscles are certainly involved for, though there may be no nervous connection between them, their physical relations are so intimate that the activity of the longitudinal fibers may perfectly well serve as a mechanical stimulus to call into action the ‘circular system. In this instance then, the longitudinal muscle of the tentacle may act independently in the beginning, but it is quickly followed by its natural opponent, the circular muscle. If the tip of a tentacle in Metridium is cut off, the tentacle contracts, the wound almost at once closes with a nipple-like formation, and the tentacle with its tip firmly puckered gradu- ally reexpands. In this condition it remains until the injured end is fully healed (Chester, 712). At the beginning of this operation, during which the tentacle is contracted, the longi- tudinal as well as the circular muscles are involved, but after the tentacle has reexpanded and before the wound has healed, a matter of a day or so, the circular fibers near the cut are the only muscular elements really active. Thus the circular muscle as well as the longitudinal may under particular circumstances give evidence of independent action. The transverse muscles of the mesenteries. These muscles connect the column wall with the cesophagus. Ifa small amount of seawater is discharged into the mouth of an expanded, resting Metridium no response is usually noticeable. If the seawater is a z hydrochloric acid solution, the actinian immediately opens the cesophagus widely and exhibits on its column a few well marked vertical grooves. These disappear gradually as the oesophagus closes. If the position of these grooves is care- fully noted, it will be found that one is always present for each siphonoglyph and that the others are distributed in accordance with the arrangement of the other pairs of complete mesenteries. The grooves thus mark the lines of attachment of these mesen- EFFECTOR SYSTEMS OF ACTINIANS 477 teries and are the result of the contraction of their transverse muscles which, are those concerned with the opening of the cesophagus. If fragments of fish meat are put on the lips of a fully expanded Metridium, they are carried into the animal by ciliary action through an cesophagus which opens widely to receive them and during this operation the column of the animal is marked by the same vertical grooves that were seen in the experiment with acidulated seawater. As the pieces of food pass into the gastrovascular cavity the grooves fade out. It is clear, then, that the transverse muscles of the complete mesenteries are concerned with the expansion of the cesophagus for the recep- tion of food. If a piece of fish meat is placed upon the tentacles of an ex- panded Metridium, these organs become characteristically stimu- lated and if the meat is removed before it is brought ‘by the tentacles to the animal’s lips, the cesophagus will still open, accompanied by the formation of vertical grooves on the column. This response could not be elicited by the application of weak acid to the tentacles. Under such circumstances a withdrawal of the oral dise took place. Thus it appears that not every form of chemical stimulus that can be applied to the tentacles is followed by an opening of the cesophagus. To appropriate stimuli, however, the tentacles and lips may act as receptors for the opening of this tube. I have found no other parts of the body of Metridium from which I could elicit this cesophageal response. The organ that would be regarded naturally as the opponent of the transverse muscles of the mesenteries is the circular mus- cle of the esophagus. Its position is such that its contraction would bring about a closure of this tube and probably such is its action, though on this point I have no direct evidence. The internal pressure of the seawater contained in the animal, slight though it is, certainly aids materially in closing the cesophagus, as can be seen in a Metridium that is feeding. After the piece of food has passed down the cesophagus of such an animal, the walls of this organ held in the beginning more or less apart, 478 G. H. PARKER approximate gradually as though under slight pressure from the inside. Transverse muscles are found not only on complete mesenteries but also on the incomplete ones, though in this situation they are rather poorly developed. Such muscles in consequence of their failure to reach the cesophagus can have nothing to do with its expansion. When a Metridium is fed, however, one can often see on its column beside the six or eight deep grooves marking the positions of the complete mesenteries, a whole series of minute grooves which, like the others, fade out as the food is swallowed. ‘These are probably due to the transverse muscles of the incomplete mesenteries but their method of formation, and the significance of the contraction that produced them, if in fact it possesses any significance at all, have not been worked out. The circular muscle of the column. 'This muscle is unquestion- ably the chief antagonist of the longitudinal muscles of the mesenteries and acts in conjunction with them inthe contraction and expansion of the animal as a whole, but it also has its own activities. If a fully expanded Metridium is freely fed, it will usually show upon its column ring-like constrictions which form near the oral disc and proceed like a peristaltic wave over the column to the pedal disc. A new constriction appears every four or five minutes. These waves, which may have to do with the movement of the food within, have been noted by Gosse (’60, p. 253) in Haleampa, and very recently by Miilleg- ger (713, p. 487) in Metridium and Sagartia, in the latter of which they have been photographed. According to Millegger they may run from the pedal to the oral pole as well as in the reverse direction. They represent without question a specialized and individual activity of the circular muscle, the receptor mecha- nism of which has not been ascertained. Closely related to this peristalsis of the column are certain responses that can be induced in animals that are partly con- tracted but still well filled with seawater. If the column ofsuch an animal is stimulated by rubbing it lightly on a particular spot with a blunt glass-rod, a constriction begins to form at the EFFECTOR SYSTEMS OF ACTINIANS 479 spot stimulated and extends slowly around the column, an operation which is completed in about half a minute. If several spots at different heights on the column are similarly stimulated, as many as three such rings can be formed at once on the same individual. From the position and method of formation of these bands of constriction it is quite clear that they are due to the contraction of bundles of fibers in the circular muscle. If in a completely contracted Metridium a spot on the ex- posed portion of the column is mechanically stimulated, a ring of constriction passes round the animal encircling the pore which marks the location of the mouth. If a radial cut is made from near the pore out to the periphery of such an animal and care is taken that this cut goes no deeper than through the column wall, the constriction resulting from a local stimulus on reach- ing this cut fails to pass across it, showing that this activity is entirely resident in the column wall. If now a fully expanded Metridium which will contract its tentacles when a given spot on the column wall is mechanically stimulated, has this spot fully anesthetized by dropping on it crystals of magnesium sulphate, a ring of constriction will still form around its column when this spot is repeatedly touched by a blunt glass-rod. As the nervous mechanism in this portion of the animal has been rendered inoperative through the magnesium sulphate, it follows that the formation of the groove must be a purely muscular operation thus demonstrating independent action on the part of the circular muscle. The radial and circular muscles of the oral disc as well as the basilar and circular muscles of the pedal disc form natural pairs of muscles that undoubtedly act in conjunction with each other in the movements of their respective parts. I have been unable, however, to find amy means by which these muscles can be brought into action individually, though judging from the results obtained in other instances it would not be surprising if such means were sooner or later discovered. This examination of the musculature of Metridium brings out very clearly the fact that its anatomical differentiation is not without physiological significance. Jordan (’08) has already 480 G. H. PARKER called attention to the striking contrast in the rate of action between the longitudinal mesenteric muscles in Metridium and such muscles as those of the sphincter or the foot in this form; the rate at which the animal can withdraw the oral disc is in the strongest contrast to that at which it covers the disc or creeps about. Jordan is inclined to ascribe this difference to the muscles themselves, but I have brought forward evidence to show that these differences in rate are dependent in part at least upon the presence or absence of nervous connections. The acontial muscles are extremely slow in response and the forma- tion of a constriction groove by the circular muscle of the col- umn is a matter of minutes. The first of these is, I believe, absolutely unassociated with nervous activity and the second can take place after nervous activity has been temporarily obliterated. It would therefore appear that these very sluggish movements are dependent upon muscle unassociated with nerve and that when these two elements are combined in a high degree of differentiation, as in the longitudinal mesenteric sys- tem, the rate of response becomes relatively high. Another matter of general importance in the nervous activity of sea-anemones is the question of reflexes. It has been generally believed that the neuromuscular system of such an animal as Metridium could exhibit tonus and rhythmic motions, but noth- ing comparable to the sharply marked individual reflexes of the higher animals (Jordan, ’08, p. 223), and this condition is in general true. But there are, nevertheless, some responses in Metridium that are strikingly like the individualized reflexes of the higher forms. For instance when acidulated sea-water js put upon the lips and the only response that follows is the opening of the cesophagus by the contraction of the transverse muscles of the complete mesenteries, a condition is presented which, since it can be revived again and again, shows all the characteristics of a highly individualized reflex. I am fully aware that this instance and others like it imply in the nervous structure of Metridium a definiteness of conduction tracts that argues against a diffuse nervous system, but I believe these cases to be exceptional and to represent merely the first signs EFFECTOR SYSTEMS OF ACTINIANS 481 of that process by which out of an undifferentiated state the highly complex nervous organization of the higher animals has arisen. In my opinion there is not the least doubt that some of the neuromuscular responses in Metridium are true reflexes, though the majority of such operations are more usually ex- hibitions of excessive tonicity or of rhythmic motion. SUMMARY 1. The effector systems of Metridium are at least four in number: the mucous, the nematocyst, the ciliary, and the muscu- lar system. , 2. The mucous, the nematocyst, and the ciliary systems are independent effectors and are not under the control of a ner- vous mechanism. . 3. The muscular system, consisting of thirteen muscles or groups of muscles, shows a variety of conditions. Some mus- cles, such as the longitudinal muscles of the acontia, are in- dependent effectors and are not under nervous control. Others, like the circular muscles of the column wall may act indepen- dently or under the influence of nerves. Still others, such as the longitudinal muscles of the mesenteries, act only in response to impulses from a relatively complex nervous mechanism. 4. Non-nervous muscular responses are carried out sluggishly and require a minute or more for completion. Nervous muscular responses are relatively rapid and may be accomplished in a second or so. 5. Notwithstanding that the whole musculature exhibits a high degree of tonicity, there are responses such as the expansion of the cesophagus by the action of the transverse muscles of the complete mesenteries which are of the nature of well individual- ized reflexes. 482 G. H. PARKER BIBLIOGRAPHY Auuapacn, L. F. 1905 Some points regarding the behavior of Metridium Biol. Bull., vol. 10, pp. 35-48. Baauioni, S. 1913 Physiologie des Nervensystems. Winterstein, Handbuch der vergleichenden Physiologie, Bd. 4, pp. 23-450. Bauer, V. 1910 Ueber die anscheinend nervése Regulierung der Flimmerbe- wegung bei den Rippenquallen. Zeit. allg. Physiol., Bd. 10, pp. 231- 248. Boun, G. 1906 Sur les courbures dues & la lumiére. Compt. rend. Soc. Biol., Paris, Ann. 1906, tome 2, pp. 420-421. CARLGREN, O. 1905 Ueber die Bedeutung der Flimmerbewegung fiir den Nahrungstransport bei den Actiniarien und Madreporarien. Biol. Centralbl., Bd. 25, pp. 308-322. CarpENTER, F. W. 1910 Feeding reactions of the rose coral (Isophyllia). Proceed. Amer. Acad. Arts Sciences, vol. 46, pp. 149-162. CuesteR, W. M. 1912 Wound closure and polarity in the tentacle of Met- ridium marginatum. Jour. Exp. Zoél., vol. 13, pp. 451-470. Cuun, C. 1881 Die Natur und Wirkungsweise der Nesselzellen bei Coelen- teraten. Zool. Anz., Bd. 4, pp. 646-650. 1891 Die Canarischen Siphonophoren. Abhandl. Senckenberg. Na- turforschenden Gesellschaft, Bd. 16, pp. 553-627. DvuERDEN, J. E. 1906 The role of mucus in corals. Quart. Jour. Micros. Science, vol. 49, pp. 591-614. Gee, W. 1913 Modifiability in the behavior of the California shore-anemone Cribrina xanthogrammica Brandt. Journ. Anim. Behavior, vol. 3, 305-328. GuasEeR, O. C., anp C. M. Sparrow. 1909 The physiology of nematocysts. Jour. Exp. Zoél., vol. 6, pp. 363-381. Gossrz, P. H. 1860 A history of the British sea-anemones and corals. Lon- don, 8 vo, 40 + 362 pp. F GroseLs, P. 1909 Untersuchungen ueber das Nervensystem der Aktinien. Arb. Zool. Inst. Wien. Tom. 17, pp. 269-308. Hapzi, J. 1909 Ueber das Nervensystem von Hydra. Arb. Zool. Inst. Wien, Tom. 17, pp. 225-268. Hertwic, O., unp R. 1879-1880 Die Actinien anatomisch und histologise mit besonderer Beriichsichtigung des Ne rvenmuskelsystems unter- sucht. Jena Zeit. Naturwissenschaft, Bd. 13, pp. 457-640; Bd. 14, pp. 39-89. Hess, C. 1913 Neue Untersuchungen zur vergleichenden Physiologie des Gesichtssinnes. Zool. Jahrb., Abt. allgem. Zool. u. Physiol., Bd. 33, pp. 387-440. Hickson, 8. J. 1883 On the ciliated groove (siphonoglyphe) in the stomo- daeum of the Aleyonarians. Philos. Trans. Roy. Soc., London, 1883, part 3, pp. 693-705. EFFECTOR SYSTEMS OF ACTINIANS 483 Jennines, H. 8. 1905 Modifiability in behavior. |. Behavior of sea-ane- mones. Jour. Exp. Zool., vol. 2, pp. 447-472. JICKELI, C. F. 1883 Der Bau der Hydroidpolypen. Morph. Jahrb., Bd. 8, pp. 373-416, pp. 580-680. JorDAN, H. 1907 Ueber reflexarme Tiere. Ein Beitrag zur vergleichenden Physiologie des zentralen Nervensystems, vornehmlich auf Grund von Versuchen an Ciona intestinalis und Oktopoden. Zeit. allg. Physiol., Bd. 7, pp. 86-135. 1908 Ueber reflexarme Tiere. II. Stadium ohne regulierende Zen- tren: Die Physiologie des Nervenmuskelsystems von Actinoloba dianthus Ellis (Fuss, Mauerblatt, Septen, Nervennetz der Mund- scheibe). Nebst einigen Versuchen an Fusus antiquus. Zeit. allg. Physiol. Bd. 8 pp. 222-266. 1912 Ueber. reflexarme Tiere (Tiere mit peripheren Nervennetzen). ILL. Die acraspen Medusen. Zeit. wiss. Zool. Bd. 101 pp. 116-138. LENDENFELD, R. y. 1883 Ueber Coelenteraten der Siidsee. III Mittheilung. Ueber Wehrpolypen und Nesselzellen. Zeitschr. wiss. Zool., Bd. 38, pp. 355-371. Mtiieccer, 8. 1913 EHigenartige Bewegungserscheinungen bei Aktinien. Blit- ter Aquar. Terrarienkunde Jahrg. 24, p. 487-488. Parker, G. H. 1896 The reactions of Metridium to food and other sub- stances. Bull. Mus. Comp. Zodél., vol. 29, pp. 107-119. 1905a The reversal of cilary movements in Metazoans. Amer. Jour. Physiol., vol. 18, p. 1-16. 1905 b The reversal of the effective stroke of the labial cilia of seaanemones by organic substances. Amer. Jour. Physiol., vol. 14, pp. 1-6. 1905 e The movements of the swimming-plates in Ctenophores, with reference to the theories of ciliary metachronism. Jour. Exp. Zodl., vol. 2, pp. 407-423. Parker, G. H., ann E. G. Tirus 1916 The structure of Metridium (Actino- loba) marginatum Milne-Edwards with special reference to its neuro- muscular mechanism. Jour. Exp. Zoél., vol. 21, pp. 483-458. ScHNEIDER, K. C. 1890 Histologie von Hydra fusea mit besonderer Beriick- sichtigung des Nervensystems der Hydropolypen. Arch. mikros. Anat., Bd. 35, pp. 321-379. 1902 Lehrbuch der vergleichenden Histologie der Tiere. Jena, 8 vo, 988 pp. Scuuuze, F. E. 1871 Ueber den Bau und die Entwicklung von Cordylophora lacustris (Allman). Leipzig, 4to, 52 pp. SHERRINGTON, C. 8. 1906 The integrative action of the nervous system. New York, 8 vo. xvi + 411 pp. Torrey, H. B. 1904 On the habits and reactions of Sagartia davisi. Biol. Bull., vol. 6, p. 2038-215. UEXxkKiLt, J. d’ 1909 a Résultats des recherches effectuées sur les tentacules de l’Anemonia sulcata au Musée Océanographique de Monaco, en décembre 1908. Bull. Inst. Océanograph., Monaco, 1909, No. 148, 3 pp. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 21, NO. 4 484 G. H. PARKER UexktLu, J. von 1909 b Umwelt und Innenwelt der Tiere. Berlin, 8 vo, 259 pp. ViaNon, P. 1901 [Ciliary curents in Actinians.] In Y. Delage et E. Hérou- ard, Traité de Zoologie Concréte, tome 2, partie 2, Les Coelentérés, p. 475. WacGner, G. 1905 On some movements and reactions of hydra. Quart. Jour. Micros. Science, vol. 48, pp. 585-622. Wirt, L. 1909 Ueber das Vorkommen kontraktiler Elemente in den Nessel- zellen der Coelenteraten. Sitzb. Abhand. Naturforsch. Gesellsch. Rostock, N. F., Bd: 1, pp: 33-52. Wourr, M. 1904 Das Nervensystem der polypoiden Hydrozoa und Seyphozoa. Zeit. allg. Physiol., Bd. 3, pp. 191-281. EXPERIMENTAL STUDIES ON THE ORIGIN OF MONSTERS! I. AN ETIOLOGY AND AN ANALYSIS OF THE MORPHOGENESIS OF MONSTERS? K. I. WERBER Osborn Zoological Laboratory, Yale University (From the Zoological Laboratories of Princeton and Yale Universities) EIGHTY-NINE FIGURES CONTENTS I, LDRO GIGLI SAW ee, coed een oe cicittc otras Peer eerie CITE emer Cee ce eee aete 486 MSP CREE COLE CG GOL UG Art cs. 8 a5)s Ga Se be tehey atlas wakes OR © Cini = Ayo scc ep ayeions S44 © 487 PMP CRAG ATOl WO MCS C Ay. siete psi ila Pn wae ee ie yyy Rel 489 dee En erop Mala lm Ch ve Lavan cece wre enact oe Mines Rive 489 pee CLE Chi Ol euler INOW Att lyse XAG a Noe wa aah oles BER RA eae 492 CD erectsormineloltactoby. DitSa-ee esac ote no a eer eousre 495 dee efecistoletherearavesicGless er emia ry orient eee 497 Pe ieetINOLP NOUS TCMADEYVOS-.intso ooo Ses oa oe See a eae cate eles Oe 498 III. The microscopic anatomy of the monsters and conclusions regarding (tavestig Tel oNrya) nVoyeteVOVERSIS Ns Gig yte 8 Geils cues o Aree aie cat eee te EIN eo erie 500 Ae iClassification: of ophthalmic teratia..:..0.. 052 c+. 0stae eet een et 500 Bite morphology of teratophthalmias. 1.22252... is 456 oie ee eds 502 ae unopiutialrinian Wilentiea: | f.20 ees - se ak es has esteacte eer ie 503 banynophbhaimiarund enticaenn fasta 4s eae ae ate meee ae 503 caCycloprasynophthalmiat serena ey acticin. sec cine enee hie 507 daGvclopiaypenmectastey terrier hgh eerie oer 510 es Monophthbalaidrasymmetricas. a)... sas aianvae ote oko ds othe 513 fee Microphthalmrapandeanophiihalmiareeseeenee ss ele eee ce eee 519 C. The morphogenetic factors underlying the eye terata............. 521 PS ssOekara S Inbirbvon theOry «00 sec. cele dees sess Seis oes ube 521 2. The defect theory of teratophthalmia.....................