if Seah relate et / 3 ag y ot okey F { thea ey t fy bpp eee et Seerte tere > t Nght Lar toy tert ty $70 Mes ee Ht A pteter ate. a O52 8 (Xe: ot At Ie ea nal ions MEE EC NLS if ene hehiat ta chen 3 He ¥ a’. wataineasys beta hy City ¢ Teter uete Satan ‘ cart, hfe’ trtee scare seein ached fy Pot Takes te sehr: Peete RUE ; Tne MMe Re aes Bete! alts f qy i maha Shou! Bs os Quik rt, ‘g 53 Df ga taky wate Tis asia Batlle Aen EMSC A SEE LE! tarsal ete be ite THE JOURNAL OF EXPERIMENTAL ZOOLOGY EDITED BY WiuuiaM FE. CastLEe Frank R. LILuiE Harvard University University of Chicago Epwin G. CoNKLIN Jacques LorB Princeton University Rockefeller Institute CuHarLes B. DAVENPORT Tomas H. Moraan Carnegie Institution Columbia University HERBERT S. JENNINGS GEORGE H. PARKER Johns Hopkins University Harvard University Epmunp B. WILSON, Columbia University and Ross G. HARRISON, Yale University Managing Editor VOLUME 17 1914 THE WISTAR INSTITUTE OF ANATOMY AND BIOLOGY PHILADELPHIA, PA. (COMPOSED AND PRINTED AT THE -WAVERLY PRESS es (Br THE WiLL1aMs & Wirxins Ci : BALTIMORE, gis S.A. we ly = Mes unant CONTENTS NOE JULY EK. J. Lunp. The relations of Bursaria to food. II. Digestion and resorption in the food vacuole, and further analysis of the process of extrusion. Eight figures and two ee eels @ se 46,6 6b ws 6)N 8 Ol uses Ss wie le els © #1. eJel= eee sls sie @) @ 0.6) 6) ep) 0)(6)|.0 8) vie 6 6) ua sive (6 \6)a | s = 0s & Ole 01018 ’CuHartes W. Metz. Chromosome studies in the Diptera. I. A preliminary survey of five different types of chromosome groups in the genus Drosophila. Diagram and twenty-six figures (plate) oe ss. a 6) nee os a ein 00 © (6) 8) 0 © @ 0 em, 0\6 8 a) 8)10.¢) @ ae eels» © vim so je) wie = = = elale ee C.M. Cuitp. Studies on the dynamics of morphogenesis and inheritance in experi- mental reproduction. VIII. Dynamic factors in head-determination in Planaria. Two figures WO \@ 0 whe a 6 6. 4))0\ 6) ee 16. 0 8 © 0 © 0,00 6 @)[6\ a,'6\0 0. e086, 00 = 00.0) 6 6 8 8)\0.u 6 o 6 «pe se 5 0 6 86 0 8 « = Seis 018 sie T. H. Morean. Two sex-linked lethal factors in Drosophila and their influence on the sex-ratio. Seven figures CHO CeCe mC Cacarer? ChUI Care r mC nny CML MCat het tery CO CU) COCuty Cs Cy Cty Ceci tI Jacques Lors. Cluster formation of spermatozoa caused by specific substances from Roscor R. Hype. Fertility and sterility in Drosophila ampelophila. I. Sterility in Drosophila with especial reference to a defect in the female and its behavior in heredity Pie \e\/e) ele s)/0) «040, «© 0) © (6) .0 (o)[e! v.08) |v. u 0's .0\ (ee) ©) 0) 0) 010) 2) 6) w 08 ©, 8.0 6 0, 6, « 6) 016! 0 wiv, 0.0 410) 0/s) 0) 9,,0 etellele ele ele) sie) = NO. 2 AUGUST Roscor R. Hype. Fertility and sterility in Drosophila ampelophila. II. Fertility in Drosophila and its behavior in heredity. Nine diagrams................-...seeee Brapiey M. Parren. A quantitative determination of the orienting reaction of the blowfly larva (Calliphora erythrocephala Meigen). ‘Twenty-four figures.......... a S. J. Hotmes. The behavior of epidermis of amphibians when cultivated outside the DOM Sem SCVMM HELIER Soh cier oie ete revere cre telve Foye ae eteceie) cists ajo-e « stcuelerenerecraicts elcid oeateiatel ake el tetarete Watpvo SHumway. The effect of thyroid on the division rate of Paramaecium. Three GINO EAC es 6 AR a REDE HIG Oe rE aOR eR TOPO crn CMC ram es aie cord o's dpie 45 61 81 1238 141 173 213 lV CONTENTS NO. 3 OCTOBER "Ty. H. Morean. A third sex-linked lethal factor in Drosophila. Three figures........ 315 HERMANN J. Mutter. A gene for the fourth chromosome of Drosophila Etuet Nicuoitson Browne. The effects of centrifuging the spermatocyte cells of Notonecta, with special reference to the mitochondria. Six figures............... 337 Roscoe R. Hype. Fertility and sterility in Drosophila ampelophila. III. Effects of crossing on fertility in Drosophila. IV. Effects on fertility of crossing within and without an inconstant stock of Drosophila. Twelve diagrams..................... 343 Eveanor L. Cuark anp Exrot R. Ciarx. On the early pulsations of the posterior lymph hearts in chick embryos: their relation to the body movements. Two charts. 373 RAYMOND PEARL AND Maynik R. Curtis. Studies on the physiology of reproduction in the domestic fowl. VIII. On some physiological effects of ligation, section, or MERON A OF, GE OVECUICE 20.5.5 am, cocteia 5 5 ees bose chee te tole UATE Oe & ORO tae eee ».. 395 NO. 4 NOVEMBER ~ x LoranpDE Loss WoopruFF AND RHODA ERDMANN. A normal periodic reorganization process without cell fusion in Paramaecium. Sixty-six figures (four double colored PLATES he pec atc talus cc ch: Sete hac ane tote cette aa eta ns a fey sees oc RA ed 425 Ross G. Harrison. The reaction of embryonic cells to solid structures. Fourteen HOUMES esc 5 Ce Rees sc SG Serie OR erie ERs cy eee SINGLET 2 os oe aE ein 521 Davin Day Wuitney. The influence of food in controlling sex in Hydatina senta..... 545 THE RELATIONS OF BURSARIA TO FOOD II. DIGESTION AND RESORPTION IN THE FOOD VACUOLE, AND FURTHER ANALYSIS OF THE PROCESS OF EXTRUSION EK. J. LUND Zoélogical Laboratory of the Johns Hopkins University EIGHT FIGURES AND TWO PLATES CONTENTS TLa Alea" EEPOLE Vo 1 Rage eerie Se RO NR Abe eh aA ee Re OR Rane Ema a i NMeatentalrandame lhodsaanuat ti scnee ecient inci Cae bits castors Nee arae 3 To what extent can yolk be considered a food for Bursaria?.................. 5 TETRA TORUS (0; Ueno een sca Moi inte Set mere a ees Oe Pett Tt na OAM 5 Ce aCe 8 1. The food vacuole; formation and physical changes..............-.--- 8 2, Chemical changes imtheimood yacuole...25-). 2.6 Sevdcn ese. +s sare ea 12 3. Effect of quantity of vitellin eaten, upon the average velocity of Co Ir Zersh 00) | Ne Re an Mee CR AA iy ae TRA ae 9 OLR ON gS ACR 15 4. Effect of congo red upon the average velocity of digestion, extrusion, SCRE OCR RE RAS Pe EMME aR aS yo Th Aig ARS Fst OmkeNe 22 Birestion ang resorphoniol fai, cage sacs os cles Sos oe eiiela tyres + 4 gees Rae 26 1. A demonstration of fat digestion and resorption in the food vocuole ORAEUTSEM PNA hoe cet eames eae Oca a ee a Rape sta ay ene Ica 26 2. Role of fat in growth and energy requirement....................5005- 28 3. Is fat formed from vitellin or starch in Bursaria?............,....++-- 29 The nature of some of the factors which bring about extrusion................ 30 i Hxperiments with paramin oil and: oltve! oil... sec casas aaa 30 2. Effect of mass or volume upon the extrusion reaction..............-. 34 pSiTTSTE E:T At PRE AAICD Ae ey, I Rei DOR aE RE ay ts ce PPAR ATS OE A Bat 38 ANT ETREY ATURE, CONSE ie ah, obi Gis Siediln o Pe He Ome ice hoeorcees Sie manera okey co .om een Gee fy INTRODUCTION To obtain exact information on the processes of metabolism in single-celled organisms, and particularly the Protozoa, experi- mental procedure must be based upon simple, well defined and reproducible conditions, so that the work may be repeated and verified. Attempts to fulfill this requirement are subject to the 1 THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 17, No. 1 gJuLy, 1914 2 E. J. LUND danger that the conditions may become so artificial as not to war- rant conclusions concerning what occurs under usual (‘normal’) conditions of the organism. Nevertheless, a certain degree of simplicity may often be obtained in the conditions of experimenta- tion and yet coincide with the usual conditions of a varied en- vironment to a sufficient degree so that the results give a definite answer to questions which are directed toward finding out what the principles are which underlie the processes. The following experiments have been carried out with an BAIS to fulfill the above requirements. In a previous paper (Lund 714) some of the most important external relations of Bursaria to food have been presented, in- cluding a demonstration of a selective extrusion among the food vacuoles. The following paper deals with the processes which take place in the food vacuole, and with the conditions under which it exists as an active system in this unicellular animal. Proteins, fats, and carbohydrates in the form of starches are present in the material eaten by Bursaria, as may easily be de- tected by direct observation or by: the aid of microchemical tests. Protein in the form of other living ciliates, flagellates, etc., is taken in and digested. Many of these Protozoa which serve as food contain fat globules and starch or amylum grains, and hence all three food substances may at times be found present in the cytoplasm of Bursaria from flourishing cultures. Since there is disagreement between results of different investi- gators, working with the same Infusoria; but more particularly because there are indications that similar forms among these animals differ in their food metabolism, the statements here made refer only to Bursaria and may or may not be valid for others, except where specifically stated. RELATIONS OF BURSARIA TO FOOD ° 3 MATERIAL AND METHODS Simple types of the three classes of food were found in lipoid- free egg yolk (vitellin), pure olein, and starch grains of various kinds. Potato starch was the most serviceable because grains of a very uniform size can be obtained by repeated decantation of — a suspension, and hence results may be expressed quantitatively in terms of unit volume. Fresh hard boiled yolk—a combina- tion, chiefly, of lipoids and protein—was also used. Olive oil of the best grade obtainable gave the same results as prepared olein. ~** From the yolk of the hen’s egg was prepared a protein (vitellin) that was perfectly lipoid-free, so far as could be determined; this was done as follows: An egg was boiled to hardness (15-20 minutes), and the granular portion of the yolk was kneaded gently to a moist, fine, floury pulp. This was placed loosely ina Soxhlet apparatus and extracted with an alcohol-ether mixture, for eight to ten hours; in this way the fat and lecithin were re- moved. The vitellin was then washed several times with fresh alcohol-ether mixture, removed from the Soxhlet and while still moist, gently kneaded to a fine white powder. This was left to dry at 25°C. for twenty-four hours. After drying was completed the residue was gently rubbed between sheets of filter-paper. When the steps mentioned above were carefully carried out, the grains of vitellin were found to be separated. This white dry powder was kept in a dry bottle in a cool place, and used as a food stock. When the animals were to be fed, a suspension of the powder was made in tap or distilled water and rubbed up; then left to settle, and the supernatant liquid with the smaller particles was poured off. This process of separating the smaller grains from the larger ones was repeated until a perfectly clear suspension ‘ of the protein replicas of the uniform hard-boiled grains was obtained. It is important to note that some eggs yield much more uni- form yolk grains after boiling, than others. Only those eggs were used in this work which were satisfactory in this respect. 4 E. J. LUND It seems highly probable from the results, that all traces of lecithin and fat had been removed by this method. Cold ether or cold alcohol extraction alone is not sufficient to remove all the lipoid content from hard boiled yolk. This will appear from results (p. 29) of experiments in which such vitellin grains were used. By this method of preparation, lipoid-free vitellin may be ob- tained in very uniform grains, and this permits one to fulfill a condition most important for precise experimentation. One can feed definite unit quantities to a single cell} and this makes it possible to attack a number of questions regarding metabolism that would otherwise be inaccessible, and to express the results in quantitative terms. Fresh hard-boiled yolk served as a protein-lipoid diet, which could be fed in the same way, for the vitellin grains’ are simply the protein matrix of the fresh hard boiled yolk grains in which the lipoids and any other alcohol-ether soluble substances are imbedded. There is, of course, no means of knowing whether the vitellin of the unextracted yolk grain does not undergo change during ether-alcohol extraction; but the fact (demonstrated by the experiments) that in the food vacuole digestion of the protein in these two forms occurs in the same way and with about equal readiness, shows that the process of fat extraction does not alter the chemical or physical nature in such a way as to interfere with the digestion and resorption of the vitellin. In all the experiments where necessary, the animals were placed in 500 ce. of tap water twenty-four hours previous to feeding. The result of this was that the food and débris con- tained in the cell had been discharged during the twenty-four hours of starvation, and hence a cell with perfectly clear cyto- plasm was obtained for the experiments. By this means a parti- cularly desirable uniformity in physiological condition of the organisms was obtained. The material was from the same wild stock cultures as that used in the experiments reported in my earlier paper. RELATIONS OF BURSARIA TO FOOD 5 In most of the experiments the organisms were fed singly, and large numbers were used, so that any individual variations play a minor part, if any, in the final results. Where temperature regulation was necessary the experiments were carried on in an oven kept constant to within 1.5°C. TO WHAT EXTENT CAN YOLK BE CONSIDERED A FOOD FOR BURSARIA As yet it has not become possible to propagate pure races of ‘Bursaria from single individuals, or in culture, by feeding it an artificial diet of vitellin or fresh yolk; so that it must be re- membered in conjunction with the facts presented at this time, that the full requirements for the maintenance of life and re- production have not been fully satisfied, and to this extent the ideal conditions have not been met. But to prove that vitellin and fresh yolk are drawn upon for the energy requirements and growth of the cell Experiment I is given: Experiment 1. Three sets of 74 individuals each were starved in tap water for 24 hours. Set A was fed with fresh hard boiled yolk in suspen- sion prepared as described above. Set B was fed vitellin; while Set C was not fed. Each individual in the three sets was left to eat as much as it would in 20 minutes. They were then picked out and washed once in tap water which had been boiled, from this they were removed to watch glasses each containing 2 cc. of boiled tap water. The water was boiled in order to kill any bacteria, or other organisms sometimes present in small numbers in the tap water. Two individuals were placed in each watch glass and these were set m moist chambers which were kept side by side. There was therefore no difference in the temperature at which the three sets remained during the experiment, although a daily fluctuation in temperature of 4 or 5°C. occurred. Such variation, however, does not affect the results of the experiment for the particular end in view. It was found that after feeding twenty minutes nearly all of those in the fresh yolk suspension (A) had eaten large quantities, while those in the vitellin suspension (B) had on the average not eaten so large a number of ‘grains as those in the fresh yolk suspension, although in both Set A and Set B, each individual had eaten more than one grain. All the indi- viduals in the three sets were treated identically except in feeding. All were normal and active at the beginning of the experiment. At the end of every 24 hours detailed records were taken for each individual, as to whether it had a normal form or had undergone dedifferentiation— 6 E. J. LUND which is under many conditions a typical reaction—and if the latter were the case, to what extent it bad proceeded. Death was taken to have occurred when the cell had ceased to move and began to disinte- grate. In this way a record of the effects of the food on the maintenance of form, degree of activity and length of life, was obtained, so that from these the means can be taken and compared. To save space, instead of giving the detailed results in the form of tables they are given by curves and averages. Curves A, B and C of figure 1 represent for the three sets respectively, the longevity or death rate. Points on the abscissa indicate the time in hours, while points on the ordinate represent the number of individuals which were still alive at the time the record was taken. The curves A and B of Sets A and B respectively, show clearly (1) that the yolk and vitellin had a definite effect in prolonging the life of the cell. The relation of curves A and B to each other will become clearer when we have considered later experiments (p. 28) which show, other things being equal, that we should expect longer life from individuals fed both lipoids and protein. By comparing the average length of life in the three sets we find that Set A lived 4.98, Set B, 5.39 and Set C 3.20 days. Sets A and B, therefore, lived on the average about two days longer than the unfed Set C. (2) The vitellin grains underwent total di- gestion and resorption in most cases, while the fresh yolk grains were generally only partially digested, this was especially the case where more than 2 or 3 grains were eaten. (3) The animals which had been fed were on the average more vigorous than those of Set C. (4) The animals of Set A grew to be larger in most cases than those of either Set B, or Set C. Many of Set A became twice as large as those of Set C, while those fed vitellin were on the average larger than any of Set C. The maximum size was reached about thirty-six hours after feeding. There can therefore be no question but that both the fresh yolk and the vitellin entered into the chain of metabolic proc- esses and were in part at least, drawn upon by the cell for its energy requirement. But very few of the organisms divided. Whether yolk or vitellin is a sufficient diet for cell division as RELATIONS OF BURSARIA TO FOOD *pey you ‘StL (Q PAIND) O 409 OTTYM “UTTTOITA poj SBM (q 9AIND) { 499 ‘yjoX YSo1J pay SBA (Y PAIND) Y OG “Yovo