-- £ 528 ' Aided by a grant from the Bache Fund of the National Academy of Sciences. > Brief reports of the results of this work have been presented and demonstra- tions of material made at the meeting of the American Society of Zoologists in Philadelphia (December 29 and 30, 1914) and at the meeting of the American Association of Anatomists at New Haven (December 28-30, 1915). A preliminary communication was also published in the Anatomical Record (ef. Werber 715 c.) 485 486 E. I. WERBER Db Delormities ofthe brane’ commen ess ia il.: ake nc pee ee mea E. The microscopic analysis of some teratomata (the ‘solitary eye’ andethe: isolated eye nn: arse: & vob ok. 2. . cys of See eee 541 F. The microscopic anatomy of some amorphous monsters.......... 553 LVe Concluding remarks. ..... see eee oes coe eee eee 558 VFS UMN RY 8 sain «50k es hee oat eRe eis (eck Soest ae 3g eee 568 Wie hiterature.eited ;..02. t.2.1eee eet ee Pee ees s vas a ee 571 I. INTRODUCTION A review of the literature shows us that we are at the present time confronted with two theories regarding the causal genesis of monsters. The first and older one, the amniotic theory, is essentially a mechanical theory, while a more recent one which is based on recent experiments on the influence of chemical alterations of the egg’s environment, might be spoken of as the chemical theory of teratogenesis. The amniotic theory maintains that terata are due to anoma- lies of the amnion, which latter, by adhering too closely to the embryo or constricting it, is thought to bring about the various well known malformations. The inadequacy of this assump- tion has in recent years been repeatedly pointed out.’ It is a well known fact that amniotic anomalies are found relatively rarely in malformed ova. In one hundred and sixty-nine patho- logical ova which Mall (’08) has examined he asserts that he has found not a single case of anomaly of the amnion. Wherever amniotic adhesions are found, Mall, I believe rightly -regards them as incidental. It is not impossible that they may be syngenetic with some terata, 1.e., due to the same causes which brought about the developmental deviation of the embryo. This might particularly apply to the experimental terata of Dareste (91) and others, in which anomalies of the amnion were found. On the whole, it might well be said, that in view, particularly, of experimental results on anamniotes, the amniotic theory would hardly seem to deserve more than historical interest. In rejecting this mechanical theory Mall traces monstrous development to faulty implantation of the ovum in a diseased uterus, which in turn makes adequate nutrition of the embryo 3 Cf. Mall (08) and Jordan (’09). ORIGIN OF MONSTERS 487 impossible. This is, essentially, a chemical theory of terato- genesis for it assumes the atypical development of the embryo to be due to lack of necessary substances. More direct support for the chemical theory of teratogeny is offered by the recent investigations of Stockard (’07, ’09, ’10 a) and McClendon (12a and b), both of whom obtained one-eyed monsters by subjecting developing teleost eggs to the action of various toxic substances, such as magnesium chloride, alcohol, ether, alkaloids, etc. These investigations have shown that some such developmental deviations as are found to occur spontane- ously may be brought about by the chemical action of various substances, and thus they suggest that atypical development in nature may be due to pathochemical alterations of the germ’s environment. The above considerations have led me to assume that in order to attack the problem of atypical development in nature effec- tively it is necessary to find the unusual chemical factors which cause the embryo in its natural environment to develop in a defective or monstrous manner. Since the metabolism is the greatest source of chemical modifi- cations of the body, I concluded that the solution of the problem of the causal genesis of monsters must be sought for in pathologic parental metabolism. | Starting from this assumption, I carried out, in the summer of 1914, some experiments on Fundulus heteroclitus, the ferti- lized eggs of which were subjected to the action of solutions of urea, butyric acid, lactic acid, acetone, sodium glycocholate and ammonium hydroxide. Conclusive results were so far obtained only with butyric acid and acetone. The (rather simple) methods employed have been described in a former paper (Werber ’15 c) ‘to which the reader is referred. | II. THE RECORDED TERATA The results which were obtained are very much alike in both series of experiments, with butryic acid and acetone. The variety of deformities being almost endless in both, it would be practically impossible to present much more than certain types Fig. 1 Normal embryo of Fundulus heteroclitus, nine days old. h., heart. Fig. 2. Synophthalmia bilentica, from j!; gram molecular butryric acid, twenty-eight days old. pc., pericardial vesicle. Fig.3 Snyophthalmia bilentica, from 7'5 gram molecular butyric acid solu- tion, twenty-eight days old. Fig. 4 Synophthalmia unilentica, from ;'; gram molec. butyric acid, twenty- four days old. Fig. 5 Cyclopia synophthalmica, from acetone solution (35 cc. to 50 ce. sea- water) twenty-six days old. Fig. 6 Cyclopia perfecta, from 7/5 gram molee. butyric acid, twenty-four days old. Jl., lens; p., pigment epithelium; h., heart. Fig. 7 Cyclopia perfecta, from acetone solution (35 cc. gram molec. to 50 ce. sea-water) twenty-five days old. Fig. 8 Cyclopia perfecta, from 4); old. e.v., ear vesicles. Fig. 9 Cyclopia perfecta, from 3‘; gram molec. butyric acid, twenty-eight days old. gram molec. butyric acid, thirteen days 488 ORIGIN OF MONSTERS 489 ol them. In a previous publication (Werber 715 ¢), I have pre- sented a brief survey of the recorded terata.. While a more complete presentation of them is attempted in this paper, I am fully convinced that it is inadequate to convey the proper im- pression of the great range of variation in the noted effects. 1. Terata of the head The malformations of this part of the body are many and unusually varied. Indeed, as will be pointed out later, this part of the body appears to be the most susceptible one to the influence of toxic solutions. The deformities of the head affect the sense organs, the brain, the mouth, and the skull. All of them are usually found to occur in various combinations. ‘The most striking ones on examination in toto are those that concern the eyes. a. The ophthalmic terata. When Fundulus eggs are subjected to the action of butyric acid or acetone in the concentrations stated above, embryos with normal eyes are of the rarest occur- rence. Cyclopia, i.e. the presence of a single median eye is found very often. I have likewise recorded in my observations a wide range of intermediate stages between two normal eyes in the typical position in the head all the way down through more or less closely approximated eyes or eyes of an apparent double composition and true cyclopia to complete anophthalmia as described so often by the older teratologists and as obtained experimentally in recent years by Spemann (’04), Lewis (09), Stockard (09, 710 b), and others. A comparison of figures with the normal embryo in figure lL is very instructive of this gradual transition from the normal eyes to various synophthalmiec or cyclopean defects. In figure 2 an embryo is seen with eyes apparently normal, but for their position. They are located on the frontal part of the head in approximation to one another and, while not being fused exter- nally, they are found to be so on examination of sections at a more posterior level. A more intimate approximation of the eyes is to be seen in figure 3. Here the eyes already are con- A490 E. I. WERBER Fig. 10 Normal Fundulus embryo, eighteen days old, five days after hatching. Fig. 11 Synophthalmia bilentica, from >‘; gram molec. butyric acid, thirty- four days old, three days after hatching. Fig. 12. Cyclopia synophthalmica, from acetone solution (35 ce. gram molec. to 50 cc. sea-water) twenty-nine days old, one day after hatching. ORIGIN OF MONSTERS 49] tiguous medially, but not fused. On microscopic examination it was found, however, that there is a fusion of the median parts of the eyes which becomes the more distinct the more posterior the section examined. In the embryo presented in figure 4 one median composite eye is to be seen, the components facing each other and enclosing one lens. Cases of cyclopia, i.e. embryos in which the single median eyes on examination in toto do not present any evidence of being composite in character are presented in figures 5 to 9 and 12. A comparison of these figures shows that the cyclopean eye may vary considerably in size as well as in other respects. It may be very large, much larger than a normal eye (fig. 5) or of about the size of the latter or even smaller, and sometimes, indeed, very minute (fig. 8). Furthermore, the cyclopean eye may have the appearance of a normal eye, or it may vary in that respect. Thus, for instance in figure 9 is seen an embryo with a cyclopean eye lacking a lens, but instead showing a very dis- tinct ventral duplicature of the pigment layer, which was verified on microscopic examination. The ectoderm above the cyclopean optie vesicle was probably defective and the optic vesicle has evidently come into too close contact with the yolk, hence the lack of a lens and the duplicature of a part of the wall of the optic cup. Or again, in figure 6 the cyclopean eye of the embryo exhibits a very striking peculiarity of another kind. Here the eye is seen to be irregular in form, anteriorly it lacks the pupil and is entirely surrounded by the pigment layer, while the lens is seen to be on the postero-lateral aspect of the eye. Besides the synophthalmic or cyclopean defect I have fre- quently found embryos with a single eye in the usual lateral position of the head (figs. 14 to 17) and of apparently normal Fig. 13 Embryo with unilaterally defective head, one normal and one rudi- mentary. eye, from acetone solution (35 cc. gram molec. to 50 cc. sea-water) eighteen days old. Fig. 14 Monophthalmia asymmetrica, from acetone solution (35 cc. gram molec. to 50 cc. sea-water), nineteen days old. Fig. 15 Monophthalmia asymmetrica, from acetone solution, (35 ce. gram molec. to 50 cc. sea-water) nineteen days old, m., mouth. 492 E. I. WERBER structure. As a serial forerunner of this type of monstrosity may be considered embryos in which one eye is normal while the other is rudimentary (fig. 13). As illustrative examples of variation in the degree of the eye defects may further be presented some cases in which both eyes are of unusually small size (microphthalmic) and often located on the dorsal side of the head (figs. 18 to 21). b. Defects of the mouth. It is a well known fact that in the eyclopean or synophthalmie embryos of man and other mammals the nose is almost invariably abnormal in shape, structure and Fig. 16 Embryo with one lateral, malformed eye with free lens, l. on eye- less side, from acetone solution (35 ec. gram molec. to 50 cc. sea-water), hatched prematurely on twenty-fourth day after fertilization. Fig. 17 Embryo with one normal and one heterotopic eye, h. e., from acetone solution (25 ce. gram molec. to 50 ce. sea-water), hatched prematurely on twenty- eighth day after fertilization. position. It has the form of a proboscis, the nasal passages are more or less blended, a rudimentary septum sometimes being present, while it often may be lacking. Its skeletal parts, if present, consist of cartilage. It is usually situated in the fore- head where it hangs down over the cyclopean eye. In many teratophthalmic embryos of my experiments a very similar deformity is exhibited by the mouth. The malformation is an unusually striking one. The mouth in the normal embryo (figs. 1 and 10) is broad and flattened and antero-median in rela- tion to the eyes which are situated laterally in the head. In synophthalmic and cyclopean (figs. 2, 3, and 12) and sometimes also in esymmetrically monophthalmic embryos the mouth has ORIGIN OF MONSTERS 493 the appearance of an elongated snout or proboscis-like structure. On sections of such embryos it was found that this ‘proboscis’- mouth is apparently capable of functioning, for the continuity of the oral cavity with the pharynx and oesophagus is nowhere interrupted. Stockard (09) who has also obtained this abnormality of the mouth in cyclopean embryos with magnesium chloride in the same material suggests that this condition is due to the fact that the single antero-median eye occupies the position normally assumed by the mouth and thus ob- structs the usual forward growth of its structures. The mouth, there- fore, remains ventro-posterior to the eye and grows downward, pre- senting the proboscis-like appearance. This interpretation of the ‘proboscis’-mouth as secondary to the ‘eyclopean’ condition does not appear to be justified, for I have recorded the occurrence of this anomaly of the mouth not only in cyclopean and synophthalmic, but also in asymmetrically monophthalmic and in some two-eyed microphthalmic embryos (fig. 21). In neither of such cases can the eye or eyes be said to occupy the position which is normally taken by the mouth. Moreover, the very opposite may occasionally be found to occur, viz., that in some asymmetrically monophthalmic embryos (fig. 15) the mouth may take the place which should normally be assumed by the lacking eye. There is good reason to believe that the ‘proboscis’-mouth results from approximation and par- tial fusion of the potential anlagen of the maxillary and man- dibular arches, following an injury which has destroyed inter- mediate parts. This becomes strikingly evident on examination of sections. Our view is further strengthened by Spemann’s (04) findings who, by constricting amphibian eggs, has produced ophthalnic deformities and the ‘proboscis’-shaped mouth in the same embryos. In Spemann’s experiments the destruction of material inter- mediate between the potential eyes, the potential maxillary and mandibular arches could not, by any means, be doubted, for here the relation between a well defined mechanical injury and the resulting morphological defect is evident. Moreover, on exami- 494 E. I. WERBER ORIGIN OF MONSTERS 495 nation of sections of his monsters Spemann (’04, p. 433) finds that the ‘Kieferbogenfortsitze’ are ‘‘in der Mitte zu einem ver- schmolzen.’’* That the defects of the mouth are syngenetic with those of the eyes is well suggested also by such embryos where the eyes are small and in a very imperfect condition and the mouth exhibits an anomaly analogous to the ‘hare lip’ or ‘cleft palate’ of man (figs. 18 and 19). This syngenesis of defects of the head is further suggested by the c. Defects of the olfactory pits. The condition of these organs in teratophthalmic embryos is strikingly similar to the defects exhibited by the eyes. All degrees of approximation and blend- Fig. 18 Microphthalmic embryo with ‘hare lip,’ from acetone solution (35 ec. gram molecular to 50 cc. sea-water), thirty-one days old, p.c. pericardial vesicle. Fig. 19 Like 18, twenty-four days old. Fig. 20 Microphthalmic embryo without fins, with club-tail, from acetone solution (40 ce. gram molec. to 50 ec. sea-water), thirteen days old. Fig. 21 Microphthalmic embryo, with ‘proboscis’-shaped mouth, supernu- merary lens, s.l., and one pectoral fin only, p.c., pericardial vesicle. Fig. 22. Embryo with one lateral rudimentary eye, free lens, /. on eyeless side, oedamatous ear vesicles and a supernumerary pectoral fin, s.f., from ace- tone solution (40 ce. gram molec. to 50 ce. sea-water), thirty-four days old. Fig. 23. Anophthalmic embryo without fins and with club-tail, from acetone solution (85 cc. gram molec. to 50 cc. sea-water), twelve days old. Fig. 24 Anophthalmic, malformed embryo with rudimentary pectoral fins and club-tail, from acetone solution (20 ce. gram molec. to 50 ce. sea-water), fourteen days old. Fig. 25 Greatly malformed embryo, with small eyes, without fins, from acetone solution (35 cc. gram molec. to 50 cc. sea-water), twelve days old. Fig. 26 Amorphous, oedematous embryo, with rudimentary eyes, from ace- tone solution (40 cc. gram molec. to 50 cc. sea-water); fourteen days old. Fig. 27 Deformed embryo, with short posterior part of body and rudimentary eyes, from acetone solution (35 cc. gram molec. to 50 ce. sea-water), thirteen days old p.c., pericardial vesicle. Fig. 28 Greatly malformed, elongate embryo with waist-like constrictions, with one vestigial eye and club-tail; from acetone solution (35 cc. gram molec. to 50 ec. sea-water), thirteen days old, p.c. pericardial vesicle. Fig. 29. Greatly malformed embryo, with waist-like constrictions, one median rudimentary eye, rudimentary pectoral fins, club-tail and a very small isolated eye, 7.e., on yolk-sac, from acetone solution (40 ce. gram molec. acetone to 50 cc. sea-water), thirteen days old. 4 My own italics. WERBER ifr 496 ORIGIN OF MONSTERS 497 ing of both pits ean be found in various synophthalmic and eyclopean embryos. This observation was made also by Stock- ard on his ‘magnesium-embryos,’ but he has not drawn the con- clusion to which the data would seem to point. In a subsequent section of this paper dealing with the microscopic anatomy of some terata the attempt is made to account for the teratoph- thalmic condition, the anomalies of the mouth and olfactory organs on the basis of a morphogenetic factor common to all of these as well as to other deformities in the head region. d. Defects of the ear vesicles. The deformities to which this organ is subject under the influence of the toxic solutions em- ployed in my experiments, vary considerably. Already on ex- amination in toto of most teratophthalmic or otherwise deformed embryos it can be seen (figs. 2, 6, 8, 22) that the ear vesicles are greatly distended and sometimes reach enormous size. In some embryos there may occur a fusion of both ear vesicles, a condi- Fig. 30 Amorphous, anophthalmic embryo, with two isolated tissue frag- ments, ¢./1. and ¢.f2., on yolk-sac; from acetone solution (35 ec. gram molec. to 50 ec. sea-water), thirteen days old, p.c., pericardial vesicle. Fig. 31 Amorphous, anophthalmic embryo from acetone solution (40 ce. gram molec. to 50 ec. sea-water), fourteen days old. Fig. 32. Dwarfed, malformed embryo with one lateral rudimentary eye, from iy gram molec. butyric acid, eighteen days old. Fig. 33 Amorphous embryo, from acetone solution (40 cc. gram molec. to 50 cc. sea-water), fifteen days old. Fig. 34 Amorphous embryo from acetone solution (40 ce. gram molee. to 50 ce. sea-water), fourteen days old. Fig. 35 Meroplastic embryo with rudimentary eyes, from acetone solution (35 cc. gram molec. to 50 ce. sea-water), eighteen days old. Fig. 36 Meroplastic embryo with one rudimentary lateral eye and oedematous ear-vesicles, e.v., from >> gram molec. butyric acid, eighteen days old. Fig. 37. Head-meroplast with rudimentary eye; four small tissue fragments, t.f., on yolk, from acetone solution (40 ce. gram molec. to 50 ce. sea-water) four- teen days old. Fig. 38 Egg with oedematous eye-teratoma (‘solitary eye’), oe., oedema of the brain fragment; from acetone solution (35 cc. gram molec. to 50 ec. sea- water), thirteen days old. Fig. 39 Egg with eye-teratoma (solitary synophthalmia), s.e., from acetone solution (40 cc. gram molec. to 50 ec. sea-water), thirteen days old. Fig. 40 Egg with ‘solitary eye’ s.e., and three blastolytic tissue fragments t./., from acetone solution (35 cc. gram molec. to 50 ec. sea-water), twelve days old. Fig. 41 Egg with eye-teratoma (‘solitary eye’), from acetone solution (35 ce. gram molec. to 50 cc. sea-water) twelve days old. 4908 E. I. WERBER tion known in human teratology as synotia. Not infrequently, however, the ear vesicles may be unusually small, and in such cases on microscopic examination it may be found that the semi- circular canals are very defective, rudimentary or diminutive in size or that one or two of them may be lacking altogether. Not many embryos could be tested for their capacity of main- taining the equilibrium while moving about, since few of them would hatch if the eggs were treated with butyric acid or acetone. However, upon several of them, which did hatch, the observa- tion was made that they could swim only in circular or spiral Fig. 42 Asymmetrically monophthalmic embryo, with club-tail, without pec- toral fins. On the yolk-sac at a distance from the embryo is seen an isolated tissue fragment ¢.f., and an isolated eye 7.e., from acetone solution (40 ec. gram molec. to 50 cc. sea-water), twelve days old. Fig. 43 Asymmetrically monophthalmie embryo with isolated eye, 7.e., from acetone solution (25 ec. gram molec. to 50 cc. sea-water), sixteen days old. Fig. 44 Amorphous embryo, in toto making the impression of malformed coalesced twins, from acetone solution (25 ec. gram molec. to 50 cc. sea-water), twelve days old. lines, or along the wall of the fingerbowl in which they were kept, while they could not move in a straight forward direction, dropping at once to the bottom of the dish, if forced to do so. This functional anomaly agrees well with the structural defects of the semicircular canals spoken of above. 