spenprarpur pL JO S}OS 9014} UI OjI] JO YASUI] oy} Surmoys soamny) T ‘SI "SYH 88z y9z ov 12 z6l gl vvl ozt 96 ZL ay vz 8 B.S. LOND well as for growth can not be answered at present. But for the questions dealt with in the present paper the important fact to establish is that the yolk or vitellin can be drawn upon for the energy requirements and growth of Bursaria. PROTEIN. DIGESTION 1. The food vacuole; formation and physical changes The passage of the grain into the body is brought about par- tially by ciliary action at the base of the buccal pouch; but large grains or masses of food are pushed through the base of the gullet and into the endoplosm, by what seem to be contractions of the wall of the base of the gullet, and also apparently by activity of the endoplasm about the gullet behind the food. During this process of swallowing, a liquid comes to be included about the food so that when the vacuole separates from the base of the gullet liquid surrounds the yolk grain. Where does the liquid enclosed with the grain come from? Part of it is derived from the external medium, as is readily determined by direct observation of the process of swallowing. But the liquid in the vacuole, when the latter is separated from the base of the gullet, is likewise partially made up of an acid secretion from the cytoplasm of the lower portion of the gullet, as will be demonstrated below. When the vacuole has formed it is usually carried toward the middle of the cell and may remain practically stationary, especially if it is large. If movement takes place, which is nearly always the case when the vacuole is small, then the vacuole may traverse any part of the cyto- plasm, and in any direction. In Bursaria there is no such regu- larity in the course of the food vacuole as has been described for Paramecium by Nirenstein (’05); and for Carchesium by Green- wood (’94). Often digestion and resorption begin and are completed in one and the same place, without any circulation of the vacuole. Residues are extruded from a small area on the dorsal side of the body. The first visible change which takes place is the absorption of the liquid which has been enclosed with the grain during the RELATIONS OF BURSARIA TO FOOD 9 formation of the vacuole. This is always definite and can generally be readily observed and followed throughout the process. This is exemplified by the following experiment: Experiment IT. Single normal individuals which had been placed in tap water 24 hours previously, were each fed a single grain of fresh yolk and the time interval noted between the separation of the vacuole from the base of the gullet and the point of complete disappearance of the liquid about the yolk grain. Table 1 shows records from ten individuals each fed one grain. From the table it will be seen that the duration of the process is relatively uniform. The point when all the liquid about the grain has been absorbed was determined to within about thirty seconds. TABLE 1 Experiment II INDIVIDUAL NUMBER 1 | 2) 3 4 | 5 5} Gi [3 | 9 0 10° AVERAGE L Number of minutes for resorption of AG al liquid about yolk grain............ 4 | 33 3 43) 7 f+] 33 5 | 4.45 When more than one grain is enclosed in the vacuole the same process of absorption takes place, following the separation of the vacuole from the base of the gullet; the vacuole membrane be- coming closely applied to the surface of the grains as is shown in figure 2. Different vacuoles in the same cell are quite independent as to the absorption of fluid. To illustrate, figure 3 is given. Here the new vacuole contained a comparatively large amount of liquid about the grain (in most cases the quantity of liquid is less), and came to be located close to an older vacuole in which the process of digestion of a Paramecium had practically been com- pleted, and which contained a considerable quantity of liquid. The vacuole containing the digested Paramecium was unaffected in size by the absorption of the liquid about the yolk grain in the newly formed vacuole. An explanation on the basis of osmotic relations within the cytoplasm and in the vacuoles, assuming the vacuole membrane 10 E. J. LUND to remain in a uniform condition, might perhaps account for the visible difference in the vacuoles in this and similar cases. But a conception of this process may more properly be obtained if we Ke, ) mer ays eel eS e, POO fy? Fig. 2 Shows the resorption of the liquid included with the yolk grains during formation of the vacuole. A, food vacuole just separated from base of gullet; B, the same, five minutes after separation. Fig. 3 Showing the independence of vacuoles with respect to resorption of liquid contents. The larger vacuole containing a partially digested Paramecium is not affected by the resorption of the fluid in the vacuole containing the fresh yolk grain. B drawn 53 minutes after A. think of it as a process of imbibition of the liquid by the colloidal cytoplasm, accompanied by changes in the permeability of the vacuole membrane. The problem here must be similar to that RELATIONS OF BURSARIA TO FOOD 11 which will confront us later regarding the absorption of the liquefied products of digestion. Within one or two hours after the grain comes to lie in the vacuole, the beginning of solution becomes apparent as a lique- faction and consequent rounding off of the corners and_ edges. This proceeds in exactly the same manner as when cubes of coagulated egg-albumen are digested by the action of pepsin- hydrochloric acid. At the end of digestion no solid remains of the vitellin grain can be seen; all that is left is more or less liquid in the vacuole. Another important visible change in the vacuole consists in the usual second appearance of liquid about the vitellin grain. It seems most in conformity with the observed facts to consider this as a consequence of the formation of soluble products of digestion, and as being due to the fact that the rate of resorption of these liquid products is less than the rate at which the lique- faction of the vitellin grain takes place. The reason for this conclusion is partly based upon the fact that in some cases the whole process of digestion of the vitellin may go on and be completed without the appearance of any visible liquid between the grain undergoing solution, and the vacuole membrane. In this case it is obvious that the rate or power of resorption is equal to or greater than the rate of solu- tion of the grain. Furthermore, it is to be expected that as digestion of the protein continues, the affective osmotic con- centration of the cleavage products increases and this obviously - may bring about an increase in the liquid contents, provided that the permeability of the vacuole membrane (resorption) does not undergo a simultaneous, proportional increase. In other words, the absence or presence of liquid during digestion of vitellin is a resultant of two sets of conditions (a) the rate of digestion and (b) the rate or power of resorption, one factor of which is the degree of permeability of the vacuole membrane. This further agrees with the fact that digestion of a vitellin grain is not always at a uniform rate. A rapid solution of the grain sometimes begins shortly after eating, changing later to a slower one. Even 12 E. J. LUND food vacules formed at the same time, of the same material, and in the same individual, may behave diversely as to the resorption of liquid, and also in rate at which digestion takes place. But this individuality of behavior in different vacuoles does not prevent the conception of equilibrium from being applied to the phenomena of resorption. Essentially the same observable physical changes take place in vacuoles containing fresh yolk as has been described for vitellin. Further evidence for this view will be given on page 25. 2. Chemical changes in the food vacuole Some of the chemical conditions in the food vacuole of Bur- saria during digestion of vitellin can be shown by the use of sensitive indicators adsorbed by the grains. Vitellin or fresh yolk grains stained with neutral red remain bright red during the whole process of digestion. Grains stained in an alkaline alcoholic solution of alizarin, quickly lose the blue color and remain colorless throughout the digestive process. Similarly, grains stained in an aqueous blue litmus solution change to red and remain red until nothing remains of the grains. These three indicators agree therefore in showing that the whole process of digestion of vitellin takes place in an acid medium, and that during no part of the process does alkalinity appear, as it does in food vacuoles of Parmecium and some other ciliates. Grains of vitellin or yolk stained with congo red become dark brown after one, two or more hours, and continue to remain | dark brown throughout digestion. In some grains no change in color from red to brown takes place, but what differences in conditions account for this has not been determined. ‘Two other indicators were used, Tropaeolin 00 and diethylaminoazo- benzene. Vitellin or fresh yolk stained with Tropaeolin 00 showed little or no change in color. Diethylaminoazobenzene likewise showed no change. This result is due to the fact that these two indicators are not sensitive to very weak acids or strong acids in high dilution. RELATIONS OF BURSARIA TO FOOD 13 The points of origin of acid as well as the rate of acidification of the vacuole contents was determined. Table 2 gives typical records of the total time for completely acidifying grains in fifteen out of fifty individuals which were fed with fresh yolk stained in blue litmus. No noticeable difference existed between the acidification of fresh yolk and that of vitellin. The time in seconds which it took to change completely the blue litmus- stained grain to red, is given by the numbers underneath the number of the individual. The average of the observations in table 2 is 38 = seconds. TABLE 2 Total time in seconds for acidifying fresh yolk grains stained in blue litmus. The table shows typical records of fifteen out of fifty individuals in which the time for acidification of each grain was taken. The grains which are represented by the numbers opposite to the brackets were included in the same vacuole. The other numbers, not opposite brackets, represent grains, each of which were in a separate vacuole, i.e., swallowed separately INDIVIDUAL NUMBER 1 2 | 3 65 6 7| 8 9| 10) 11. 12 13 14 45 | | | | | Ist grain......... 15 20 ‘ 25) 20, 65 45 60 80 4515 ‘ | 45 (35 | 55 Od grain..........| 35 30 | 20 15 40 40 10 30 20 30 {30| )60 }35 | 30 3d grain..........| 50) | 55 | | 15 80 | 50 20, 30 40| \50 |35 | 25 4th grain......... ali’ <40 | 15 | | 25) 60 45| |40| [35 Sth grain......... | | so] | 20 hole Weal | 60} {30 | Gil grain’. ...>.:.: | 2p | 30 | Fy et | 45 | 7th grain......... | | | 20 Wa 75 SED Pram 4.2% 4,54 | | 180, he. 30 It was evident that the time of acidification depended mainly if not solely upon three conditions, as follows: (a) the size of the grain; the larger the grain, other things being equal, the longer it took for the last trace of blue color to disappear. For example; grain number 8 of individual number 5, table 2, was a very large grain; while grain 2 of individual number 8 was very small, hence the difference in time. (b) The physical consistency of the grain must be assumed to determine its permeability and hence its time of acidification. (c) The concentration of the acid secreted into the vacuole. @. 14 E. J. LUND It has so far not been possible to determine the relation between the number of grains eaten and the time of acidification of each, because litmus is slightly toxic to Bursaria and hence stained grains of even fresh yolk are not eaten very readily. The arrow at a in figure 4 shows the point at which the acid first begins to be secreted into the forming vacuole. This is indicated by the sudden appearance of a faint pink color about the edge of the erain which gradually increases. Figure A in plate 1 shows the degree of acidity reached before the grain leaves the gullet. Fig. 4 Outline drawing of Bursaria showing steps in the formation of food vacuoles and resorption of the liquid enclosed; a, place at which acid secretion begins. Grains of relative size shown will not as a rule be rejected after they have passed the point indicated by b. As the grain passes through the mouth surrounded by the mixture of the acid and the external medium, the change toward red progresses rapidly until the grain’s interior has been reached by the acid (plate 1, figs. A, B and C). When the acid had reached the center of the grain the times noted in table 2 were taken. There is no way of telling whether acid is continuously poured into the vacuole after it has left the end of the gullet. No evidence was obtained as to the nature of the acid secreted, nor as to whether it is in combination or in the free condition. * RELATIONS OF BURSARIA TO FOOD 15 The physical and chemical changes taking place in the vacuole containing vitellin or fresh yolk grains, which have been de- scribed above, are briefly summarized in plate 1. Here the series of figures from A to L, inclusive (series 1 and 2), show the usual course of the process of digestion and resorption of fresh yolk when the latter is retained throughout the process. Series 1 and 3 show the usual course of digestion and resorption of vitel- lin, when the rate of resorption is less than or equal to the rate of digestion. Series 1 and 4 show the same when the rate or power of resorption is equal to or greater than the rate of digestion of the vitellin grain. The successive figures do not represent the condition at equal intervals of time, but are typical stages in the process. Frequently there occurs an alternate presence and absence of liquid or smaller variations in the quantity of liquid around the grain during the stages from H to L, inclusive, and from H’ or H” to L’ or L”, inclusive. Thus the figures from H, H’ and H” to L, L’ and L” respectively, represent this process only as it proceeds typically in the majority of cases. 3. Effect of quantity of vitellin eaten upon the average velocity of digestion Since each cell could be fed just the desired number of grains of practically uniform size, the effect of the quantity of protein upon the average rate of digestion, and upon certain other processes could be determined. Experiments arranged to find out what the relation is between rate of digestion and the quantity eaten, were tried with both fresh yolk and vitellin. Fresh yolk was found unsatisfactory for this purpose, since the grains were generally not retained but extruded before digestion was completed. This was especially true when a large quantity of fresh yolk had been eaten (cf. Experiment VIII, p. 34). Under usual conditions vitellin fed to -the animals was not extruded, and this was especially true if care was taken not to let the animals eat too many grains. The average limit in the number of grains that could be eaten and retained -was determined from a large number of observations 16 E.. J. LUND during the work. In these it was found that when normal, active animals were fed vitellin under what seemed to be the best experimental conditions, five to seven or even nine grains were retained, and digestion of the whole mass continued to com- pletion, leaving no residues. In the experiments, to determine the average rate of digestion of vitellin the maximum number of grains fed was six, and it was found that with this number as the maximum extrusion very rarely took place. Rise in temperature accelerates the digestive process, but the variations due to temperature were eliminated, since all the sets of individuals were side by side in an oven kept at 25 to 26°C., and the fluctuations in temperature due to taking the moist chambers from the oven while the records were taken, were the same for each set of individuals. The organisms used in each experiment were from the same healthy culture. They were starved in tap water twenty-four hours previous to the beginning of the experiments, so that a clear cell was obtained. When the animals were to be fed, a large number were removed from the 500 ce. dish of tap water in which they had been starved, to an 8 cc. dish containing tap water. Some of the vitellin suspension prepared as described above (p. 3), was added to the tap water in the dish containing the animals. The quantity of vitellin suspension added varied according to whether it was de- sired to feed a small or large number of grains to each individual. If a large number of individuals were to be fed only one grain, a very weak suspension was added; the result was that the rate of feeding was slow, and hence gave sufficient time to remove the individuals as soon as one grain had been swallowed. As feed- ing went on, the individuals that had eaten the desired number of grains were picked out with a pipette and placed in separate dishes containing tap water. In this way, under favorable con- ditions, it was possible to obtain several sets of individuals - simultaneously, each individual of each set having eaten a defi- nite number of protein grains. After feeding, the animals were RELATIONS OF BURSARIA TO FOOD 17 removed from the dishes with tap water, and two individuals were placed in each watch glass in order to lessen the labor in taking the records. Each watch glass contained 2 ec. of tap water. The watch glasses were placed in moist chambers. In order to avoid the effect of individual variations in the rate of digestion of equal numbers of grains, forty-eight individuals were used in each set. Effects upon the result due to variation in size of the grains were therefore also practically avoided. Digestion was taken to be complete when all the solid contents of the vacuole had become luquefied. The results of two experiments are given. Experiment III, February 12, 1913: Table 3. Each individual of Set A, Set B and Set C was fed 1, 3 and 6 grains respectively. Exami- nation of each individual was made and records taken, 33, 5, 64, 8, 93, 11, 123, 14 and 22 hours from the time of feeding. The rate of digestion and resorption when a large number of grains have been fed is in many cases slow toward the end of the process; so that it is sometimes difficult to tell at just what time (within 1 or 14 hours) the last remains of solid protein disappear. For this reason eighteen hours was taken as the average time of complete digestion in those individuals of Set C which had small protein residues at the end of fourteen hours and none at twenty- two hours. All of Set C showed complete digestion at the end of twenty-two hours, In each of the three sets (table 3) individuals numbered 2 show a longer time for digestion than their partners numbered 1. This has no significance, because it is due to arbitrarily calling the first individual that had completed digestion, number 1, and the last one, number 2; when the record was taken. The same is true for table 4, Experiment IV. It will be seen from table 3 that the average number of hours which it takes to complete digestion in the three sets A, B and C, is not directly proportional to the number of grains eaten, i.e., as 1:3:6. Before further consideration of these results the next experiment will be described. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 17, No.1 LUND J. E. 18 | +21 e] ace { SI {ISI SI | FIST ZIT IST IST Tal STIST IST \8T |8T IST | SIS 11 EIST key] Ged fcc pee “6 ON [eNpriAtpuy ST ST (26 ST IPL [$208 SZrrt IPT \IT |$OTECTiIL [PT [PT IIT IFT 26 |ST 16 £6 ISL SI “T (ON [enprAtpuy UI[JOPIA JO SUIBAS Q PoF YOO ‘S[BNpIAIpul qyS10-44.10,.—9 4ag +61'6 eee G | IIIT | PIPL | 8 FT 1 8 | 8 | 8 [fet 8 26 26 \26 PI S21] 8 6 F146 opal ah ol ee a GON [BNprAtpuy \ G (26 IIT |§9 \IT $9 IT [$9 | ¢ 9 | ITO 16 t 8 FL 86 179° S EO 6 | GIS | SP T ON [enptAtpuy UI][OJIA JO SuIBId E POF Yovo “SyeNpIAIpUl 4ys1o-A410.J.—_ 409 — * = 90°2 ee fo G |%6 £9 [6 { g 29 | 8 [9 | 818) 8 6 \%6 £9 26 EZIK6 | 8 16 £9) WeeleG@eln ss “-g ON [BUptAtpuy : £9 |¢ \¢ |€9 |¢ 19 ¢|¢ 29|¢/)|¢ \$9 #9 | 8) ¢ #9) ¢ £9 |$9 | g [$9 29 | Gig}: T (ON [@nprAtpuy NOILSASIG . ¥Z | &B| 22 | 12 02 Or sr LT | Of | ST | #1 ‘et | al II ‘or GaSe; 9 |) el 8 |] te oe @LaTdWOO OL | SHUNOH 2 u pS ee oa aWzOu ‘ON SSVID HOLVM UT[[P}TA JO ULeIS [ poy Youo ‘spenprArpur yysIo-A7.10.J—V 498 uornjsabip ajajdwoo sof sunoy fo saqunN—sI6] “Gl fsonigay ITI quauriad xy € ATAVL 19 RELATIONS OF BURSARIA TO FOOD ie l. (ee sage 02 TOGIEGG 02 SI tog 2GG| OcleGalScc sGaitoolsS1 is ed 0z|227|2zz\¢ Ss] eee G ON [BNprAtpuy le | 16 | 16 al iat a TSTFST| FT ST EST] 8 PUEST TSI | | — eM AST el) one | Pedtedted ade eel ale LI| LUEST LL ra) aITOgtA jo uoIsn.}xo [eyaed jo osnvoaq SuOTyB[No[vo WOIJ poz}IUIO XY PoYIVUL AMNOF SUTT[9}IA JO SUTeAS g poj Youe ‘spenprtArpur 4YSI0-A7.10 T— 09 | ee | Wes 26 | PLZ6 ESTIIT ST 1 fT eeuct Est 21| TUEStlestl ZT) °° °G ‘ON [BNpiAtpuy 1i$6 \$21| 8 | 118 | 8 /$6 \$6 29 | TI] I1eeT/g6 |¢ PL FL ESI on aes T ON [enptarpuy LL GT G19 Ka Tt 6 SST/E81le Wi Gl 8 cree tiscris6 IT aa mIN RIN us rei UI[JOPIA JO SUIBIS E Pof YOO ‘STBNpTAIpur JYste-Aj10,J—D og ext | 2zq LAE fOH 8 | Ft] FUZT Foule6 | TI FZTeeT| 8 |k6 wr esT{¥oTezTeoTeeT| HI) STPT | TIIesT| "GON [eNpiArpuy é "8 16 | 8 |£6 [E2T/pT | 8 6 |f6 [£9 [£6 [9 [£6 Ss [£9 | TI] IT] 8| 8 Ferg6 [IT f6 Kal "LON [enprarpuy | | | | | : | | UI][O}IA JO SUTBID Z PSF YOR ‘S[VNPIAIPUr yYsto-AQIO—_ JOG SYNOHUTAWON | 1VLOL | | fr a [i 9¢°8 | ite | 8 |EZI/IL 186 | LUECTEG 6 [£9 | 8 | 8 26 | 8 \f6 | 8 IIT IT 26 | 8 $9 | 8 | TT 11 8 mniay "SON [eNprlArpuy = | isis |8 te ge kelsicisgolsisiolsis ¢ is 18 | ¢ 39 6 6 a a ca ‘TON [8nprarpuy Reioee e eae = eae s | e | a NOWSaDId | ¥ | &@ | G2 | Te | OZ | GT | SE | AT | OF | ST | H el BL HE OG | Pe) ae ae N) te | if GLATdNOOOL | SHN0H |— ts GOVuaaAV | ON SSVIS HOLVM UI[JOFIA JO UIBIS | poy POV “S[ENPTATpUr yysTe-A4410,J—VW 49g uonsabip ajajdwos sof sinoy fo LaqQUunn—SI6] ‘8] hannagay : AT quamiuaday % ATAVL 20 E. J. LUND Experiment IV: Table 4. In this the procedure was the same as in the previous experiment, but the animals were taken from a different culture, and instead of 3 sets of 48 individuals each, 4 sets of 48 individ- uals each were used. Each individual of Sets A, B, C and D were fed 1, 2, 3 and 6 grains, respectively. Four individuals (marked X in the table) of Set C extruded part of the six grains and are therefore not counted in the results. The temperature was 25 to 26°C. Examinations were made 33, 5, 63, 8, 93, 11, 123, 14, 153, 17, 183, 20 and 223 hours after feeding. In this experiment the possible slight error in the average time for complete digestion in Set C, Experiment III, due to the approximation of the time of complete digestion in some of these individuals, is eliminated, because observations were made regularly throughout the whole 223 hours. And the conditions were on the whole better than in Experiment III. The results of both experiments agree in showing that the aver- age total amount of vitellin digested per unit of time is greater in individuals fed a larger quantity than in those fed a smaller one, i.e., this quantity is greatest in Set D> Set C> Set B> Set A. The average time, found by experiment, for the complete digestion of one grain or its equivalent in volume or mass, in the sets of individuals of the preceding two experiments is given in column 2 of table 5. Now if we establish a limiting case, by supposing that the average quantity of vitellin digested per unit of time, is a func- tion only of the amount of surface of the substrate exposed to the action of the digestive agent, then we should expect to find (provided further that the total surface of each grain in the individuals fed more than one grain was exposed to the action of the digestive agent), that individuals fed 6, 3 and 2 grains each, would complete digestion of all in the same length of time as it takes to digest one grain. Hence, if it takes 8.56 hours (Experi- ment IV) to digest one grain, then in Sets B, C and D we should expect to find that the quantity of vitellin equal to the volume of 1 grain, would undergo digestion in the times shown in table 5, column 3. RELATIONS OF BURSARIA TO FOOD PA The results are uniformly less than those found by experiment (column 2). Now since we have constructed a limiting case for the maxi- mum average rate of digestion on the assumption that the rate is in direct proportion to the surface of substrate exposed, we may on the same assumption establish a limiting case for the mini- mum average rate of digestion, by assuming that the surfaces exposed to digestive action in the four sets, are to each other as the surfaces of spheres of vitellin having volumes 1, 2, 3 and 6. Calculating this on the basis of the rate of digestion found by experiment for one grain (Set A), we obtain the values given in column 4, table 5. It will be noted that these values, based on the above assumption are all greater than the values found by experiment. The results of both experiments agree. However, the intermediate position of the obtained results (column 2, table 5) does not exclude the possibility that the digestive process TABLE 5 Summary of and calculations for the average velocity of digestion of vitellin, based on results of Experiments III and IV, tables 3 and 4 COLUMN 1 2 3 4 5 6 ses [gaze [ggees jeez | 2 miele Sea 8 BEa7s B33 EXPERIMENT Ser so) g.2° ents |phosa | BE a3 233 aUn aS ae gae nae = (= ao 500) Sok ROO QHRZAGD| S Bo | g88 [5o253\sge2 aise. =| 22 es | go | gEGSElgegeeeseset) =! go | gaa. [atest esos ziags, | © e3 BGes 2. O88 BLoORS a Rims 3 33 286s | Goss Sloas | Sore 4 Zi < 6) 0 |< > | hours hours | hours hours Set A...) 1 7.06 | | 7.06 ts ae Set Bu: 3 5:05), |, “2850 4280, || 4506), aes) i pa Set C...| 6 DBD Me Tek |). Sc88 a: | 124846: | RGnee | = == ee | (| Set A... 1 8.56 | 8.56 IV. Febru-}| Set B...| 2 5.62 499) | 6.76 6.04 7.96 ary 18,1913 )| Set C...| 3 4.25 2.85 | 5.93 4.94 | 7.38 Seti D:.2)/ 116 3.04 W742, | 4S70N Wa oe4Ou, W744 pi, E. J. LUND really conforms to the law of a heterogeneous chemical reaction (between a solid and a liquid). The values in column 5, table 5, were obtained by substitution in Arrhenius’ formula t=kV M for the rate of digestion in the dog as found by London.? The values obtained by means of this formula agree well, within the limits of experimental error, with those found by experiment in the case of Bursaria. In column 6, table 5, are given the values of & in the formula t=kV M. The agreement is not so good in Experiment III as in Experiment IV for reasons given above, although the variation in the constant k of both experiments, and especially that of Experiment IV, is well within the limits of experimental error. If it were possible, it would be interesting to determine what the value of k would be for other food substances. The experiments show that in spite of individual variations among the organisms and the variation in the process in different vacuoles in the same individual which were mentioned above, the sum total gives a definite result, and shows that the average rate of digestion of vitellin under rigid experimental conditions follows a definite law. 4. Effect of congo red upon the average velocity of digestion, extrusion, etc. | In order to discover some of the changes which take place in the reactions of the cell when protein (vitellin grains) was changed by letting it adsorb a substance from solution, the following experiment was carried out. Congo red was chosen since its toxicity to Bursaria is low when compared to that of many dyes, 1 The intermediate position of the values found suggests agreement with the results of the experiments of Bayliss (’04), who found that ‘‘with concentrations of casein up to about 4 per cent, the velocity of digestion is proportional to the concentration of the substrate . . . . . whilst with more than 8 per cent, inverse proportionality sets in’’ (cited from Euler: General chemistry of the Enzymes, p. 186). Furthermore the results also suggest the possibility that the quantity of active agent produced in the cell or vacuole is not directly propor- tional to the quantity of vitellin eaten. 