2. The amorphous embryos This group extends over a wide range of monstrous embryos. It begins with forms which on examination in toto in their shape, size and structural peculiarities only faintly suggest the resem- ORIGIN OF MONSTERS 499 blance with fish embryos of a corresponding age, and goes all the way down to such forms where such gross-morphological similarity is entirely obliterated (figs. 24 to 34). Thus embryos may be found with rudimentary eyes in which the anterior part of the body has widened out to a very striking degree, with club-fins and a club-tail. Not uneommon is the occurrence of greatly elongated and misshapen embryos often with waistlike constrictions (figs. 28 and 29), with rudimentary fins or (sometimes) dislocated fins (figs. 29 and 31). This group would comprise also some remarkably misshapen and dwarfed embryos (figs. 25, 26, 27 and 32) and finally embryos whose form would almost seem to suggest a similarity to some invertebrate animals (figs. 33 and 34). 3. Meroplastic embryos Very great numbers of eggs were found both in butyric acid and acetone experiments in which parts of the bodies have developed, while the rest of the germ has apparently suffered destruction (‘Meroplasts’-Roux, |. c.). In such meroplastic embryos the same wide range of variation obtains in the degree of defect, as in all other monsters recorded above. Common to nearly all of them, however, is that, usually it is the anterior part of the body that develops, while the pos- terior part is lacking. The variations in defect concern the organs of the meroplasts eas well, as the quantity of what has developed. A glance at figures 35 and 36 shows that the eyes mey be variously defective and the same obtains for the ear- vesicles. Likewise the shape and size of the meroplasts are very inconstant. The meroplastic embryo may correspond to more than the anterior half of an embryo or be just about half of the embryo’s body. The latter ones recall the hemiembryos which Roux (95) and later other investigators have obtained by injur- ing one of the first two blastomeres of the frog’s egg. Finally the meroplasts should be mentioned which are less then half of the body (figs. 37 to 41). Not infrequently all that can be seen to have developed is a more or less malformed head, recognized only by the presence of a rudimentary eye (fig. 37). THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 21, NO. 4 500 E. I. WERBER Even smaller, usually amorphous, meroplasts can be found in some eggs as the only evidence of development. As the most significant of all the meroplasts recorded may be regarded eggs in which nothing could be found on the yolk-sac besides a very small tissue fragment with an eye (‘solitary eye’) (figs. 38 to 41). The morphology of such ova and their ontomechanical signifi- cance will be discussed under a subsequent heading. * * Of otber deformities recorded in these experiments several dcuble monsters would seem to deserve description. Owing, however, to limitations of space, the presentation of these obser- vations is reserved for a paper (soon to be published) which is to deal with a rather large number of various duplicities recorded in more recent experiments. IiI. THE MICROSCOPIC ANATOMY OF THE MONSTERS AND CON- CLUSIONS REGARDING THEIR MORPHOGENESIS This part of the work has yielded some very encouraging results. It has, as I hope to make clear, contributed not only a basis for rational interpretation of the morphogenesis of some monsters, but also disclosed some general principles which may underly the genesis of all malformations recorded in my experi- ments, and possibly of such as occur spontaneously in a state of nature. While for this reason an extensive study of these terata would seem very desirable, I have so far not been able to study more than a few common types. A more extensive treat- ment of these experimental terata will follow this communica- tion in further studies. In the present study I have given most attention to the mor- phology of ophthalmic terata. A. CLASSIFICATION OF OPHTHALMIC TERATA There being, as I have mentioned before, a very great variety of these deformities, a classification of them, although of neces- sity arbitrary to some extent, would seem desirable. For, it is obvious that the term ‘cyclopia’ in its present use does not ORIGIN OF MONSTERS D01 define any one of the conditions which it is meant to cover. A similar objection must be made to the term ‘semi-cyclopia’ (Gemmill, ’12) for, the terata, which this term would define, are of many kinds, and while gross-morphologically similar, they mostly differ in important structural detail. It is also hardly necessary to emphasize that the term ‘hour-glass eye,’ so com- monly employed, is wholly unscientific, for it utterly lacks the precision that is necessary to define a morphological condition. Fairly serviceable classifications of the various eye monsters have been made by Vrolick (’49), Kundrat (’82) and Bock (’89). However, these classifications being based on mammalian terata and considering not only defects of the eye but also the corre- lated defects of the nose and skull are lacking in precision (even if applied to mammals only), and do not well lend themselves to other vertebrates. A rational classification which would con- sider the deformities of both these organs appears to be well- nigh impossible. I have, therefore, attempted the following simple classification of the teratophthalmic (‘cyclopean’) monsters with the con- dition of single- or two-eyedness respectively as the only and basic morphological criterion: fa. bilentica I. Synophthalmia \b a aie ie (a. synophthalmica. lb. perfecta. III. Monophthalmia asymmetrica (s. lateralis) Il. Cyclopia s. Monophthalmia mediana Under ‘synophthalmia’ will be classified cases of either more or less approximated, closely approximated eyes, or eyes so fused that the composite character of the organ is easily discernible in toto. If such an optic organ should possess two symmetri- cally placed lenses, the term ‘synophthalmia bilentica’ will be applied, while, if'on more intimate fusion of the eye components only one lens should be present, ‘synophthalmia unilentica’ will be used as the descriptive term. The term ‘cyclopia’ will be applied where a single median eye is present, which on examination in toto does not present the appearance of a composite eve. If, however, such an eye on 502 E. I. WERBER microscopic examination of sections is found to consist of two blended eye components, I shall term it ‘eyclopia synophthal- mica.’ To this category belong by far the greatest number of eyclopean monsters. Of rare occurrence in my experiments (as well as in the cases of spontaneous cyclopia described by various authors) were found to be cyclopean eyes which on microscopic examination were single throughout and nowhere suggesting the possibility of their being composite in character. Such cases I shall hence- forth term ‘perfect (‘true’) eyclopia’ (Schwalbe’s ‘cyelopia com- pleta’). In perfect cyclopia the eye is, as a rule, more defective than in synophthalmic cyclopia. It may be very small in size (microphthalmic) or some of its structures may be more or less defective or even lacking entirely. In perfect cyclopia the de- feets of the brain and other defects are usually much greater than in synophthalmic cyclopia. The term ‘Monophthalmia asymmetrica’ has been in use since its introduction by Ahlfeld (80-82) and would seem to need no further comments. The embryos of this group possess one eye in the usual lateral position of the head. B. THE MORPHOLOGY OF TERATOPHTHALMIA A fairly extensive study of teratophthalmic embryos in sec- tions has enabled me to make many observations which directly or indirectly point to certain dynamic factors underlying their formation. It was found that these embryos sustain at an early stage of development an injury mainly in a restricted area of the anterior end of the future embryo’s body which eventually leads to the formation of the terata of the eye. In the follow- ing I shall now present anatomical deseriptions of various types of teratophthalmia and such evidence will be pointed out as may reasonably be adduced to the interpretation of their morpho- genesis. Our description will begin with bilentic synophthalmia, where the defect is yet relatively slight and take up successively the more extreme malformations of the ‘cyclocephalic group’ through cyclopia all the way down to anophthalmia. The anatomy of asymmetric monophthalmia will also be considered, ORIGIN OF MONSTERS 503 which, while being outside of the ‘cyclocephalic group,’ yet ex- hibits ample evidence that it owes its formation to like morpho- genetic factors. a. Synophthalmia bilentica. One of the embryos which we have chosen for the presentation of this malformation is seen in toto in figure 3 (p. 488). The eyes are large, not fused, but very closely adjacent. On microscopic examination it is seen that in anterior sections the eyes are separate (fig. 63). If, how- ever, the whole series be examined, it is found that at about the level at which the lens appears in the sections, the medial mar- gins of the eye-bulbs begin to blend; and when followed more posteriorwards this fusion becomes so intimate as to manifest itself in a unition of the retina of one eye with that of the other. The eyes appear to be otherwise normal in structure and two optic nerves are present which are seen to enter the optic lobes after having formed a chiasma (fig. 64). Examination of the entire series of sections of this embryo suggests that the injury sustained by it which is responsible for its ophthalmic malformation was apparently restricted to the most anterior part of the future embryo’s body. This is evi- denced by the following data. The abnormalities such as are found to characterize this embryo concern the mouth, the olfac- tory organ and the most anterior part of the brain, viz., the fore-brain. As seen in figure 3 the mouth is a typical proboscis, and the olfactory pits are in sections seen to be perfectly blended into one large pit (fig. 63). The fore-brain is abnormal in struc- ture and unpaired (fig. 63), while the mid- and hind-brain are bilaterally symmetrical and apparently normal in other respects (fig. 64). On following out the whole series of sections no other abnormality can be detected. The injury sustained is thus very clearly seen to be restricted to the embryo’s anterior end. Its probable nature and its morphogenetic consequences will be pointed out in the course of the following description of other embryos where a like teratogenetic principle seems to obtain. b. Synophthalmia unilentica. The embryo selected for the description of this deformity is from the same experiment as the preceding one and was twenty-four days old when killed. In 504 E. I. WERBER toto (fig. 4) it almost made the appearance of a perfectly cyclo- pean monster, there being only a single median eye present with one centrally located lens. The pigment wall of the eye-bulb, however, was strikingly abnormal, because presenting a figure similar to two C’s blended at their opposite ends, it indicated Fig. 45 Camera lucida drawing of a transverse section through the eye of the embryo in figure 4. ¢.s., tissue spaces; f., fibrin. X 1733 the composition of the eye of the ophthalmoblastic materials of both sides. On microscopic examination of this embryo the following find- ings were noted. The eye is composed of two incomplete optic cups facing each other and enclosing a single lens of about the usual size (fig. 45). The cornea and iris are apparently normally developed. The anterior chamber of the eye is absent and the ORIGIN OF MONSTERS D05 vitreous body only barely suggested. The retina, while being well differentiated, is abnormal in some of its layers, which are to a considerable extent intermingled with the fibrous layer. Some of the retinal layers of the two optic cups are continuous in sections through the anterior part of the eye (fig. 45) while no such continuity can be observed more posteriorly (fig. 46). Two optic nerves are seen to pass out of the eye in few and loose SRA eS -, c= = = Fig. 46 Camera lucida drawing of a more posterior section through the eye of the same embryo (region of optic lobes). t.s., tissue spaces; e.oe., extra- cerebral oedema; 0.c., optic cross, mb., mandible. 173}. 506 E. I. WERBER bundles of fibers and to enter the opposite sides of the brain after having formed an indistinct cross (fig. 46). The incompleteness of the fused optic cups and the other abnormalities exhibited by this eye are. probably due to the circumstance that at a very early stage of development a large part of the ophthalmoblastic material had undergone destructive changes and has suffered partial elimination owing to the chemi- eal action of the toxic solution and physical action (increased osmotic pressure) due to subsequent transfer to pure sea-water. The injury sustained by the embryo has apparently been the severest at the most anterior point of the chief body axis, dimin- ishing gradually posteriorly. The following data would seem to substantiate this interpretation. The malformation is restricted mainly to the eyes and the anterior part of the head. For, examination of sections shows that the fore-brain of this embryo is unpaired and otherwise defective, while the rest of the brain is, when followed in sections posteriorwards, seen to gradually present more and more distinctly the condition of bilateral symmetry. The mid-brain and hind-brain while being bilateral, exhibit, however, a certain other abnormality. The injury bere was apparently restricted to the blood and lymph vessels, the earliest anlagen of which seem to have been arrested in their development. This condition can be recognized by the great number of large, clear and empty spaces (figs. 45 and 46) in the tissues of the posterior parts of the brain which in the living embryo have apparently been filled with fluid owing to the existing Imperfection in the circulation. A condition of oedema has thus apparently resulted from lack of drainage, an analogue of which is also represented by the oedematous distension of the cranial cavity in the region of the fore-brain and mid-brain. No other abnormalities of the posterior parts of the brain or any other part of the embryo can be noted, which suggests the con- clusion that the anterior part of the embryo’s body is the most sensitive one and thus subject to the highest degree of injury. c. Cyclopia synophthalmica. The distinction between this de- formity and the preceding one is based on morphological differ- ences existing between the two types of a genetically similar ORIGIN OF MONSTERS 507 malformation. These differences are apparently such of degree only, and probably due to differences in the degree of intensity of action of the same morphogenetic factors. Synophthalmic cyclopia is characterized by the presence in the embryo of a single median eye usually of a larger size than the normal eye, not composite in its appearance in toto but very markedly so on microscopic examination of sections. The fol- lowing two cases may illustrate this deformity. In figure 5 (p. 488) is presented an embryo which has a single, very large, well-formed, median eye, which on examination in toto reveals nothing that would suggest its composite character. On microscopic examination of transverse sections the most anterior sections still present the appearance of a solid non- composite eye, while at about the level at which the lens begins to appear in the sections, the eye-cup discloses its composite character the more the further the sections are followed out posteriorwards, until at the level of the optic lobes this condition of fusion is seen to be very striking. In this region, as well as somewhat anterior and posterior to it, the optic cup presents the appearance of the horizontal section of a funnel (fig. 65). The small end of this funnel is blind and enclosed by the brain, into the substance of which it is seen to dip to a remarkable depth. This funnel-shaped eye is unusually large and the brain is strikingly small. The structures of the optic cup are well differentiated as far as the large part of the ‘funnel’ is concerned, for here the pigment layer and all layers of the retina are present in their typical appearance. In the small end, however, there are only slight traces of the rods-and-cones layer between the outer margins of the ‘funnel’ and the brain, while of the other parts of the retina the fibrous and ganglionic layers are present throughout and are seen to be in continuation with these parts of the large end. The interpretation of the morphogenesis of this eyclopean eye is facilitated by examination of sections of the entire head. The following conditions are revealed by it. The olfactory pits are fused. The fore-brain is unusually short and unpaired, while the mid-brain is bilaterally symmetrical but not distinctly divided 508 BE. I. WERBER into two hemispheres. This is very striking on comparison of figure 64 with figure 65. Both illustrate sections of approxi- mately the same region. While in the former (the case of bilentic synophthalmia described above) the separation of the two hemi- spheres is very distinct, in this cyclopean embryo (fig. 65) there is an apparent rupture of the solid brain mass caused evidently by the expansive growth of the eye which it tightly encloses. This is a break rather than a natural division and it can be fol- lowed throughout the entire mid-brain, while the hind-brain is distinctly bilobed. No other abnormalities were found in this embryo. These data suggest that the ovum at an early stage of its development has sustained an injury at the anterior end of the future embryo’s chief body axis. The injury apparently con- sisted in a destructive elimination of a small, very sharply pointed wedge of tissue (the point directed posteriorwards), comprising the future interocular area, possibly a small part of the opthal- moblastic material, and, evidently, also a considerable part of the future brain. The coalescence of the wound surfaces has caused an approximation and subsequent fusion of parts, which in turn eventually resulted in cyclopia. The other case of synophthalmic cyclopia concerns an embryo (fig. 12) (p. 490) whose head exhibits some striking features. It is relatively very small (microcephalus), the mouth is a wide open, typical proboscis and the eye is single, median in position, very large and betrays on examination in toto no evidence whatever of being composite in character. In anterior sections of this embryo (fig. 66) the appearance of a normal, transversely sectioned, eye is presented, and the synophthalmic character is revealed only by examination of sec- tions at a more posterior level. However, if the brain is exam- ined in anterior sections the nature of the process is disclosed to which the malformation secondarily owes its origin. The forebrain is unpaired and very small, for, in size it hardly exceeds that of the eye. The most remarkable feature, however, which the brain at this level presents, is a fragment of retina which is fused with it at its lower extreme right and just above the orbit ORIGIN OF MONSTERS 509 of the eye. The significance of this retinal fragment is at once recognized when the condition of the whole eye is considered. On following the sections posteriorwards the eye more and more appears to be oval in shape. More posteriorly yet, the eye widens out enormously (fig. 67) while the brain is at this point very distorted and strikingly small in size. A few sections further posteriorly, the shape of the eye is still practically the same, but its size has diminished somewhat while that of the brain has increased. The latter which is now in the region of the optic lobes is very distinctly bilobed, the right hemisphere, although somewhat distorted, is, however, complete, while of the left one about a half is wanting, the place of this lacking part of Fig. 47 Diagrammatic outline reconstruction of the cyclopean eye of the embryo in figure 12. the brain being occupied by the larger part of the eye, which at this level is horse-shoe-shaped in cross-section. The very last sections of the eye prove unmistakably that it is composite, for its base consists of two separate optic cups. The nature of this eye is best understood from the diagram- matic outline reconstruction which is attempted in figure 47. It is a heart-shaped body with the apex directed frontalwards and the base cerebralwards. The question now arises in what way this peculiar malforma- tion came about. The intracerebral retinal fragment which was feferred to above (fig. 66) as well as the shape of the eye and the defects of the brain point to an answer which, I think, contains a high degree of probability. Here, too, the injury sustained by 510 E. I. WERBER the embryo was restricted to its most anterior portion. Owing to this lesion resulting from a process of destruction which I term blastolysis, a wedge of blastema (with the sharp point directed posteriorwards) was eliminated, and the subsequent coalescence of the wound surfaces has caused the earliest optic anlagen to fuse. The eliminated wedge-shaped piece of tissue contained ap- parently the future interocular area, much more of the anterior parts of the ophthalmoblestic material of both sides than of their posterior parts, and a part of the potential brain. This would account for the perfect fusion of the anterior part of the eye components, as well as for the incomplete fusion of their posterior parts, of which probably very little had been lost. That the process which caused the injury was apparently one of dissociation and dispersion (blastolysis) would seem to be evi- denced by the retinal fragment, which can only secondarily have come to fuse with the brain, 1.e., after the ophthalmoblastic frag- ment which has given rise to it had been dislocated cerebralwards from its natural position. d. Cyclopia perfecta. This monstrostity represents a very high degree of ophthalmic malformation. In its morphogenesis it differs from synophthalmic cyclopia, for the perfectly cyclopean eye is genetically a single eye. However, the evidence which I have been able to find, points to the same dynamic process, namely, blastolytic action of the altered environment as the factor responsible for its formation. A few examples may now be presented. In figure 6 (p. 488) is presented an embryo which is possessed of a single median eye and small defective fins, partly obscured from view by the very large oedematous ear vesicles. The whole body is distended and the tail is finless and club-shaped. The examination of the eye discloses a very remarkable con- dition. Already in toto (fig. 6) the eye is seen to be abnormal in shape and in the position of the lens, which is situated postero- laterally, instead of being anteriorly and in the center of the eye. The front of the eye is entirely closed over by the pig- ment layer, so that the condition presented might almost be considered as a rotation of the polar axis of the organ. ORIGIN OF MONSTERS 511 The microscopic examination of the sectioned embryo con- firms the macroscopic appearance. The whole eye appears in forty-nine sections of 6u thickness. Up to the twenty-second section no trace of a lens is seen, and in these anterior sections the condition is presented which normally obtains for posterior sections, i.e. the view is one of the base of an eye where it is entirely encircled by the pigment layer. In the section illus- trated by figure 68 of Plate 1 the retinal layers are defective and the brain, while being bilobed, yet appears to be greatly dis- torted. Following the sections posteriorwards we see in the twenty-second section (fig. 69) the beginning of the lens in its abnormal position. It extends up to the forty-seventh section inclusively and it is thus seen to occupy about the posterior nine- sixteenth of the eye. No optic nerve can be found, although the fibrous layer of the retina is fairly well developed. The mouth is absent in this embryo and the pharynx comes into view in the last sections through the eye, which is seen to be partly enclosed by the defective mandibular arches and projecting for the greater part into the distended pericardial vesicle. The optic cup is C-shaped; it strikingly suggests the simi- larity to a component of a synophthalmic eye and points to the morphogenesis of this monstrosity. Owing to blastolysis the future interocular area of the early embryo, the entire ophthal- moblastie material of one side and a part of it of the other side were destroyed as well as also the earliest anlage of one olfactory pit. The subsequent approximation of the wound surfaces has moved the remainder of the one uninjured ophthalmic anlage out of its original position so that the incomplete optic cup which has developed from it, has turned at an angle of about 90° in relation to its axis. The apparent heterotopia of the lens and the pigment-enclosed front of the eye are evidently due solely to this secondary change in position of the remaining part of one optic anlage. No traces of a fusion could be found any- where, and the presence of a single (non-fused) olfactory pit (fig. 68) would seem to strengthen the evidence that the cyclo- pean eye of this embryo has developed from the ophthalmo- blastic material of one side only. pil? E. I. WERBER The injury sustained by the embryo was a severe one, much severer than is usually found in either synophthalmia or synoph- thalmic cyclopia. This is evidenced by the defective and oedem- atous condition of the brain (fig. 69), and the curving of the head, to which is due the appearance in one section of the eye, medulla, and semicircular canals. Yet the morphogenetic fac- tor which brought about the embryo’s deformity was, no doubt, the same as in synophthalmic monsters. A case of perfect cyclopia, where the malformation is far more extreme than in the preceding one is presented in figure 8 (p. 488). The embryo is seen to be extremely deformed. It has a single, median, unusually small eye, and greatly distended ear vesicles (ef. fig. 71); the entire body is oedematous, all fins are lacking and the tail is club-shaped. On microscopic examination (fig. 70) the eye is seen to be genetically single, there being no indication whatever of its having been formed out of optic anlagen of both sides. It is very rudimentary in structure, the pigment layer and the rela- tively very large lens being its best developed parts, while the retina is very defective, the optic nerve, the iris, anterior chamber and the vitreous body lacking altogether. The brain is lateral to the optic cup of this minute eye instead of being dorsal as should be expected from the position of a cyclopean eye. - This distortion of the relation between the eye and the brain, the defective and unpaired condition of the latter throughout (cf. fig. 70) and the deformities of the rest of the body suggest that a severe injury was sustained by the entire embryo. However, the defects are most extreme at the embryo’s anterior end, which again points to the conclusion that in this case, too, the degree of injury was the highest at the most anterior part of the early embryonic anlage, diminishing posteriorly along the chief body axis. The eye arose from a fragment of one potential optic anlage, the remainder of which and the entire other poten- tial optic anlage as well as a part of the future brain and the potential olfactory pits having suffered destruction. Owing to subsequent processes of regulation the surviving ophthalmo- blastic fragment has come to occupy the median position in the defective head where it developed into the rudimentary eye. ORIGIN OF MONSTERS oo One more case of perfect cyclopia may now be described. On examination of the embryo in toto (fig. 7, p. 488) it was seen that the cyclopean eye is lacking the lens. This may be attrib- uted to the circumstance that the optic vesicle came into con- tact with damaged head ectoderm. On examination of sections (fig. 48) conditions are found which again point to blastolysis (dissociation and dispersion) and subse- quent regulation as the factors responsible for the formation of perfect cyclopia. The retina is defective, its rods-and-cones layer being fairly well developed on one side and less so on the other side of the optic cup. The ganglionic granular and fibrous layers appear to be scattered and intermingled. Several insular accumulations of retinal cells surrounding fibers can be observed resembling the ‘retinal rosettes’ recently described by Nehl (714). The oedema evidenced by numerous large tissue spaces in the eye and in the brain (which latter is highly defective and un- paired), and the distension of the cranial cavity are very likely due secondarily to blastolytic action, the blood vessels, owing to destruction of embryonic material, having failed to develop into a continuous system of drainage. Nothing can be seen in any of the sections that would indicate a fusion of two optic anlagen. Only one optic nerve is seen to pass out of the eye and enter the brain. The complete absence of the olfactory pits and the mouth (the latter coming into view in sections behind the eye) also strengthens the evidence for blastolytic action (dissociation and dispersion) of the environmental modification employed in the experiment. Owing to this action, evidently, the ophthalmo- blastic material of one side has been destroyed, while that of the other side has, through subsequent reparation, come to occupy a median position. e. Monophthalmia asymmetrica. I shall now attempt to show that the same morphogenetic factor (blastolysis) is responsible also for the genesis of other cases of teratophthalmia. They are exemplified by embryos in which both eyes are present in the typical later position in the head, one of them being normal, while the other is small in size and rudimentary in structure, by embryos in which one of the eyes is dislocated (ophthalmic ectopia), and by asymmetrically monophthalmic embryos. 514 E. I. WERBER e The first case now to be described is that of an embryo with one small rudimentary eye, the other eye being apparently nor- mal (fig. 13, p. 490). The embryo is very small as compared with normal embryos of the same age after hatching, its head is curved towards the side of the rudimentary eye, but no other abnormalities could be noticed. Fig. 48 Camera lucida drawing of a transverse section through the eye region of the perfectly cyclopean embryo in figure 7, r.c., rods and cones; r.r., ‘retinal rosettes’; o.n., optic nerve; /., fibrin; e.0e., extracerebral oedema; pl., plasma in pericardium into which the eye dips. 1733 ORIGIN OF MONSTERS SES In microscopic sections the embryo’s right eye appears to be perfectly normal in every respect, while the left eye appears in the sections posterior to it and extending into the yolk-sac, is diminutive in size, but well differentiated in essential structural details. The optic nerve of the small eye is seen entering the brain independently of the same nerve of the other (normal) eye. On examination of the entire series of sections the probable nature of the injury sustained by the embryo is gradually dis- closed. It is restricted mainly to one side and the abnormalities concern the olfactory pit, the eye, the brain, and the ear vesicle. While the olfactory pit of the uninjured side is normal and in its typical position antero-median to the normal eye, there are three olfactory pits on the abnormal side, all of which are closely approximated (fig. 72). This points to a fragmentation of the potential rhino-ectoderm at a pre-differential stage of develop- ment. The condition of the eye of this side has been described; its probable manner of formation is suggested by a careful exami- nation of the brain. The latter in successive sections is found to consist of two hemispheres, but is strikingly asymmetrical in regard to the position occupied by them in relation to the chief body axis, the right hemisphere preceding in all sections the left one which is pushed posteriorwards (fig. 73). This agrees well with the posteriorward dislocation of the eye on the same side and is probably due to regulation after a sustained unilateral lesion. Striking evidence of a process of dissociation (blastoly- sis) is found on examination of sections at the level of the optic lobe of this side (fig. 74). The latter, which, like the whole hemisphere of this side, is posterior to the one of the right side, is seen to be at the same level with the posterior part of the left (smaller) eye, with the right ear, the heart, and the yolk of the body cavity. It is incomplete, and looks as if a part of it had been broken off. To the right from it (in the figure) there is a wide cleft. Between it and the head integument there is a large fragment of tissue (o. |. f.) which has the appearance of 2 small optic cup at an early stage of differentiation, but on careful examination of all sections is found at its most ventral point to be in connection with the optic lobe. This tissue fragment must THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 21; NO. 4 516 E. I. WERBER thus be considered as a part of the optic lobe, from which it was for the greater part delaminated. The ear of the left side is lacking entirely, and it stands to reason that its absence is due to blastolytic elimination of the otoblastic material of this side. Very striking evidence of blastolysis can be observed in the embryo in figure 17 (p. 492). It hatched prematurely on the twenty-eighth day, when it was drawn and kiJled. The yolk- sac is still very large and the embryo is curved. Only the left eye is in the usual position in the head while the right eye is found to be strangely dislocated. It is seen extending between the right mandibular arch and the yolk, in which it is well im- bedded, covered by the yolk-saec. Examination of sections reveals the following conditions. The left eye is well differentiated, but not quite normal. It lacks the anterior chamber and the vitreous body, while the iris is rudimentary. An optic nerve is present and can be followed out to enter the hemisphere of the opposite side. The brain is apparently normal everywhere. In the sections at the level of the medulla and semicircular canals there comes into view the heterotopic eye (fig. 75). It is enclosed between the yolk and the yolk sac and is in proximity to the heart. Between the heart and the dislocated eye very large lymphocytes are seen. The eye is as well differentiated as the orthotopic left eye and is oval in shape, which may perhaps be due to its extra-orbital development. It lacks a pupil and the vitreous body, while the iris also appears to be defective. No optic nerve could be found. An important clue to the genesis of these malformations is furnished by the condition of the olfactory pits. The latter can be followed out very satisfactorily in the most anterior sec- tions. Antero-median from the left (orthotopic) eye there is one olfactoroy pit—in its normal position. On the right side, however, (where the eye has been dislocated) there can be seen jn sections somewhat more posterior, as many as four minute olfactory pits. It seems obvious that these four small olfactory pits have developed from four fragments of the dissociated potential anlage of the olfactory pit of this side. Dissociation ORIGIN OF MONSTERS 517 has, then, on this side affected both the potential olfactory and eye anlagen, fragmenting the one and delaminating and dislo- cating the other. The effect is an increased number of olfac- tory pits and ophthalmic heterotopia. The same morphogenetic factor was apparently responsible also for the conditions found in the embryo in figure 16, p. 492. Its head is unusually large and suggests oedema. The left eye is lacking entirely and a free lens is found in its place. The right eye is defective, the optic cup being C-shaped and the large lens greatly protruding, owing to the absence of an anterior chamber. In the most anterior sections only the ‘independent’ lens is seen which is not yet fully differentiated. In the next sections there comes into view the lens of the right eye and two olfac- tory pits. The latter are observed to be so closely approximated as to be partly contingent on their median borders, and to form an angle of about 90 degrees. More posterior sections show that the optic:cup of the eye is only anteriorly C-shaped, while it is complete, though small, in its posterior part. The part of the section where the (lacking) left eye should be, is occupied by very loose mesenchyme the interstices of which are filled with plasma. The retina is very well differentiated and one optic nerve can be traced to its entrance into the brain hemi- sphere of the opposite side. A very remarkable feature is pre- sented by the sections through the base of the eye (fig. 76). Here, ventral to the brain and lateral from the eye from which it is entirely separated, there can be observed a very small optic cup-fragment with all layers (pigment and retinal-rods and cones) perfectly differentiated. This is evidently a dislocated but fully differentiated remnant of the destroyed ophthalmo- blastic material of the left side. Essentially similar conditions were found in the embryo illus- trated by figure 15 (p. 490). A ventral view of it is presented in which it can be seen that the right eye is lacking and its place is occupied by the mouth. The left fin is about half the normal size, and no other abnormalities were found on examination of the embryo in toto. 518 E. I. WERBER The microscopic examination of sections revealed on the left side (right side in the illustration, fig. 77) a practically normal eye. On the right side in the place of the eye is seen the mouth cavity.. The ophthalmoblastic material which was to form the right eye has been largely destroyed and only asmall fragment of it has been left, which has developed into an optic cup- fragment, seen enclosed in the cranial cavity at the base of the brain. This fragment of the optic cup shows all layers, includ- ing the fibrous layer, of the retina well differentiated. The brain is bilobed and no abnormalities can be found in it excepting its oblique position in the head with régard to the main body axis. This is either secondary to or syngenetic with the herterotopia of the mouth, which in turn is due to the destruc- tion of the ophthalmoblastic material of one side. The elimina- tion of the latter has allowed the mouth to expand in the direc- tion of least resistance so far, as to occupy the exact position of the lacking eye, while the excessive expansion of the mouth on this side may have caused distortion of the brain in its relation to the body axis. Very interesting conditions are found also in the embryo, a dorsal view of which is presented in figure 14 (p. 490). The left eye is lacking and there is a very distinct invagina- tion where the eye should be. On examination of the ventral side the mouth could be seen