2 This is, =kVM; where t is the time for complete digestion, M the amount by weight eaten, and k a constant depending upon the nature of the food. RELATIONS OF BURSARIA TO FOOD 23 and since vitellin grains readily adsorb it in considerable quantity, depending upon the length of time during which the grains are left in the aqueous solution of the dye. _The results of the experiment clearly show: (a) that the ad- sorbed congo red very markedly interferes with or prevents di- gestion of the parts of the vitellin grain toewhich the congo red has been adsorbed, and (b) that it brings about a condition which leads sooner or later to an extrusion of the contents of the vacuole; (ec) that it may exert a greater or less toxic effect from within the vacuole, upon the cell, which may lead to an earlier death than if the unstained vitellin had been eaten. In short: the chem- ical nature of the substance taken into the vacuole of Bursaria may determine in various ways what many of the conditions and reactions of these organisms will be, and especially, in this connection, the action of the digestive agent and the process of extrusion.? The following are the results in brief. Experiment VI. Grains of vitellin of uniform size, prepared as de- scribed above, were stained twenty minutes in a deep red aqueous solu- tion of congo red. They were then washed several times until no more stain could be washed out. An unstained portion of the same vitellin sample was washed the same number of times in tap water; this was fed to the control individuals. Forty normal individuals previously starved in tap water for 24 hours were each fed one grain of the stained vitellin. Similarly a control set of 40 individuals were fed, each one grain of the unstained vitellin. All the conditions and material were the same in the two sets, except that the individuals of one set was fed stained grains, those of the other set unstained ones. Both sets of individuals were kept in watch glasses, each containing 4 cc. of spring water. Two individuals were placed in each watch glass. Records of each individual were taken 2, 4, 6, 7, 9, 11, 13 and 22 hours after feeding. Table 6 gives the results. Table 6 (a) shows that the partially digested or undigested congo red stained grains were to a large extent extruded, while in the control set, all retained the unstained 3 [.have here used the term ‘chemical nature,’ and in my previous paper (Lund 14) such térms as ‘toxicity’ (p. 29), ‘specific chemical properties’ (p. 41), with a general meaning. The finer distinction between the effects due to chemical and physical properties of a substance, as for example its solubility in the plasma membranes, state of colloidal aggregation, chemical reaction with the protoplasm, etc., remains an open question. 24 E. J. LUND protein grains. This, therefore, shows that the chemical nature of the substance in the vacuole may, and does in this case, deter- mine whether it shall be extruded or not. Table 6 (b) shows the effect of the adsorbed dye upon the process of digestion of vitellin. The stain retards and sometimes even prevents digestive action. ‘This was seen from the fact that the corners and edges of the grains where little or no unstained yolk exists, persist more markedly than those of the unstained grains. The stained grains, nevertheless, often became smaller by shrinkage, caused apparently by solution of the protein from the deeper parts of the vitellin grain, which may have been stained less deeply. That the adsorbed dye interferes, to some extent, with the maintenance of the normal form and outline of the animal is shown by the results of table 6 (c). The fact that, for example, the total number of normal individuals of those fed a stained grain has increased rather than decreased at the eleven-hour record, since the nine-hour record was taken, is due to regulation of form, either by closing and opening the oral pouch, or by undergoing TABLE 6 Showing the results of feeding one grain of vitellin stained in congo red, to each of a set of forty individuals. For comparison (control) another set of forty individuals were each fed 1 grain of unstained vitellin TIME IN HOURS AFTER FEEDING 2{ 4] 6| 7| 9/11] 13| 2 (a) Total number of individ- Stained: (8.0) ah gh eol ale). gi aghoetes uals that had extruded } need 000000 0 0 grains before death J ae oar (Dy Bete member ian nas aiceac dae oo 0000000 uals that had completed’ | Ty, stained | 0 1 4! 10 22| 28| 37 digestion of grains DLs aiaaie tment Cc) iota) mune 08 avi cried eee 38) 35) 27 26) 15| 24! 23] 14 uals that were Hormal ini) 77 ieined|. 0. 37| 38| 37, 38 34 331 35! 29 Stamedeuoaeeneeee o.0 0 0 0 3 11 Unstained........ 000 0 0 OF O 2 form J (d) Total number of individ- uals dead (e) Total number of individ-| uals with a perceptible | | Stamed.).-..0\:... 700 0 3 2 0 0 amount of liquid 2 Unstained........| O| 4 vacuole ~J = ies) — Ww — ler) w bo «- RELATIONS OF BURSARIA TO FOOD 25 more fundamental morphological changes. Set B> Set A. This becomes more apparent if we calculate the percentage of the total number of grains fed which was present in the individuals at the times when the records were taken. In order to bring out these relations more clearly the curves in figure 8 were notes from the percentages given in table 7. Now, from what has thus far been said, and from the records, we have no proof that the loss in the number of grains which did RELATIONS OF BURSARIA TO FOOD 30 occur was not, in part at least, due to complete digestion and resorption of the missing grains. But this difficulty ‘is almost wholly avoided in this experiment, because the process of diges- tion and complete resorption of fresh yolk is much slower than that of vitellin alone. This is mainly due to the fact that the fat of the fresh yolk is digested and resorbed at a slower rate than the vitellin of the yolk grain (cf. plate 1, figs. H—L, and figs. H’-L’). This fact is further shown in the records of the total numbers of grains present in Set A, table 7, for twenty individuals out of forty-eight, still contained remains of the one grain fed tharty hours previously, this should be compared with the average time for digestion of one grain of vitellin in Experiments III and IV, table 5, column 2. Now, from these facts regarding the difference in rates of digestion of one grain of vitellin and fresh yolk, it is not to be expected that the individuals of Sets B and C would show as rapid a rate of digestion as if they had been fed an equal number of grains of vitellin; and therefore it becomes clear that if the comparatively small total number of grains pres- TABLE 7 Summary of results of Experiment VIII (January 24, 1913) showing the effect of the quantity of fresh yolk eaten, upon the course of the extrusion reaction in Bursarva. Three sets of forty-eight individuals each were used. Each individual in the Sets A, Band C, was fed 1, 3 and 6 grains, respectively | HOURS AFTER FEEDING 24 30 Ohad lace Zena | te tae tow * < — = Sa $ 7; 5 al | > | {| Total number of grains) | | | ee 2: || present........... .. | 48 | 47| 46 | 34 | a2 | a2 | 32 | 30 | 28 | 20 ss 4 Per cent present of to- | | viduals fed | Palla Penh aaa | | | | 1 grain each | Ce SPAS | | UI isd somite age? « (100 97.9+ 95.8+ 70.8+66.6-+66.6+60.6+62.5+ 56.3 41.6+ | ‘ i ae wad ies == l Set B. Forty-{ | Total number of grains | | | | eight indi-]| present.............. 144 | 137) 97 | 44 | 44 | 43 | 31 | 29 | 27 19 viduals sal Per cent present of total | 3grainseach|| numberof grainsfed. 100 95.1+ 67.34-(80.5-+)30.5-+ 29.8 21.5 10.14 18.7+ Ae | 49 | 30 26 17 10 Set C. Forty- {| Total number of grains | eight indi-|| present.............. | 271/219 | 165 | 98 | viduals fed {| Per cent present of total | | | many grains number present at | | | | i each a three hours...2.5...