to be of a shape approaching the ‘proboscis’ type. Its position was very near to what was to be the place of the lacking eye. No other abnormalities could be observed in the embryo in toto. Microscopic examination of sections reveals a normal olfac- tory pit and normal eye on the right side. In sections through the posterior third of the eye (fig. 78) the mouth appears almost exactly in what was to be the place of the eye. An unusually small left olfactory pit comes into view at this level and a minute ‘independent’ lens is noted on the maxilla. Two more minute lenses are found on sections still more posteriorly (fig. 79). The origin of these lenses on the eyeless side I am in- clined to consider as due to contact of remnants of the destroyed ophthalmoblastic material of this side with the ectodermal epithelium. ORIGIN OF MONSTERS 519 The brain is bilobed, but asymetric in regard to its relation to the chief body axis, the hemisphere of the side possessing the eye preceding in sections that of the side lacking the eye. The lesion sustained by the early embryo was evidently re- stricted to the left side of its anteriormost part, where it has eliminated the entire ophthalmoblastic material, minute remnants of which have apparently stimulated the differentiation of free lenses. Owing to subsequent processes of regulation, the in- jured side of the head suffered a posteriorward displacement. This would account for the small size and unusual position of the left olfactory pit as well as for the asymmetry in the posi- tion of the left hemisphere of the brain. f. Microphthalmia and Anophthalmia. Microscopic examina- tion of sections of embryos with very small, rudimentary eyes, or such in which no eyes can be detected in toto have likewise disclosed conditions which point to the action of the same dynamic factor that was found to underly the formation of all other eye terata described above. The following two examples may suffice: On cross sections (fig. 80) of the microphthalmic embryo illus- trated in figure 20 (p. 494) the following view is presented. One eye, while being very small, is seen to be fairly well developed, the retinal layers of the optic cup and the lens being well differ- entiated. The optic nerve of this eye is very clearly seen to enter the optic lobe of the opposite side. The other eye is much more defective. It consists of a rather poorly differentiated optic cup enclosing a lens. One side of the wall of this optic cup is invaginated and partly surrounds another lens. Median- wards from and near to this eye is an undifferentiated mass of apparently ophthalmic tissue with a large well differentiated lens. The embryo then, as we see, possesses virtually three eyes, one of which in structure approaches the norm. Of the other two eyes the first is poorly differentiated and possessed of two lenses, while the second is represented by an undifferentiated optic vesicle with a lens. The ophthalmoblastic material of both sides has suffered lesions, which resulted in the breaking up of one optic anlage 520 E. I. WERBER into two and in the small size of the other eye due to destruc- tion of a part of the potential eye anlage of this side. The injury sustained was a rather severe one for it affected also the brain, the bilaterality of which is obscured (as seen in fig. 80), and the rest of the body. However, it is at the anterior end of the embryo’s body where most damage seems to have resulted from the process of destructive dissociation (blastolysis). The general defects are usually even more extreme in embryos in which on examination in toto only a small rudiment of an eye, like a fragment of the pigment epithelium, is found, or where no eyes at all can be detected (ef. figs. 23 and 24). As a rule, Fig. 49 Camera lucida drawing of a transverse section through the head of the embryo in figure 23. o0.c., optic cup; l., lens; br., brain; e.v., ear vesicle. Som25! it is found on microscopic examination of sections that most anophthalmic embryos possess poorly differentiated and deeply buried eye anlagen, sometimes with a profusion of very small lenses. In figure 49, which is a transverse section through the anterior head region of the embryo in figure 23, two optic vesicles of unequal size can be seen with lenses of corresponding sizes. The simultaneous appearance in the same section of a rudimen- tary ear vesicle, the unequal size and proximity to each other of the rudimentary optic cups as well as the distortion of the brain would seem to well warrant the assumption of blastolysis as the morphogenetic factor responsible for the defects of this embryo. ORIGIN OF MONSTERS 5Dt C. THE MORPHOGENETIC FACTORS UNDERLYING THE ORIGIN OF EYE TWRATA 1. Stockard’s inhibition theory The analysis of the morphogenesis of ophthalmic terata has been attempted repeatedly and with varying success. Good reviews of the opinions of the earlier writers on the subject were already given by Spemann (’04), v. Hippel (’09) and Schwalbe and Josephy (13), and to these the reader may be referred. No distinction is made by most authors between the various degrees of the synophthalmic condition and cyclopia, i.e. the presence of asingle median eye. Thus, their interpretation of the morpho- genesis of cyclopia, applies to the whole so-called ‘eyclencephalic’ group. Two views have been advanced to account for the morpho- genesis of these monstrosities. According to the one represented by Huschke (32), Dareste (91) and very recently advocated by Stockard (09, 710 a, 718) ‘cyclopia’ is a condition in which the separation of what they consider the originally single optic anlage has been inhibited. Opposed to this view is the theory of fusion of two optic vesicles as underlying the formation of ‘eyclopia,’ which was originally advanced by Meckel (’26), and which in a modified form has recently been advocated by most investigators of the subject and particularly by Spemann. This author (Spemann ’03, ’04) has with an entirely different object in view constricted eggs of Triton taeniatus by placing in the two-cell stage a ligature around the first cleavage furrow in relation to which it was somewhat oblique. As a result of this operation he obtained embryos in which the anterior end was doubled to a greater or less degree, depending upon the degree of constriction. In many of the embryos thus treated Spemann observed that one of the doubled heads (or parts of the head) thus resulting was normal while the other which, owing to the oblique ligature was narrower, had defective, usually synophthalmic and sometimes cyclopean eyes. Since the eye deformity was mostly found on the part distal from the embryo’s 522 E. I. WERBER main body axis, Spemann (’03, ’04) concluded that the ‘cyelo- pean’ deformity produced by him was due to a defect because it resulted from the destruction of the area which would nor- mally be the area between the eyes. While not inclined to support unreservedly the hypothesis originally advanced by Meckel (’26), namely that the synophthalmic or cyclopean con- dition results from a secondary fusion of two originally separate optic vesicles, Spemann leans very strongly (’04, pp. 440-441) to the view advanced by Fischel (’03) according to which cyclopia might result from a fusion at a very early stage of development of two originally separate masses of cells which were to form the eyes but had fused before they began to undergo the process of differen- tiation into these organs. With this latter view agree well the observations made by Stockard (09) which I can confirm from my own experience and which were confirmed also by Lewis (09), namely that the synophthalmic and monophthalmic de- formities can be recognized as such already in the stage of the optic vesicle, in other words that a ‘twin optic vesicle’ or ‘cy- clopean optic vesicle,’ if these expressions be permitted, comes off from the brain directly as such. The experiments of W. H. Lewis (09) have also a very impor- tant direct bearing on the subject of morphogenesis of terato- phthalmia. This author employed the method of pricking the anterior end of Fundulus eggs in the embryonic shield stage. Various synophthalmic and one-eyed monsters resulted from these operations, depending on the degree and exact localization of the injury inflicted. From these results Lewis concluded that in the cases of ‘cyclo- pia’ certain cells have been destroyed by pricking, which would normally form the area between the eys. Owing to this elimi- nation of tissue, ‘‘the repair, taking place after the operation, consists of a closing together of the parts left behind : and rudiments are thus brought into contact that normally are quite widely separated, those of the eyes, for example.” This is essentially an assumption of a fusion of early optic anlagen before their differentiation into optic vesicles, as underlying the morphogenesis of cyclopean and synophthalmic monsters. ORIGIN OF MONSTERS 523 Very recently Mall (’08) and Whitehead (’09) have advanced essentially identical views in confirmation of Lewis’ (’09) con- clusions. Thus, as we see, the recent authors practically all agree upon the mode of formation of the ‘cyclopean’ eye. However, the fusion theory of ‘cyclopia’ does not altogether lack opponents, as the most ardent of whom we must now regard Stockard. For a critique of this author’s views the reader must be re- ferred to Spemann’s (12a and b) excellent discussion. While on the basis of evidence in my possession much could be added to the latter, I shall for the present confine myself largely to the discussion of the arguments which Stockard has most recently (13) brought forth in defense of his views on the morphogenesis of ‘eyclopia.’ . Concluding from his experiments with magnesium chloride on Fundulus eggs and also from experiments with alcohol, ether, and chloroform-acetone solutions on the same material and with very similar results, Stockard (’09, ’10a, p. 387) concluded that “the evidence strongly indicates that the ophthalmic abnormal- ities produced in these experiments are the result of an anaes- thetic action during the early developmental stages.” The fallacy of this hypothesis of anaesthetic inhibition has become obvious since the work of McClendon who has recently (12a) shown that ‘cyclopia’ and other malformations of the eye can be produced by various other, non-anaesthetic sub- stances. Thus in his most recent publication, Stockard (13) no longer speaks of ‘anaesthetic’ action,’ but instead he con- siders his experimental eye terata as ‘‘the result of a weakened® development”’ which is brought about by the toxic solutions. These solutions, he argues, ‘‘all tend to suppress or arrest the development of the eye material in the brain.” (718, p. 271). Stockard (713, p. 254) believes that the eye anlage in the medullary plate is primarily median and single and normally separates into two almost equal growth regions, which develop in lateral directions reaching further and further out ®’ My own italics. 524 E. I. WERBER until finally the optic vesicles come into contact with the ectoderm at the sides of the head. On the basis of this hypothesis he now (’13, p. 273) offers the following explanation for the morphogenesis of synophthalmic and one-eyed monsters: the median eye anlage does not widen or spread later- ally but is arrested in its primary condition; thus the growth centers are not sufficiently separated and only a single center exists, and even more than this, the arrest is to such an extent that the entire or normal amount of optic material does not differentiate. Hence one finds a median cyclopean eye consisting of an amount of eye material far below that normally present. This for cyclopia. In the various degrees of synophthalmia he assumes that the “developmental vigor’ is less suppressed, less ‘weakened.’ Here the separation of the single anlage into two ‘growth centers’ is inhibited only to a certain (varying) degree and depending on the variation in the degree of inhibition various synophthalmic conditions, such as the ‘‘eyclopean eye showing distinctly its double composition,” the “hour-glass eye or incom- plete cyclopia,’’ approximation of two separate eyes, etc., result. For the genesis of lateral monophthalmia, finally, Stockard (13) makes the following suggestion: The growth centers representing the two future eyes of an individual are rarely equally vigorous. . . . . It might be that at some critical point in development one of the future eye centers is affected after the growth centers had begun to localize in more or less lateral positions. It is very difficult to understand why (in the same experi- ment!) in some embryos the inhibition of one of the potential eyes should begin after the division of the single anlage (mon- ophthalmia asymmetrica), while in other embryos this single eye anlage should be inhibited before its division into two parts takes place (‘cyclopia’). That the latter is primarily single and median in position in the medullary plate Stockard now regards as a fact, which he thinks he has established by experiments described in his 1913 paper. However, it seems more than probable that the method ORIGIN OF MONSTERS Vinh which he employed in removing fragments of tissue from the antero-median and the antero-lateral portion of the medullary plate has led to errors which Stockard apparently must have overlooked. The unavoidable inaccuracy inherent in the method of mechanically removing minute fragments is obvious. Even with such refined methods as Spemann has employed and with the experience and skill of the latter in such operations, this inaccuracy can not be entirely avoided, as Spemann himself has repeatedly pointed out. It must also be borne in mind that the whole area from which the eyes can presumably arise, is in the stage, at which Stockard performed his removal operations, rela- tively very small. How could any satisfactory degree of pre- cision be attained in operations performed on this small area with ‘fine scissors’ (Stockard’s method)? Is it not probable, that while cutting out the antero-median region of the medullary plate, Stockard has evidently removed also lateral material? Or, that, while attempting to cut away antero-lateral parts of the medullary plate he evidently removed too little, being anxious to avoid inaccuracies resulting from cuts carried too far from a presumably correct position? In the medullary plate, the region under discussion is—at least on superficial examination—mor- phologically homogeneous, and the mapping out of morpho- genetic areas for analytic experiments is beset with well-nigh insurmountable difficulties, even if the instrument used for the operation be the finest conceivable. This can readily be seen from the results which Stockard reports to have obtained from his experiments. Of nine embryos in which ‘narrow strips’ were removed from the antero-median part of the medullary plate, “four . . . . failed entirely to develop eyes.’’ Of the five other embryos only in one the eyes developed ‘“‘to an extent approaching the normal’ and ‘‘four . . . . individuals pos- sessed highly defective eyes” (13, p. 288). Do such results warrant any conclusions at all? Are not the four cases, where defective eyes resulted from the removal of the antero-median tissue, at least as conclusive for the lateral position of two optic anlagen in the medullary plate as the four eyeless embryos for the median position of a single optic anlage? 526 E. I. WERBER Is it not safe to hold that in these experiments the results can not be read aright, owing to the great probability of an error introduced by the inadequate method of the experiment? It would seem obvious that, if experiments be performed on the said area at this stage of development with an instrument as crude as the one employed by Stockard, the deductions from the ‘results are bound to be either erroneous or at least very unsafe in almost every case. For the validity of this claim of the ‘single eye anlage’ and its ‘median position in the medullary plate’ Stockard (13, pp. 274— 276) attempts another proof. He holds that the position of the optic cross outside and below the brain would be inconceivable, if the optic anlagen should be lateral in position and other brain tissue be present between them. For, then, he concludes, the optic stalks instead of having a ‘‘median origin and connection” would be ‘“‘attached to lateral regions of the brain from which the optic vesicles pushed out.’ According to his diagram pre- sented in figure 8 (p. 276) “. . . . in the course of develop- ment the fibers of the optic nerve following the stalk reach the lateral position and must enter the brain and continue within its tissue in order to meet the nerve of the opposite side and form the cross or chiasma. Brain tissue would lie beneath the optic chiasma” and ‘‘this condition is never found in any normal vertebrate.” These deductions would have to be regarded as very important if they were correct. But, as will be seen from the following, they will not hold good. At the outset it should be said that the optic vesicles do not ‘push out.’ Instead they are being pushed out by the pri- moridum of the brain. The parts of the brain anlage (and not of the eye vesicles, as Stockard suggests) most directly concerned in this process of pushing out the anlage for the eye, elongate more and more, until they have attained the form of optic stalks at the time when the differentiating eye vesicles have reached their final positions in the head. Their origin is just as much from tissue dorso-lateral as from ventro-median to the optic vesicle. It is true that this view is, as yet, not based on experi- ORIGIN OF MONSTERS 527 mental evidence, but even lacking this important basis, it would seem far more reasonable and safer and far less arbitrary than the unwarranted claim that the ‘optic anlage’ is median and that the optic stalks are a part of it. Granting, however, that my view is correct—and there is at least a very high degree of probability in it—it is easy to understand that optic stalk tissue may be partly median in origin while the optic vesicles come from antero-lateral regions of the medullary plate. In this case, then, it is also evident that the position of the optic stalks and later of the optic nerves and chiasma ventral to and outside of the brain can not, by any means, be regarded as evidence of the median origin of the eye anlagen. Stockard’s own diagrams (figs. 6 and 7, p. 275)* which are to prove that the eye anlage is primarily median in position would rather seem to support my arguments for the lateral origin of the latter, while his diagram in figure 8 (p. 276) most decidedly portrays a condition which is impossible, not, because the optic stalk tissue is (partly) of median origin, but because it is not a part of the eye vesicle. From what has been said so far, it is obvious that Stockard’s assumption of the single condition of the optic anlage and its median position in the medullary plate as a basis for the morpho- genesis of synophthalmia and cyclopia is untenable. From a study of my own abundant material in sections as well as from a careful scrutiny of the views presented by pre- vious writers on the subject I have convinced myself that a rational analysis of the morphogenesis of synophthalmia and synoph- thalmic cyclopia must be based on Spemann’s (’04, 712) and Lewis’ (09) theory of a fusion of early (pre-vesicular) eye anlagen, due to a defect of intermediate tissue. This conclusion I have reached in spite of the indisputable fact that the fusion theory is in- adequate in the case of perfect cyclopia and that the nature of the defect that precedes the fusion has not yet been made quite clear. In the following I shall attempt an analysis of the morpho- genesis of teratophthalmia which is based on a recent physio- ® Owing to the great importance of the subject here discussed the reader is advised to consult Stockard’s (’13) diagrams, without which this discussion may perhaps not be quite intelligible. 528 E. I. WERBER logical discovery of Child’s as an important key to the under- standing of the nature of chemical defects, and on the assump- tion of fusion of. two originally separate early optic anlagen as underlying the formation of synophthalmic and cyclopean mon- strosities. 2. The defect-theory of teratophthalmia When Fundulus eggs are subjected to the influence of toxic solutions of a perceptibly injurious (but not lethal) concen- tration for a certain (not lethal) length of time, it can usually be found that among the many monstrous embryos which will develop, a certain rather large number will exhibit deformities of the eyes, while other parts of their bodies may appear not, or only very slightly, to deviate from the norm. In such embryos the eye deformities are often the most, if not the only, striking ones. Since the entire eggs were subjected to the influ- ence of the toxic solution, it appears rather puzzling that the developmental product of the egg should show the effect of the treatment only at such a locally restricted area. In experiments in which mechanical methods (such as pricking—Lewis ’09, or constriction—Spemann ’04) are employed the local deformation can readily be accounted for by the locally restricted lesion which has been caused mechanically. In the chemical experi- ment, however, the treatment is not restricted to a part of the egg, and yet the effect is so often a restricted one. How can this be accounted for? There is, so far as I am aware, only one known fact, which will account for this interesting phenomenon. This is Child’s impor- tant discovery of the high susceptibility of the animal pole to noxious influences. This author (Child 711, ’12, 718, 714) has shown that if a ciliate infusorium or a planarian be subjected to the influence of lethal solutions of certain toxic substances the disintegration resulting in the death of the animal will proceed gradually from the anterior towards the posterior end of the body. The same results were obtained also in other adult invertebrates. In this manner a definite gradient of susceptibility was demonstrated to exist along the chief body axis. ORIGIN OF MONSTERS 529 Of quite particular interest in connection with my own results reported. in this paper are Child’s (15 b) experiments on starfish eges. If these were subjected at early stages of development to the influence of solutions of potassium cyanide not strong enough to kill the eggs, the resulting larvae exhibited the detrimental effect of the sojourn in the toxic solution mainly in the apical region.’ Here, as well as in my experiments on Fundulus the part of the egg destroyed by chemical action is the one which corresponds potentially to the embryo’s apical (animal) pole. Very recently Child (15a) has concluded that this primary gradient from the ‘animal pole’ to the ‘vegetative pole’ is very general in organic life and that its demonstraton is only difficult in higher animals, where, owing to complex organization, the results of the ‘resistance method,’ by which they can be demon- strated in invertebrates, are obscured. In the early embryo, however, before the differentiation of or- gans, when the physiological conditions are yet relatively simple, the assumption of such an ‘axial gradient’ in the susceptibility of the fish egg would seem to be justified. If, however, this gradient of metabolic reactions exists in the fish egg, then there exists also a point’ of highest susceptibility and accordingly of least resistance; and this point (the animal pole) is the potential anterior end of the embryo’s body. Grant- ing this, however, it is no longer difficult to understamd why the effect of the toxic solution on Fundulus eggs should, as it so often does, manifest itself in defects of organs of the anterior end of the body, and most particularly the eyes, the mouth, the olfactory pits, and the forebrain, while the rest of the body may suffer very little from the sojourn in the solution. Many observations point to the conclusion that this injury at the apical pole which results in terata of the eyes, is caused mainly by a process of disintegration and dissociation which I have termed blastolysis. It is not easy to understand just what 7 Very recently Painter (’15) observed that in Ascaris eggs which have by an accident come under the influence of carbon dioxide ‘‘roughly 33 per cent of the embryos (in 54 cases out of 165 examined for the point)’’ have sustained severe injuries at the anterior end. 530 E. I. WERBER chemical reactions may underlie this process. They obviously depend entirely on the chemical nature of the solution employed. Thus it would seem reasonable to expect that for mstance they are quite different in butyric acid solutions from those of mag- nesium chloride solutions employed by Stockard (’09) or alkaloid solutions employed by McClendon (12b). The action of some solutions may dissolve, while that of other solutions may coagu- late or precipitate certain substances of the egg. However, no matter what this action may be, it certainly results in a chemical alteration, which will be the more intense, the higher the con- centration of the solution or the longer the time of exposure. Accordingly, if the action be a slight or moderate one, the chemi- cal alteration may result only in an inhibition of certain groups of cells possessing a high degree of susceptibility, i.e. these cells may continue dividing and differentiating up to a certain point, beyond which, owing to exhaustion of their chemical capacity, they are unable to proceed. Or the action may be strong enough to cause, by chemical alteration, a check of the most important physiological processes (cell metabolism and cell division) of these embryonic cells, which would result in their disintegration. No matter what chemical solution be employed, if it only is injurious to life, it will in this way, have a destructive effect. If non-lethal concentrations and lengths of exposure be em- ployed, this destructive process will largely be restricted to the animal pole of the egg, i.e., to that part of the egg which in normal development would correspond to a certain area at the anterior end of the potential embryo’s body. Since, according to the rule of the ‘axial gradient’ this destructive process— chemical blastolysis—begins at the animal pole of the egg and cannot proceed further, owing to insufficient strength of the solution or to timely transfer to a normal environment (pure sea-water), its effect is eventually noted in deformities at the embryo’s anterior end of the body. According to what has been said so far the sequence of events leading to the deformities of the eye is, then, more or less the following. ORIGIN OF MONSTERS O31 A small part of the egg in the earliest stages of development corresponding to a restricted area at the anterior end of the future embryo’s body sustains a chemical lesion, 1.e., it becomes so altered chemically as to be incapable of the reactions necessary for its further normal development. The part of the egg thus incapacitated, is potentially the region anterior to and between the future optic anlagen or even the region of those anlagen. This affected area goes on developing up to a certain point beyond which, owing to the exhaustion of its chemical capacities, it loses its correlation with the whole, i.e., with the rest of the embryo-forming material and becomes eliminated by dissocia- tion. Or, the affected area may, at that critical point of chemi- cal incapacity, be permanently arrested and retain the charac- teristics of this early stage of development (some cases of anophthalmia), while the rest of the embryo may develop and differentiate further. The size of this restricted area of blastolytic lesion at the anterior end of the potential embryo’s body axis is probably subject to considerable variation. Thus it may comprise the mass of cells which would normally correspond to the future interocular area and cause an approximation of the potential optic anlagen or, it may extend over the latter ones and elimi- nate parts of them, while the uninjured parts would fuse after an approximation resulting from the healing of the wound, and form any one of the various degrees of the synophthalmic con- dition. Again, the injured region may comprise parts of the ophthalmoblastic material of both sides and very little of the potential interocular area. The remnants of the optic anlagen may develop and differentiate fully into eyes of strikingly small size of the microphthalmic monsters. Or, the lesion may com- prise the whole of one optic anlage and little or no material of the future interocular area. In that case the embryo will develop into a perfectly cyclopean embryo, if, owing to subsequent regu- lation, the uninjured potential optic anlage is shifted median- wards, or into an asymmetrically monophthalmic monster, if no such change in the position of the uninjured or less injured ophthalmoblastic material takes place. Wherever in such cases THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 21, NO. 4 5a2 E. I. WERBER the sustained injury is of a still higher degree, i.e., if only a fragment of one optic anlage survives, the cyclopean or asym- metrically monophthalmiec eye, formed from it, will be of a corre- spondingly small size. And, finally, if the entire ophthelmo- blastic material is destroyed by the blastolytic process, the defect will result in anophthalmia vera. At this point it should be noted that there is yet another kind of anophthalmia, which is not due to blastolytic destruction but apparently to an inhibition. In anophthalmic embryos of the latter category there are always found on microscopic examina- tion eyes which owing to the exhaustion of their chemomorphic capacity, have remained at a very early stage of development (figs. 23 and 24, p. 494 and fig. 49, p. 520). This, however, is, I believe, the only instance where an ophthalmic deformity is due to inhibition, while all synophthalmic and one-eyed conditions owe their origin to a defect brought about by blastolytic elimination. For a better understanding of the morphogenesis of ophthalmic terata it is also necessary to consider the time at which blastoly- sis brings about the changes leading to their formation. Since it can be shown that in Fundulus eggs under the influence of toxic solutions the chemical injury follows the rule of the ‘axial gradi- ent,’ i.e., it proceeds from the anterior end posteriorwards, it seems safe to assume that the blastolytic lesion is sustained at a very early, primitive stage of development, long before differen- tiation of organs has yet begun. In a previous paper (Werber 715, p. 558) I have advanced the view that the blastolytic injury is sustained before the for- mation of the embryonic shield. ‘To be more precise, 1t may be said that this elimination of destroyed material most likely occurs at a late ‘Randwulst’-stage of the germ-ring. For it is at this stage that very important events take place which eventu- ally lead to the organization of the embryonic body and differen- tiation of tissues and organs. At that time of transformation of the blastoderm into the embryo which is accompanied by rather active movements or shiftings of the embryo-forming material, the contact between the sound, unaltered part of the ORIGIN OF MONSTERS 533 latter and the chemically destroyed one is loosened more and more until it is lost entirely and the altered fragment breaks off. This will at once become apparent if we recall the facts regard- ing the formation of the embryo in teleosts so well established by the investigations of Kopsch (96, 04). According to this author the formation of the embryo comes about in the following manner (’96, p. 120): An dem zelligen Randring (der Keimscheibe 24 Stunden nach der Bildung des ersten Umschlages) muss man zwei Bezirke unterscheiden. Einen embryobildenden und einen nicht (direct) embryobildenden. An dem embryobildenden Bezirk, welcher. an der Stelle der ersten Einstiilpung gelegen ist, haben wir weiter zu unterscheiden einen der Medianlinie naiher gelegenen Teil, dessen Zellen den Kopf des Embryos bilden, und jederseits lateral von diesem Bezirke Zellengruppen, welche im Laufe der Entwickelung in der Medianlinie zusammenkommen, und den Knopf bilden. Der Knopf stellt ein Wachstumcentrum vor, von welchem Rumpf und Schwanz gebildet werden, wobei Zellen des nicht direct zum Aufbau des Embryos verwendeten Teiles des Rand- ringes im Laufe der Umwachsung des Dotters zum Knopf gelangen und dort ebenfalls zur Bildung des Embryos benutzt werden. In both the area which is to form the head (K) and in the parts which are to form the ‘Knopf’ (figs. 50 and 51, p. 534) the organs of the head and trunk are potentially contained. For experiments of Kopsch have shown that if parts of any one of these areas be destroyed the organs which these eliminated fragments potentially represent, will be lacking (Kopsch ’96, pe 121). Fully in accord with these data is the view which I have expressed above, namely that terata of the head owe their origin to blastolytic injury at the anterior part of the area which is to form the head. This area would correspond to the area K in Kopsch’s diagram and the blastolytie elimination of an antero- median chemically incapacitated fragment of this area must be assumed as taking place before the ‘Knopf’ has yet been formed, or thereabout. The eliminated fragment would contain parts of the materiai potentially representing maxillary and mandibu- lar arches, the olfactory pits, the tissues of the interocular area and very often more or less of the ophthalmoblastic material. Fol- ot E. I. WERBER lowing a suggestion made by Schwalbe and Josephy (13, p. 205- 206) I have attempted to portray this elimination diagrammati- cally, employing Kopsch’s diagram as a basis (fig. 50). In the dotted area K, which is the primordium of the head, the two horizontally barred circles represent groups of cells which would normally develop into the mouth and olfactory pits, while the two cross-barred circles immediately below repre- sent the ophthalmoblastic material of both sides. If it now be imagined that blastolysis has eliminated a wedge of tissue which Fig. 50 Diagram of germ-ring with thickened embryo-formating area (‘Keim- randwulst’ )—modified from Kopsch. Fig. 51 Diagram of the early steps in the formation of the embryo from the thickened area of the germ-ring (from Kopsch). as seen in the diagram, is directed with its point posteriorly, while it is broadest anteriorly, this wedge would contain, more or less, the potential interocular area, parts of the potential mouth and the potential olfactory pits of both sides. A con- siderable part of the potential forebrain would in this way also be eliminated. The gap thus resulting will soon close, owing to a contraction brought about by the elongation which accom- panies the transformation of the head primordium of this stage into the head. What would be the final result of the elimination of such a wedge of tissue and the subsequent closure of the wound? 100. come into view and also a part of an obliquely cut gut-coil. A few sections more posteriorly the plasma-filled space between the two parts of the embryo has disappeared from view and its place is now taken up largely by the obliquely sectioned intes- tine which stretches across between the two parts. The latter are now continuous. The spinal cord of the tail-part is now restricted to the region dorsal to the notochord and can be ORIGIN OF MONSTERS 555 observed to gradually assume a shape approaching the normal. At this level one of the ear vesicles has disappeared from view while that of the other side (a very large one) is very prominent in the section. Laterally and ventrally from this ear vesicle (fig. 62) is seen a rudimentary optic cup. In the last sections the spinal cord runs across between the two parts of the embryo. It is particularly in these sections that the true nature of the monster is disclosed. From them we see that the posterior half of the embryo’s body is in relation to the embryo’s longitudinal axis turned at an angle of almost 180°. An oedematous blasto- dermic cavity filled with plasma and extending between the two so disarranged parts of the embryo has given the latter the ap- pearance in toto of coalesced very malformed twins. What is, now, the probable morphogenesis of this monster? The presence of one eye only would, according to what has been stated in the preceding, point to blastolytic injury sustained by the early embryonic primordium. This injury seems to be evi- denced by the following facts. The ophthalmoblastic material of one side has, owing to chemical alteration, been somewhat inhibited in development and has given rise to an imperfect eye. The ophthalmoblastic cells of the other side have suffered great destruction (owing to increased osmotic pressure after chemical alteration?) and a remnant of them has been fragmented into two parts. Of the latter one has made some initial steps in development (having attained the shape and to some degree the structure of an optic cup) and stimulated the development of a lens, while the other fragment has been dislocated and developed into a rudimentary optic cup at the posterior level of the ear vesicles. The latter are both deformed and oede- matous. The deformities of the brain are such as to point to a hap-hazard regulation of a brain primordium after blastolytic lesion. The oedematous blastodermic cavity filled with plasma suggests that no continuous blood circulation existed, the failure of blood vessels to develop being due to chemical alteration— a condition which can be found nearly always in embryos which have sustained blastolytic injury of a high degree. The strange space relation of the posterior half of the body to the anterior would apparently indicate the action of osmotic pressure, which 556 E. I. WERBER while not sufficiently great to cause a rupture of the two halves of the primordium, has only shifted one of them out of its normal position on the yolk. Briefly, it may be said, that the effect of blastolysis, due to chemical alteration combined with osmotic pressure, was in this case a highly deformed amorphous embryo, which as figure 44 shows, can in toto easily be mistaken. for coalesced malformed twins. To blastolysis by chemical alteration and increased osmotic pressure are evidently due also the conditions found in the following embryo now to be described. As can be seen from figure 30 the embryo is very deformed and only slightly suggests the body form of a fish. The irregu- larity of form pertains particularly to the head and the trunk, which latter appears to be greatly distended. There is no clear indication of the true nature of the embryo, namely that it is, as examination of sections shows, acase of a partial, fused duplicity. ; The most anterior (transverse) sections show an anophthalmic and sharply pointed, highly oedematous head. Between the head and the yolk-sac there is a space filled with plasma, (fig. 85) which can be recognized as the enlarged, oedematous pericardial vesicle. A lens is seen situated laterally, but there is no trace of an optic cup. Practically at the same level with the lens there appears ventro-lateraJly to it a transversely sectioned tube- like structure (gut?) which can be followed only in eight sec- tions of 7u thickness. Following the sections posteriorwards the oedematous brain is seen to increase in size and to be highly deformed. Still more posteriorly, i.e. at the posterior level of the lens the brain mass makes the impression of two fused, malformed brains, there being between the components a distinct lamella. In sections through this double brain mass there can be seen at a distance from the embryo a transversely cut fragment of nervous tissue which makes the impression of a deformed spinal cord. It corresponds to the tissue fragment (¢f.,) in figure 30. Between this fragment, the embryo and the yolk, there is a cavity filled with a plasma-like mass (fig. 86). ORIGIN OF MONSTERS 557 A second tissue fragment (tf. of figure 30) can be seen in