- 100 81.14 00.8+ 30.1++|18.0 | 11.0} 9.5 6:2=F) (30 36 E. J. LUND HRS. Fig. 8 Curves plotted from percentages in table 7, Experiment VIII, showing the effect of mass or volume of fresh yolk eaten, upon the extrusion reaction. Curves A, B and C are plotted from Sets A, B and C, respectively. The points on the ordinates represent the percentage of total number of grains fed, which still remained in the vacuoles at the time indicated by the same points on the abscissa. RELATIONS OF BURSARIA TO FOOD 30 ent, towards the end of the experiment, in the individuals of Sets B and C, was caused by rapid and complete digestion, then the effect of mass on the rate of digestion for fresh yolk would have to be relatively vastly more pronounced with fresh yolk grains than is true for vitellin; and this is not the case. The first parts of the curves (fig. 8) bring out the pointsclearly, for here the error, that would be caused by the loss of grains due to complete digestion and not to extrusion, is practically avoided, since the rapid fall in the number of grains takes place during the first part of the experiment, before the digestive process could have been effective in causing total disappearance of the yolk grains; especially is this true for Set C, Curve C. On the other hand, Curve A which we should expect to be markedly affected by digestion, does in fact show only a slight fall, compared to that of Curves B and C. It will be noted that the amounts of fall of the curves show the same sequence as the number of grains fed. The, results therefore actually demonstrate beyond question, that the quantity of fresh yolk eaten is a determining factor in the process of extrusion. The greater the mass or volume the more effective is the stimulus from the contents in the vacuole. A quantitative or intensity factor as well as a qualitative factor therefore enters and determines whether or not extrusion of fresh yolk shall take place. Now from the results of Experiment VIII we have as yet no evidence showing in what way the mass or volume affects the process; 1.e., whether it is an intensity effect due to chemical or mechanical stimulus or both. Another important fact in connection with the extrusion re- action is that the stimulus from the contents of the vacuole is most effective in bringing about extrusion during a rather limited period (4 to 6 hours with fresh yolk grains) after the formation of the vacuole. This will be seen from the sudden drop in the curves in figure 8, with subsequent tendency to retention of the remains of the yolk. There is a process of functional adjust- ment (loss of irritability?) with the continued action of the stimu- lus from the contents of the vacuole. Roughly it may be ex- 38 E. J.. LUND pressed by saying that a reaction by digestion is substituted for one by extrusion. “These facts appear in.a clearer way when the course of extrusion is studied in single individuals. A further analysis of these responses must be kept for a future time, since the data from other experiments in which were used other substances are not sufficiently complete. SUMMARY 1. A quantitative method (p. 3) was worked out, by means of which it was possible to study the processes in the food vacuole of Bursaria. 2. Yolk and vitellin may be drawn upon as food for the energy requirements and growth of Bursaria. 3. The liquid of the newly formed food vacuole is partly made up of the external medium, and partly of an acid secreted by the base of the buccal pouch. After a few minutes this liquid is resorbed and the vacuole membrane becomes applied to the yolk grain. The vacuole contents remain acid in reaction throughout the process of digestion of vitellin and yolk grains. 4. Sooner or later after the initial resorption of liquid about the grain, digestion begins. Digestion may or may not result in the second appearance of liquid in the vacuole, according to the principle that whenever the rate of solution—this perhaps in part depending upon the concentration of the cleavage agent— is greater than the rate of resorption, then the liquid products of digestion accumulate more or less about the grain, while if the rate of solution of the grain is slower than the rate of resorp- tion, then the products of digestion are removed as fast as they are formed. Equilibrium between these processes in the vacuole may be established during digestion of vitellin with much, little, or no liquid present in the vacuole. 5. The average time for complete digestion of vitellen in Bur- saria was found to be directly proportional to the square root of the quantity of vitellin eaten, i.e., the relation expressed by Arrhenius’ formula t=kV y was found to hold to within the limits of experimental error. ‘ RELATIONS OF BURSARIA TO FOOD 39 6. Congo red adsorbed by vitellin grains and fed to Bursaria interferes with or prevents digestion of the parts of the vitellin grain to which the dye has been adsorbed and causes an early extrusion. 7. Olein is digested and resorbed by Bursaria while paraffin oil is not affected. Lipoids and fats play an important rdéle in promoting growth in Bursaria. No evidence was obtained for the formation of stainable lipoid from pure vitellin. Starch or amylum grains are not digested. 8. The time of extrusion is determined by the quality (chemical) and the quantity or intensity (chemical, physical or both) of the stimulus from within the vacuole by the substance eaten. 9. The maximum tendency.to respond by extrusion to the stimulus from the vacuole contents, exists within a limited time (4 to 6 hours with fresh yolk) after feeding. LITERATURE CITED ARRHENIUS, 8. 1909 Die Gesetze der Verdauung und Resorption. Zeitschr. f. Physiol. Chem., Bd. 63, pp. 323-377. GREENWOOD, M. 1894 On the constitution and mode of formation of ‘food vacuoles’ in Infusoria as illustrated by the history of the processes of digestion in Carchesium polypinum. Phil. Trans. Roy. Soc., London, vol. 185-B, pp. 355. Lunn, E. J. 1914 The relations of Bursaria to food. I. Selection in feeding and in extrusion. Jour. Exp. Zoél., vol. 16, pp. 1-52. NIRENSTEIN, E 1915 Beitrige zur Ernahrungs-physiologie der Protisten. Zeits. f. allgem. Physiol., Bd. 5, pp. 485-510. 1910 . Uber Fettverdauung und Fettspeicherung bei Infusorien. Zeits. f. allgem. Physiol., Bd. 10, pp. 137-149. Stanrewicz, M. W. 1910 Etudes expérimentales sur la digestion de la graisse dans les Infusoires ciliés. Bull. Intern. Acad. des Sc. de Cracovie, Série B, pp. 199-215. PLATE 1 EXPLANATION OF FIGURES Typical stages in the process of digestion and resorption in a food vacuole of Bursaria containing a single grain of fresh yolk or vitellin. A Degree of acidity attained by the vacuole as shown by litmus at time of separation from the end of the gullet. The pink edges indicate that acid has been secreted from wall of gullet; figures B and C, further stages in same process. A to E, inclusive, show course of resorption of liquid enclosed during formation of vacuole. G to L, inclusive, typical stages of digestion and resorption of fresh yolk (protein and lipoid). G and H’ to L’, inclusive, typical stages in digestion and resorption of vitellin- liquid present in the vacuole. G and H” to L’’, the same as H’ to L’ except that products of digestion are resorbed as soon as they are formed. Change indicated by figures under z lasted on the average 38 seconds; change shown by figures under y about 43 to 6 minutes. 40 RELATIONS OF BURSARIA TO FOOD PLATE 1 BE. J. LUND oO ae) y Y ite ro =) A a “ ES w x ¥ Nae COO SOSEO 4] PLATE 2 EXPLANATION OF FIGURES A and B_ Low power, camera lucida drawings of two sister cells stained to show fat 42 hours after separation. Cell A was not fed; cell B was fed olive oil 18 hours after separation from sister cell. The large drop of oil in B represents that re- maining in the vacuole. a and b, high power, camera lucida drawings in one focal plane of area indicated by squares in A and B, respectively. 42 RELATIONS OF BURSARIA TO FQOD PLATE 2 E. J. LUND 43 4 Loe aaa 9 ve oe a x ms i . i ¥. ‘ . WE, ‘ rr. - . ’ ’ é i* ‘ 8 veiw 5 oe ~ , 7 * . . 3 ne . . i . ? . ‘ . . + ‘ an) Lo a « q bh 8: 2 . a ry "* ¥ y ow bg. CHROMOSOME STUDIES IN THE DIPTERA J. A PRELIMINARY SURVEY OF FIVE DIFFERENT TYPES OF CHROMO- SOME GROUPS IN THE GENUS DROSOPHILA CHARLES W. METZ *+ From the Department of Zoédlogy, Columbia University DIAGRAM AND TWENTY-SIX FIGURES (PLATE) To the student of chromosomes and chromosome behavior a cytological study of the Diptera presents many features of inter- est. Some of these have been described or mentioned by Miss N. M. Stevens in connection with her work on the sex chromo- somes,! but aside from this I know of no serious attempts at such a study. When compared with the detailed and critical works on the chromosomes of the Hemiptera, Orthoptera, and Coleoptera, this lack of knowledge of the Diptera is surprising, and can be explained only by the fact that the latter are in many ways un- satisfactory objects for cytological study. The difficulties in such a study are admittedly numerous, but notwithstanding, much can be accomplished, and the interest of the questions in- volved seems to me to more than justify the extra effort expended in their investigation. For this reason a series of such studies has been undertaken, the extent of which will depend upon the time and facilities available. Some are under way at present and others will be taken up subse- quently. The results I hope to present in a series of papers, to which the present is introductory. 