more posterior sections on the side of the yolk opposite the embryo. This fragment proves on careful examination to be a very rudi- mentary eye, of which only the lens can be recognized with certainty. The larger one of the two brain components (a) is more oede- matous and its staining reaction with haematoxylin is weaker than that of the other component (b). This fusion of two brains into one mass becomes very distinct in still more posterior sec- tions where component a now gradually diminishes in size, until it entirely disappears from view while the component 6 has encroached upon its place in the section. Several sections further posteriorly (fig. 86) there is seen a large optic cup at the ventral part of the brain component b. Retinal cells (rods and cones) of this optic cup can be recognized in some sections. The following is the complete picture presented at this level. Almost half of the section through the embryo is taken up by brain component 6. Laterally from it is seen a transversely cut notochord, while ventral to it is the optic cup, and ventrally from the latter are seen cross-sectioned coils of an intestine. In the intestines a fibrin-like mass can be noticed, which points to defects in the circulatory system. Following the series more posteriorly there can be seen gradually to appear the spinal cord above the notochord. It is at first on one side fused with brain component 6b while in sections at a more posterior level (fig. 87) it is separated from the latter by a mass of mesodermal cells (poorly differentiated myotomes?). In still further sections the brain component b disappears entirely while a large cavity lined with endothelial cells and filled with plasma is seen to follow it in the sections. The last sections are cross sections of a deformed tail. At the outset of the description of this monster it has been stated that the conditions found on microscopic examination point to blastolysis as a factor underlying its formation. While of course, a complete analysis of the morphogenesis of this monster is practically impossible, the two extra-embryonic tissue fragments would seem to point decidedly to blastolytic action. Considerable difficulty, however, is presented by the interpre- 558 E. I. WERBER tation of the double, fused brain. Here an additional assump- tion may be made, namely that fragments of the blastolyzed embryonic primordium have secondarily come to fuse owing to shifting caused by increased osmotic pressure. Such spatial dis- tortions and shiftings can be recognized in many duplicities. In our present case this hap-hazard regulation and fusion of parts would seem to be evidenced by the unusual position of an optic cup on the ventral side of the posterior part of the brain (ef. fig. 86). That amorphous monsters result from blastolytic injury of a high degree is well illustrated also by the anophthalmic and greatly malformed embryo in figure 31. Besides the defects and malformations which are seen on examination of the embryo in toto more are disclosed when sec- tions are examined. There is no trace of eyes, and the nervous system is lacking almost entirely. In transverse sections (fig. 88) through the region of the head only some muscles, connective tissue, cartilage (of maxillary or mandibular arches?) and very rudimentary, closely approximated ear vesicles can be seen. In a few, more posterior sections there can be seen what might be taken for a very faint suggestion of nervous tissue in a very poorly differentiated stage. It is evidently this exceedingly small remnant (only seven sections of 7y thickness) of the nervous system that has given rise to sensory part of the ves- tigial ear vesicles. Somewhat further posteriorly transversely sectioned coils of the intestine come into view. No trace can be found of either the notochord or the spinal cord. The fins, which also come into view at this level, are oedematous and filled with plasma (fig. 89). A few sections further posteriorly there is seen ventrally from the intestine a large coelomic cavity containing some fibrin. This cavity enclosed by the body wall is practically all that can be seen in the last sections of the monster. IV. CONCLUDING REMARKS It has been shown in the foregoing that an almost endless variety of monstrosities can be produced in fish, if eggs in the very earliest stages of development are subjected to the action ORIGIN OF MONSTERS 559 of some toxic products of pathologic metabolism. The mon- strosities thus produced resemble very much those occurring spontaneously in man and in other mammals. In other words, effects of unknown processes occurring in nature have been practi- cally duplicated on a very largescale by the laboratory experiment. In the latter the initial cause is known, because controlled by the experimenter. Some further steps in the sequence of events can, as I have recently convinced myself, partly be observed, and partly deduced from anatomical examinations of great num- bers of monsters. The results of the latter agree well with those of anatomical examinations of the spontaneous monsters of mammals. Provided that in nature the same or very similar causes are responsible for the origin of monsters, the whole problem of teratogenesis in nature is—theoretically at least— within the control of the experimenter. On the other hand, the experimental results recorded by pre- vious work in embryology and teratology all seem to point to the conclusion that of teratogenic factors thereare many, indeed almost as many probably as there may be agents injurious to organic existence. On what basis then, may it properly be asked, can experiments be performed with the aim of control of the causes, if the latter ones can only be a few out of a very great number of possible ones? The problem, while obviously beset with many difficulties, is, however, not quite as elusive as it would at first appear. For, while there is probably an endless number of factors which, if acting on the egg, might cause it to develop in an atypical manner, it is, for our purposes, necessary to consider only those which can reasonably be conceived as acting in the mammalian body under certain unusual conditions. The latter are, as we shall presently see, relatively few. These factors must be either of a physical or of a chemical nature. Of the physical factors which might interfere with the typical development in utero of mammals only pressure and an increase in temperature could be imagined. The former has, indeed, for a long time been regarded as the cause underlying the origin of monsters. The inadequacy of this mechanical theory of terato- 560 E. I. WERBER genesis has been repeatedly pointed out, and notably by Mall — (l.c.). Neither is it probable that such an increase in the tem- perature of a mammalian female as would yet allow it to survive’ could materially affect an ovum in _ utero. If, after what has been said, we dismiss the possibility of physical factors as the underlying causes of atypical develop- ment, there would remain only the group of chemical factors. Of the latter the number of those which can reasonably be imagined to be present in the mammalian’s body under certain conditions, is rather limited. They might be some extraneous poisonous substances which may chance to gain entrance into the human (or other mammalian) body, they could also be bac- terial toxines, or, finally, they may be autogenous poisons of the body, such as some toxic products of a disturbed metabolism. Of the extraneous poisons which may find a way into the body without killing the individual, but only exerting an injurious in- fluence on it, or on a developing ovum contained in the latter’s uterus, there are probably not many. Alcohol and someother drugs, to which individuals may habitually be addicted, would probably occupy an important place among these substances. Lead-poisoning and phosphorus-poisoning might perhaps also be considered under this heading. However, while in the light of recent data, there is hardly any doubt left regarding the deleterious effect of parental alcoholism on the offspring (ef. Stockard 712) no cases of human monsters are yet known of an alcoholic parentage. The same would apply also to the occu- pational diseases of lead-poisoning and phosphor-poisoning. It cannot be denied that these poisons probably are capable of alter- ing the typical course of embryonic development, but, evidently, they seem to be acting in a degree not sufficient to cause the development to become monstrous. Or, possibly the strength of action of these poisons necessary to produce monsters might kill the parent or bring about its sterility, and thus, no occasion may exist for human ova to be under the influence of such strong action of these poisons as would very materially alter the course of their development. The conspicuous lack of data regarding the rela- tion between alcoholism and ‘industrial diseases’ on the one ORIGIN OF MONSTERS 561 hand and the occurrence of monsters in man on the other hand would seem to point to the correctness of our conctusions. Besides, many monsters similar to human monsters are found in other mammals (and in sauropsids) which are neither ex- ‘ fe is posed to the dangers of alcoholism nor to those of ‘industrial ‘diseases.’ These considerations have led me to believe that poisons such as alcohol and other drugs as well as poisons of the ‘occupational diseases’ only very rarely, if ever, may lead to monstrous development in nature. The percentage of monsters whose origin may be due to these causes must, at best, be re- garded as so small, as to be practically negligible. Let us now turn to the next known chemical modification of the mammalian body, namely to the toxines of infectious diseases. Is there any likelihood that such toxines, when present in the blood of a female, would have such a deleterious effect on the ovum in its uterus, as to seriously derange the course of its development? This question is as yet difficult to answer, for the existing data on the subject are too insufficient and too indefinite. But even if future researches should prove the probability of a deleterious effect of maternal infectious diseases on the embryo in utero, which at present seems doubtful, there would remain only one other great source of chemical alteration of the ovum’s environment, namely the products of pathologic metabolism. Of the latter there is a considerable number known at the present time, and a better knowledge of the pathological chemistry of disturbances of, particularly, proteid metabolism may add to this number. Already Forster (65) has advanced the hypothetical view that chemical alteration of the maternal blood may be one of the causes underlying the origin of monsters. This view although —strangely enough—almost entirely overlooked is now, however, not insupportable. For, recent advances in pathological chem- istry have made us familiar with the chemical changes which our organism is subject to under certain pathological conditions. At the same time a better insight has been gained into the close physiological relation existing between the developing embryo and the mother. Thus, in the light of our present knowledge 562 E. I. WERBER we must admit that Forster’s idea although at the time ex- pressed in a rather vague manner, is characterized by great foresight. One gains this impression particularly from Wolff’s (13) excellent summary on the biological relationship between embryo and mother during pregnancy. This relation is, according to Wolff, particularly intimate in man, for the nutrition of the embryo is directly from the mother’s blood. It is also of especial interest in such cases where, owing to pathological conditions, some toxic products of metabolism ac- cumulate in the mother’s blood. That the embryo is influenced by such alterations of maternal blood, is, according to this author, very plainly evidenced by clinical as well as experimental pa- thology. Thus it has, for instance, been found that the fetus of a nephritic mother may suffer from oedema and ascites (Sitzen- frey 710). The same obtains also in the animal experiment, if pregnancy is induced in a nephrotomized female. As a very interesting illustration of the intimacy of physio- logical relation between fetus and mother in man may be con- sidered the observations recorded by Kehrer.’ According to this author bile acids and bile pigments can in the animal experi- ment not cross the placental barrier and thus they never reach the embryo, while in man the observation can be made that children of women suffering from jaundice during pregnancy may have jaundice at birth. Such data would seem to leave no doubt that themammatlian embryo is sometimes subject to influences of a disturbed maternal metabolism. Whether all products of a deranged metabolism will have a deleterious effect on the embryo is a matter which, of course, needs investigation. In the preceding I have shown that if two substances known to occur in disturbances or carbo- hydrate metabolism, viz. butyric acid and acetone, be allowed to act on fertilized teleost eggs, the latter will develop into various monsters very strikingly resembling the monsters of man and other mammals. Even teratomata have thus been produced. ‘These results would seem to leave little room for doubt that mammalian monsters may often be due to a coinci- ® Quoted from Wolff (14). ORIGIN OF MONSTERS 563 dence of pregnancy with at least such disturbances of metabolism as diabetes. This coincidence has been found to be fatal to the embryo. Thus, according to Seitz (13) diabetes melitus has for a long time been considered as causing sterility in women. It is quite possible, I think, that in such individuals the ovum, due to toxic changes, might be expelled soon after conception, that such early abortions might have been mistaken for menstruation and that while perhaps no sterility existed a similar effect has been mistaken for it. For, to quote Seitz: erst als Duncan iiber mehrere Falle von Graviditit bei Zuckerkrankheit berichtete, dnderten sich die Ansichten und nach den vorliegenden Statistiken iiber insgesamt 427 Frauen in eeschlechts- fahigen Alter darf man annehmen, dass bei rund 5 procent der dia- betischen Frauen Schwangerschaft eintritt. The influence of diabetes on the development of the offspring is according to Seitz very disastrous: Besonders ungiinstig ist die Riickwirkung des Diabetes auf des Kind. Ubereinstimmend berichten die Statistiken, dass rund 50 procent aller Kinder intrauterin zugrunde gehen und zwar kann der Tod des Kindes jederzeit eintreten, sowohl in den ersten Monaten der Schwanger- schaft als auch noch gegen Ende. .. . Ein weiterer Teil der Kinder wird friihzeitig und schwach geboren; wiederholt wurde auch Erkrankung an Polyurie, an kongenitalem Diabetes und an Hydro- cephalus beobachtet. The data quoted above are, obviously, insufficient and, com- ing, as they do, from clinicians, they leave us altogether in the dark regarding the morphological effects of maternal diabetes on the embryo. The desirability of comprehensive data regard- ing fetuses aborted by diabetic mothers as well as those suffering from other diseases of metabolism can not be urged enough upon physicians, as they may not infrequently chance to observe these experiments of nature. But even more desirable would seem to be experiments on mammals. Here the various patho- logical conditions of metabolism must be imitated by experiment as closely as possible, and the animals so diseased mated in THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 21, NO. 4 564 E. I. WERBER various combinations. The difficulties of such investigations are numerous, even if the most important one did not exist, namely that the experimental pathology of metabolism is prac- tically yet in its infancy. But such difficulties may perhaps in time be at least partially overcome. And while the complete solution of the problem of the causal factors underlying the origin of monsters may yet be distant, well planned experiments and careful analysis of results may at least furnish direct evidence for the correctness of our assumption that monstrous develop- ment is primarly due to parental metabolic toxaemia, an hy- pothesis which in view of the noted results of our experiments would seem to be well justified. Our experiments have, besides, pointed out the direction for investigations into the morphogenesis of monsters. In the pre- ceding pages I have shown that microscopic analysis of all monsters discloses evidence of an action on the egg treated with butyric acid and acetone, which tends to dissociate (or disrupt) the germ’s substance. This action, blastolysis, is a complex com- ponent in the sense of Roux (’95), being the result of the col- lective action of a number of factors potent in a varying degree in all eggs. Some observations made in experiments during the summer of 1915 seem to suggest that the most important direct factors whose action results in blastolysis are the toxic effect of the chemical modification of the environment and an increase in osmotic pressure mainly after transfer of the eggs to pure sea- water. The first materially alters or destroys parts of the germ, while the latter may be regarded as a very forceful dissociating agent which disrupts the injured germ. The alteration brought about by the action of butyric acid or acetone seems to be due to the solvent action of these acids. To these conclusions would seem to point the fact that some hours after transfer of the eggs from butyric acid or acetone into pure sea-water there could in- variably be found a sediment at the bottom of the dish in which they were kept. The sediment was of a slimy consistency in the dishes con- taining eggs which had been treated with butyric acid, while ORIGIN OF MONSTERS 565 in the dishes into which acetone-treated eggs had been trans- ferred the sediment made the impression of a granular precipi- tate. Chemical analysis of these sediments may eventually dis- close the nature of the chemical alteration of the eggs treated with these organic acids. One rather evident effect which they have on the egg is that they increase its permeability. For on transfer from the solu- tion to pure sea-water the eggs, owing to increased imbibition of sea-water, swell to an unusual size; after some time, however, they return to approximately the normal size and occasionally some eggs may be found considerably below the size of the normal, untreated egg. This indicates that the increase in per- meability has called forth an increase in endosmotic pressure (imbibition) which allowed enough sea-water to enter the egg to dilute the substances dissolved by the acids, and a subsequent increase in (ex)osmotic pressure (shrinkage) owing to which the dissolved substances have passed out of the eggs and formed the sediment on the bottom. . Besides these two factors (toxicity of the medium and osmotic pressure) which are directly concerned with blastolysis there seem to be some other factors, to which apparently is due the enormous variation in the blastolytic effect and consequently the variation in the morphological deviations from the typical development. While some observations have been made which may bear on the nature of these factors, I can make no definite suggestion at this time and the determination of these factors must be deferred to future experiments. However, while the analysis of blastolytic action into its com- ponent factors must thus for the present time remain incom- plete, the evidence of this action is indeed very striking in practically all deformed embryos of my experiments. Considering the fact that all these terata have been produced by experimentally induced blastolysis from eggs which would in a normal environment have given rise to normal embryos, it _ would no longer seem to be necessary to assume ‘germinal varia- tion’ as the cause underlying the origin of ophthalmic monsters and various duplicate twins, as this has been postulated by 566 E. I. WERBER some authors and notably by v. Hippel (’09) and H. H. Wilder (08). The latter author, led by the ‘“‘symmetry and regularity in anatomical details’? of some monsters (cyclopia, diplopagus) has come to look upon them as “beings as orderly and perfect in their development as are the usual and normal types of being.” In contradistinction to deformed embryos (true monsters) he even proposes the term ‘“‘Cosmobion (plural cosmobia)”’ to desig- nate such ‘regular’ symmetrical monsters, the underlying primary cause of which he assumes to be germinal variation. Quite apart from the circumstance that the cause of ‘germinal varia- tion’ would still have to be explained, its assumption would seem unnecessary, because it can be demonstrated that of any given batch of eggs of approximately the same (early) stage of develop- ment those left in a normal environment will develop into beings typical for their species, while those subjected to the influence of agents which induce blastolysis will develop into monsters of which some may come very near to Wilder’s own standard of ‘cosmobia.’!” In our experiments such monsters have developed under the blastolyzing influence of some products of pathologic metabo- lism which might be imagined to be acting on the mammalian ovum during its uterine as well as pre-uterine existence: In the latter case we would be presented with what might, in a restricted sense of the word, be called ‘germinal variation.’ This germinal deviation, however, being due to a pathological cause, it would seem unwarranted to regard the developmental products of such ova, no matter how symmetrical and well formed they might be, as ‘cosmobia’ (‘orderly living beings’). Besides, it is not en- tirely improbable that. true ‘cosmobia’ might be produced, if it only were possible to imitate exactly the environmental modifi- cations which underlie the origin of monsters in nature. This degree of accuracy is, however, not yet attained in our experiments. 10 One cannot but admire the refinement of Wilder’s morphological specula- tions. But, with all due respect for this accomplished morphological philosopher, it is difficult to accept his theory of cosmobia. For, being, as it is, of necessity based on the assumption of germinal variation (of an apathological nature) it leads to the pessimistic conclusion that the problem of the origin of monsters (or at least of ‘symmetrical monsters’—‘cosmobia’) is beyond control. ORIGIN OF MONSTERS 567 The assumption of the action of parental metabolic toxaemia on the (uterine or pre-uterine) ovum might also reasonably be extended to the male germ cell, which in fertilizing a normal ovum might cause it to develop into a monster. In this case the chemically altered spermatozoon would probably act in a manner very similar to the germ cell of another species. For, as it will be remembered, Moenkhaus (’04) has demonstrated that if eggs of Fundulus heteroclitus be fertilized with the sperm of Menidia various deformities will result. This experiment has since been repeated by Loeb (’15) and myself" and most recently by Reagan and Thorington (’15). Foreign sperm has evidently a toxic effect on the egg and thus deranges its development from the typical course. In analogy, the same may be true for the chemically altered spermatozoon of the same species. That thus modified spermatozoa, if fertilizing normal ova, may cause them to develop into degenerate individuals appears to be probable from statistical data on children of male alcoholics (Stockard ’12). Very recently Stockard (14) has demonstrated that the progeny of experimentally alcoholized guinea pigs, which had been mated to a normal female, was degenerate and deformed, thus sug- gesting that the chemical injury sustained by the chromatic substance of the spermatozoa had a deleterious effect on the normal ova of healthy females. Since male individuals are just as subject to disturbances of metabolism (although possibly less frequently) it is not improbable that the sperm of males suffer- ing from metabolic toxaemia may bring about abnormal develop- ment when fertilizing normal ova. The effects of this union of normal chromatin with chemically altered chromatin may even be unnoticeable in the first generation of offspring and appear in the second and further consecutive generations as in the above mentioned breeding experiments of Stockard with guinea pigs. This author even reports that the results of mating _ descendants from an alcoholized father and a healthy mother deteriorate markedly with each consecutive generation. These data are very important for our considerations. For, if toxic 11 Not published. 568 E. I. WERBER products of pathologic metabolism should have an injurious influence on the male sex-cells—and there is every reason to believe that they do have such an effect—the chance for a terato- genic influence of parental metabolic toxaemia on the offspring is greatly enhanced. Elsewhere I (Werber 715 b) have pointed out the bearing which these conclusions and the results of my experiments with products of metabolic toxaemia on the teleost ovum may have on some, so far elusive, problems of medicine. Not less evident is their possible significance for eugenics and the biology of the race. From the point of view, however, of the embryologist and pathologist our hypothesis and the results of the present experi- mental study based on it would seem to offer a rational basis for the solution of the old problems of the etiology and the morphogenesis of terata occurring spontaneously in many ani- mals and notably in man and other mammals. These problems may, I think, now be, at least partly accessi- ble to experimental control. V. SUMMARY 1. Recent results of investigations in experimental ernbryology and teratology pointed to the conclusion that the primary causes underlying the origin of monsters in man and other mammals are of a chemical nature. They also suggested that in the latter, particularly, these striking deviations from the develop- mental norm are due to autogenous chemical modification of the parental blood during disturbances of metabolism. 2. On the basis of this hypothesis experiments were performed on fertilized eggs of Fundulus heteroclitus which were subjected to the action of some substances of certain metabolic toxaemias. 3. Positive results were obtained with particularly two sub- stances, which occur in toxaemia due to disturbances of carbo- . hydrate metabolism, namely butyric acid and acetone. 4. A very great variety of monsters has resulted from these experiments, analogous to human and other mammalian mon- sters. The deformities concern the eyes (cyclopia, synophthal- ORIGIN OF MONSTERS 569 mia, monophthalmia asymmetrica, and anophthalmia), the ear vesicles (rudimentary structure, or lacking, or presence of one vesicle only, or synotia), the olfactory pits, the mouth, the cen- tral nervous system, the heart and blood vessels, the fins (un- paired: pectoral fins, absence of pectoral fins or all fins, club- tail, etc.), and body form. 5. Oedematous conditions were found in many embryos lack- ing a continuous system of blood circulation in various parts of the body, which were greatly distended and contained plasma or fibrin and, not infrequently, many lymphocytes. This con- dition of hydrops is most frequently found in the head. It may be intracerebral or extracerebral and suggests an analogy with the congenital internal and external hydrocephalus of man, which latter may perhaps also be due to developmental imper- fections of the blood-vascular system. 6. Blastolytic action of the chemically modified environment is assumed as a morphogenetic principle common to all terata of these experiments. Blastolysis either destroys part or all of the germ’s substance, or it may split off and disperse parts of the latter. 7. The nature of blastolysis is two-fold, namely chemical and osmotic. a. Chemical blastolysis 1s a process of chemical alteration (by solvent or precipitating or coagulating action) of the germ’s sub- stance. This alteration results in dissociation or disintegration of parts of the latter (defect), or, occasionally in a decrease of the germ’s chemical capacity for development and differentiation (inhibition). b. Osmotic blastolysis sets in while the eggs are under the in- fluence of the toxic solutions employed and again (more so) on their transfer from these solutions to pure sea-water. It results from the increase of permeability which allows sea-water to enter the eggs. The wmbibition of sea-water by the eggs which swell rapidly, calls forth a fragmentation of the germ and dispersion of its parts which at this stage are yet capable of further independent develop- ment and differentiation. 8. All eye terata (cyclopia, synophthalmia, monophthalmia later- alis, anophthalmia) are due to a defect, viz., to blastolytic elimination 570 E. I. WERBER of a fragment of, either, ophthalmoblastic or potential interocular material and not to an inhibition as was held by Huschke and Dareste and is now postulated by Stockard. Only such cases of anophthalmia, where on microscopic exami- nation rudiments (ill-differentiated optic vesicles or cups) are found, form an exception to this rule. Here an inhibition is assumed, due to a decrease of the chemical capacity for develop- ment (chemical exhaustion). 9. The frequent occurrence in these experiments of terata of the eyes (or the anterior part of the head) only is regarded as being due to the highest degree of susceptibility of that part of the earliest embryonic primordium, which eventually becomes the embryo’s anterior end (animal pole). This assumption sup- ported by many data, is based on Child’s discovery of a definite susceptibility gradient (‘metabolic gradient’-Child) along the chief body axis of many animals from various phyla. 10. The numerous meroplasts recorded and especially such teratomata as the ‘solitary eye’ and the ‘isolated eye’ point to a very high degree of capability of parts of the embryonic pri- mordium for independent development and differentiation (self- differentiation—Roux). 11. While the occurrence in these experiments of various duplicities would seem to point to a relatively high prospective potency of parts of the teleost egg, the latter decreases very rapidly, for at an early stage (the Randwulst-stage or thereabout) these parts are already specifically predetermined as early (un- differentiated) anlagen of some organs. If at this stage a frag- ment be eliminated the result will be a defect of a corresponding organ or part of the body. Accordingly the teratogenetic time limit must be regarded as very brief, and especially so in the case of duplicities. 12. The results obtained tend to justify the hypothesis on which the experiments were based, namely that parental meta- bolic toxaemia may be the cause, or, at least, the chief cause underlying the origin of monsters. DECEMBER 24, 1915 ORIGIN OF MONSTERS 5A LITERATURE CITED AHLFELD 1880-82 Die Missbildungen des Menschen. Bock 1889 Beschreibung eines atypischen Cyclops. Klin. Monatsblitter f. Augenheilkunde, vol. 27. (Quoted from Schwalbe and Josephy.) Cuttp, C. N. 1911 Studies on the dynamics of morphogenesis and inheritance in experimental reproduction. I. The axial gradient in Planaria Doro- tocephala as a limiting factor in regulation. Jour. Exp. Zoél., vol. 10. 1912 Studies on the dynamics of morphogenesis ete. III. Jour. Exp. Zool. vol. 11. 1913 a Studies on the dynamics of morphogenesis, etc. V. Jour. Exp. Zool., vol. 14. 1913 b Studies on the dynamics of morphogenesis, ete. VI. Arch. Entwmech. Vol. 37. 1914 The axial gradient in ciliate infusoria. Biol. Bulletin, vol. 26. 1915 a A dynamic conception of the organic individual. Proc. Nat. Acad. Sciences, vol. 1, p. 164-172. 1915 b Axial gradients in the early development of the starfish. American Jour. Physiol., vol. 37. DarestE, C. 1891 Recherches sur la production artificielle des Monstruosites ou Essais des Teratologie experimentale. 2d edition. Ernst, P. 1909 Missbildungen des Nervensystems. In Schwalbe’s Die Mor- phologie der Missbildungen des Menschen und der Tiere. III. Teil. Fiscnen, A. 1903 Uber den gegenwiirtigen Stand der experimentellen Teratolo- gie. Verhandlungen der deutschen Pathologischen Gesellschaft. V. Forster, A. 1865 Die Missbildungen des Menschen. Genwi, J. F. 1912 Teratology of fishes. Glasgow. Harrison, R.G. 1907 Observations on the living nerve fiber. Proceed. Society Exp. Biol. and Medicine, vol. 4. 1910 The outgrowth of the nerve fiber as a mode of protoplasmic movement. Jour. Exp. Zool., vol. 9. Hippey, v., E. 1909 Die Missbildungen des Auges. In Schwalbe’s Die Miss- bildungen des Menschen und der Tiere, III. Teil. Huscuxe, EK. 1832 Uber die erste Entwicklung des Auges und die damit zusam- menhingende Cyclopie. Archiv. f. Anat. u. Physiol., vol. 6. JorpaNn, H. EK. 1909 A study of pathological cat embryos. Anat. Rec., vol. 3. Koprscu, Fr. 1896 Experimentelle Untersuchungen iiber den Keimhautraud der Salmoniden. Verhandl. Anat. Gesellsch. zu Berlin 1896. Anatom. Anz., vol. 12: Kunprat, H. 1882 Arrhenocephalie als typische Art von Missbildung. (Quoted from Schwalbe and Josephy). Lewis, W.H. 1909 The experimental production of cyclopia in the fish embryo (Fundulus heteroclitus). Anat. Ree., vol. 3. Logs, Jacques 1915 The blindness of the cave fauna and the artificial produc- tion of blind fish embryos by heterogeneous hybridization and by low temperatures. Biol. Bulletin, vol. 29. pp E. I. WERBER McCtenpon, J. F. 1912a An attempt toward the physical chemistry of the production of one-eyed monstrosities. Am. Jour. Physiology, vol. 29. 1912b The effects of alkaloids on the development of fish (Fundulus) eggs. Am. Jour. Physiol., vol. 31. Matt, F. P. 1908 A study of the causes underlying the origin of human mon- sters. Jour. Morph., vol. 19. Mecket, J. F. 1826 Uber die Verschmelzungsbildungen. Archiv f. Anat. u. Physiol., vol. 1. Moenxnuaus, W. J. 1904 The development of the hybrids between Fundulus heteroclitus and Menidia notata with especial reference to the behavior of the maternal and paternal chromatin. Am. Jour. Anat., vol. 3. Nest, F. 1914 Netzhautelemente im Opticusstamm. Studien z. Pathol. d. Entwicklung, herausgegeben von R. Mayer und E. Schwalbe. vol. 1. Panter, T. S. 1915 The effects of carbon dioxide on the eggs of Ascaris. Jour. Exp. Zool., vol. 19. Reagan, T. P. and Toortneton, J. Monror 1915 The vascularization of the embryonic body of hybrid teleosts without circulation. Anat. Rec., vol. 10, no. 2. Roux, W. 1895 Gesammelte Abhandlungen zur Entwicklungsmechanik der Organismen, Bd. 2. ScuwauBe, E. 1906-07 Die Morphologie der Missbildungen des Menschen und der Tiere. I. und II. Teil. Scuwause, E. und Josepny, H. 1913 Die Cyclopie. In Schwalbe’s Die Mor- phologie der Missbildungen ete. III. Teil. Serrz, L. 1913 Innere Sekretion und Schwangerschaft. Sirzenrrey, A. 1910 Odem der Placenta und kongenitale akute Nephritis mit hochgradigem universellen -Odem bei Zwillingen, die von einer an akuter Nephritis leidenden Mutter stammen. Zentralbl. f. Gynakol., vol, 34. Spemann, H. 1900 Experimentelle Erzeugung zweiképfiger Embryonen. Sitz.- Ber. d. phys.-med. Ges. Wirzburg Jahrg., 1901. 1903 Entwicklungsphysiologische Studien am Tritonei III. Arch. f. Entwmech., vol. 16. 1904 Uber experimentell erzeugte Doppelbildungen mit cyclopischem Defect. Zool Jahrb. Suppl., vol. 7. 1912a Entwicklung umgedrehter Hirnteile bei Amphibienembryonen. Zool. Jahrb. Suppl., vol. 15. 1912b Zur Entwicklung des Wirbeltierauges. Zool. Jahrb., vol. 32, Abt. f. allg. Zool. u. Physiol. SrockarD, C. R. 1907 The artificial production of a single median cyclopean eye in the fish by means of sea-water solutions of magnesium chloride. Arch. Entwmech., 23. 1909 The development of artificially produced cyclopean fish, ‘the magnesium embryo.’ Jour. Exp. Zool., vol. 6. 1910 a The influence of alcohol and other anaeshetics on embryonic development. Am. Jour. Anat., vol. 10. ORIGIN OF MONSTERS aye: SrocxarD, C.R. 1910b The experimental production of various eye abnormali- ties and an analysis of the development of the primary parts of the eye. Arch. f. vergleich. Ophthalmologie, vol. 1. 1912 An experimental study of racial degeneration in mammals treated with alcohol. Archives of Internal Medicine, vol. 10. 1913 An experimental study of the position of the optic anlage Amblystoma punctatum, with a discussion of certain eye defects. Am. Jour. Anat., vol. 15. 1914 A study of further generations of mammals from ancestors treated with alcohol. Proceed. Soc. Exp. Biol. and Medicine, vol. in@ Vrouicx, W. 1849 Tabulae ad illustrandam embryogenesin. Amsterdam. Werser, E.I. 1915a Is defective and monstrous development due to parental metabolic toxaemia? Anat. Rec., vol. 9, pp. 133-137. 1915 b Is pathologic metabolism in the parental organism responsible for defective and monstrous development of the offspring? Johns Hopkins Hospital Bulletin, vol. 26. _ 1915 ¢ Experimental studies aiming at the control of defective and monstrous development. A survey of recorded monstrosities with special attention to the ophthalmic defects. Anat. Rec., vol. 9. WuitTeneEaD, R. H. 1909 A case of cyclopia. Anat. Rec., vol. 3. Wiupver, H. H. 1908 The morphology of cosmobia. Am. Jour. Anat., vol. 8. Wotrr, B. 1914 Biologische Beziehungen zwischen Mutter und Kind wihrend der Schwangerschaft. In: Studien z. Pathol. d. Entwicklung, herausg. von R. Mayer und E. Schwalbe, vol. 1. PLATE 1 EXPLANATION OF FIGURES 63 Photomicrograph of a transverse section through the eyes and forebrain of the embryo in figure 3. fbr., unilobed small forebrain; o.p., fused olfactory pit. X 116. 64 Photomicrograph of a more posterior section (region of optic lobes) of the same embryo. 0.l., optic lobes; 0.c., optic cross. X 116. 65 Photomicrograph of a transverse section through the eye of the cyclopean embryo in figure 5. o.l., optic lobes. X 135. 66 Photomicrograph of a transverse section through the eye of the cyclopean embryo in figure12. fbr., forebrain; o.c.f., blastolytic optic cup fragment. X 116 67 A more posterior section (midbrain) of the same embryo, to demonstrate the large size of the fused eye and the small size of the defective midbrain. mb., midbrain; e.c., ear cartilage; m., mouth. X 116. 68 Photomicrograph of a transverse section through the anterior part of the eye of the cyclopean embryo in figure 6. p., pigment epithelium, r.c., rods and cones; 0.p., olfactory pit; br., brain. X 116. 69 A more posterior section through embryo in figure 6. m.o., medulla ob- longata; oe., oedema; pl., plasma of pericardium, into which the eye dips; p., pigment epithelium; 7., retina. X 116. 70 Photomicrograph of a transverse section through the eye of the cyclopean embryo in figure 8, fbr., forebrain; y., yolk. X 116. ORIGIN OF MONSTERS E. I. WERBER PLATE 1 o I or or a o PLATE 2 EXPLANATION OF FIGURES 71 Section through the region of the ear vesticles of the embryo in figure 8. The semicircular canals are on both sides all fused into one, m.o., medulla oblongata; ph., pharynx. X 116. 72 Photomicrograph of a part of a transverse section through the head of the embryo in figure 13 showing three small fused olfactory pits o.p. on the deformed side, o0.c., oral cavity. X 116. 73 A section more posterior of the same embryo. .e., the last part of the normal eye; 7.e., rudimentary eye; o.p., one of the three olfactory pits on the abnormal side; 0.l., optic lobe. X 116. 74 Photomicrograph of a transverse section of the embryo in figure 13, more posterior to the section in figure 73. s.c.c., semicircular canals; 0.l., optic lobe; o.l.f., optic lobe fragment; h., heart. X 116. 75 a fe ee lst a pate! a — iF 3. ta) Qoeb.F ». bidte i Ro ai *- 9c + a 9... aaa) ee Seba kcanea Seiea" peters ee = a Seah a tstats® siisaees 6.9, mat 2. A ee er bio “ed “ot pl etek * ho he iz aly * 7 733 ae 5 = oJ oe : ohetsres reed oe “8. Sale see OA. 0-8-0 O_8. aise! * CeCe 0 Hi aL) aa)e) *8 “i e ~ ble Sen esd Sete es « setae 9-95 Ft eke i dee * Reed ig 16 Rie TD ¥ mat ONL Pid e iSieeatecss ; nit p eB Ng! ve ‘ “+ Ay Ste eRe CRY Setateteiate _%.5 2 3-808 ik —_—s