1So far as known to the writer, the papers of Dr. Stevens are the only ones dealing with the chromosomes of the Diptera. They are four in number: The chromosomes of Drosophila ampelophila. 1907. Proc. VII Internat. Zodl. Cong. eee 131 45 34.0 Lethalamimia tures: eee eee 814 323 40.0 White mimiatures.<7 cc eee ene ee 994 397 40.0 SEX-LINKED LETHALS IN DROSOPHILA 97 THE SECOND LETHAL FACTOR In an experiment with certain stock which had been inbred for three years, a pair produced: 73 females and 16 males (= 5:1) These numbers represent the total output of this pair, or at least all the flies that were produced from one bottle. (A) Twenty-two of the seventy-eight females were mated to white miniature (of which two pairs produced nothing). (B) Twenty- nine of the seventy-eight females were mated to eosin vermilion males (of which six pairs produced nothing). (C) Twenty- four females of the seventy-eight were mated to eosin miniature males (of which four pairs produced nothing). The experi- ments with white and with eosin should give similar results, since white (w) and eosin (w®) are allelomorphs. The sixteen males were mated to eosin miniature, white miniature, and eosin vermilion females, but every one of these males proved sterile. The output of the three lots of females mated to the respective males mentioned above is shown in A, B and C of table 10. Under each column, A, B, C, the progeny is provisionally classi- fied, first, into those that give apEren Ea a 1:1 ratio, and then those that gave approximately a 2:1 ratio (or higher). A horizontal line separates these two classes. The classification into these two groups may appear arbitrary, since the ratios in the two classes come very near to each other in some cases. For the present it will suffice to explain that the 2:1 ratio is expected if one lethal is present, the 1:1 ratio if no lethal is present. From table 10 one can get no clue as to the location of the lethal factor that is assumed to be upsetting the normal sex- ratio, but by breeding the females to a male that carries the same sex-linked factors as their fathers, the location of the lethal factor is revealed. 98 T. H. MORGAN Crosses with white miniature males A number of daughters were taken from No. 12 (indicated by a star in table 10) and bred in pairs to white miniature males. Since No. 12 was carried further, I have selected it first for illustration. The data are arranged in table 11, under two headings, viz., those pairs that gave a sex-ratio of 1:1, and those that gave 2:1. The data from pairs with a 1:1 ratio show nothing except the association of white and miniature. The data from the 2:1 ratio, in which the second lethal is present, show no males in the TABLE 10 A B Cc : g ey) ig (ies ¢ 2 | alow Bi heen a ee a 20) Jaa aa uf 11 | 60% 44a a ee One 8 Soy CGnimeot | 1.3 VIL | 53891) 44) 1.25 a Te selec Bilin BOr ql made cl) 128 10a 8 9) ee |) AS 9 aoe ate G64 ab tet eS as ‘San a 54 | 53) | ate Say Vian ee . | ales TAH 63 il, G40 || wal m |. 24 | 20 hates 9 | 69 | 49 | 1.4 TE | 358, 0H) 5G aad Dp | 67. or aie IS 55 Gl Sd. 1) 084i" ey a ey levee ih oe ieee eke | Baert Veda 1G I: gilts sli el 0 tes GV) BAS alee 8, 65 |.-47' | aaa DZ) GA ede | 8 XV Ova aaa ole de \t1 calc 1S 8) 1 068.0|" Oa eV I Ost 83. eT este 5S. ah” 8a" |) seed | DE 670 56) | 12a a e794) Seoimn eae PAGO T i) 48 ES XK | 4d 88 hy eS be®s 6 | 46, Aisa ATMO 35, eae SOD BY | BRIS Aa 2 ky Meo) ea eames Cee AN OA NSS ONERATINT |) ROI | 40h ele ee NGS one an ne TOR S938: ily 155 saaieail| eaenaealll imine 53 ci + 19) hoe 11 Geez, eae 119 T | 79 Ny 8% eral hes ale G2: eso maren ig |p EP Ne SRS} LOOM ee Pee) te Oye 80 | 31 2.8 APG da | 268i aeeeo TV) 64:71 926-9) 204) ni? 58 20 ees NS Sa avec he Mes: a i BA Vil TA ST 288. oo: Mae) | 2Senn amare! 19 Sb asehi eee Vii 16 5S SAE) Gg a | Bly Cie 20) || «BS el WAS GCA VOY Dba GSei eta Sen aaondiely bat 10 3p oe8 RIT |< 531) eee | VAL | YOGI 30a one DERI) (20) A ae SEX-LINKED LETHALS IN DROSOPHILA 99 red-long class, which is the lethal class and, as will appear, also the double cross-over class. One of the single cross-over classes is red miniature; the other, white long. Both are smaller than the non-cross-over class, white miniature. TABLE 11 (From No. 12 of table 10) | ¢ ee cptetres 52 8 ae pare fe a Ro: eee ese eM ag. OL SRS a Ge INO Bte les Gen RENCE nate eee Wes Re Ae eae 1 i aay oe NP SSS aha 28 © ead 87 he 20 led 3 eeeO itl 12 chew 5 Sees Ally 10y [wk cleats 7 7 DCE PO Tig i Gagne 0h) ay i oS ee: 8 osmogl se Ne TS an) ied 28. The30". | Or Gees 9 Soll | 2Gua Id) ie Se ie 22n lee27>!, | ISNA) 2S oie 13 Set See elt Nt ale Ae OO lk 2 > Ate ME TeRi a I) wal A 15 25 Paget sient) UontGalk 1S) 0 280! Tse aG a ieee 18 27 | 26 | 14 | 16 | 38 | 33 |, 13 | 20 8 19 DA S260 OI, 10 sh BF) Ae 21 SO i G2ep i oelh Se FID. eebrohy 10...) 6: ales 22 aa 251 PSP Ge 2a 08 | 9. | 13 i 26 | 26.92 250 ota 22 1 O20) 22 S| IES 29 ee ea ey ee er ae | Eee Total | 392 | 344 | 162 | 163 | 245 | 332 | 159 | 132 2 BO lags 155 tO! SOKO ee 3, ) 14 Wieoen 4* OS ae” 16 late 0 | 19 3 Til 239 5* AG 5 tly AA N20 lo? AO) Vek 6 | 25 6 Dita FO 12 Oo 9 ase aS aee 11 kee 16s 2 100.) 1a (Pa ee oa mee 2) Fase Ps 12 ier He2e en alse | a2) Or pat 5 8.0 (2M 14 Pe NOs bee. OC AO OG" ae aerye O°.) alte 16* ae te 8. | Ole On det. a.) CO -es 17 ae lag). 98 | ia | 0 | or 3 1) 61). ees 20 Depa ade i leh sO) St 2 | -0:, | aaa 23 38 | 34 | 18 | 19 0/535 3 lent: || 24 Bf) 24 Te io 0) |, 3h 5 7. ete 25 2th | 31, oh ES tO 0", 19 2) |. - 2) lees 27 De 27 cea UW eel Os STZ ae Sy A eed 28 I es a Va 2 i= Us ee | 0 30 Bole, |e 20) 7 0 | 33 5) 6 eee 31 Si eon re tay | ade Ot aer k aade eG) a / aes Se Btn = = | | | 0 | 436 | 54 | 104 | . 100 T. H. MORGAN : In each generation the females that gave the highest sex- ratio were selected to breed from. Whatever disadvantage this method of selection may have, it at least makes more probable the retention of a 2:1 ratio. TABLE 12 (From No. 16 of table 11) Pieanomteoeslok. . No. ae |} Be | ce) BE) 08 | Bede cee ee ) 2° EF a ES ge BA a HS ae a 146 495 | to 494 | 0 i 30) 7 1) aoc) iaee b* 532-32 1). Ber 19 0 | 36 9 1 |) oe c* 17 a7 "49, 908) ag 0 | 26 6 4 |e d a7 Say a ety ee ati @ i. 29 3 5) eg e Ban ee 9 ee Seay a) 0 | 29 2 4 | S38 f Loew! 3004 20 Ge 15 0 | 20 2 9. |. asae g 55-| 49 | 20 | 28 0 | 66 9 |) 14) 9) ae h 7 ae; 3 0 | 40 6 1 || Some i 20 | 20 Geile ao Onn a 4 A 117586 Total 367 | 327 | 164 | 159 0 |290 | 48 | 45 | | | TABLE 13 (From No. b of table 12) Ges fe | | % a) (ee ear eee en es No. Wee Cod BEE PY x Bas treeilN) 4.06 a gene eam nets < Cleese ee lc se | 1 Be 98 eG. eats 0 | 19 6 | 3 ee 2 35 Wh 530,. 4 alo ets 0 | 36 9 | at) joe 3 39) |. 30 sod RIG al) LOro aon AM 2 aa 4 Wetec es Pass ol eee a ee | 6. |. 2 ae 5 |} 36 | 44 | 21 Gy 1, 20" | 228 4 Oo | ag 6 be36 nse | c16 Sl, Oe 1 988 5 1 igo 7 AA 9G «| Sener a. O71. 35 3 | Sores 8 46°) 49 | 25°] 25 NO) 44 Ce peo 9 | 43 | 34 9 6 0 | $35 eet 1 es: 10 Bigot a5. 41" Sa 6 Oi. || Kok) Hee ol) 3a 11 35h. 3801) 15 laws 0 |) 45 | 4 1 {oan 12 908 1 87 | 27 G90 0 | 2 | 10 | 10 | 25 Total | 467 | 410 | 230 | 162 | 0 | 304 | 51 | 47 | 2.6 | | | SEX-LINKED LETHALS IN DROSOPHILA 101 From lot 16 of table 11 several virgin long-red females were mated in pairs to white miniature males. Their offspring are shown in table 12. Each pair shows in its offspring that the second lethal was present in the long winged red mother. A further generation was reared in pairs from red females of lots b and ¢ of table 12. The offspring recorded in tables 13 and 14 show that the 2:1 sex-ratio continues. Returning to table 11 there are records of two lots taken respectively from No. 4 and No. 5. The offspring from these pairs are here shown in tables 15a and 15b. Both show that lethal II is present as indicated by the absence (or rarity) of the red long males. Before examining the preceding tables 10 to 15 in detail it may be helpful to have in mind certain theoretical relations. The location of factors for white (w) and for miniature (m) have been already determined. If we are dealing here with a lethal factor, it is essential to determine its position in relation to white and to miniature. For reasons that will appear later it is here assumed that this lethal lies between white and minia- ture (fig. C). TABLE 14 (From No. c of table 12) o Oo | Bl ee D D % | D Oi sean é ae Srl eee | eee g a Sr eGo AiGeli mete: [al6. wit cet enely ee Ge llaae el PI SA bee lee hes REO ee me ee 1 27 uy 27 \a A | 26 0 | 39 Bean was | «28 2 26 | 35 6 |215 Oh e238 Wo Sb. |. 2a 3 bao ( -59-" | 730, (7 932 0 | 56 7 2 | oe 4 61 | 49 | 39 | 13 0 | 44 7 31 toe 5 40-9981 16> | V16\='|.~ 0) 1-30 D 5 || aaa 6 29 | 35 7 eats 0 | 26 3 7. |e 7 praG, 4 le 20) «|g 0°| 50 6 9. |Wg3 8 SA LU EM ls O2 |F 32 1) £0 7 | ears 9 BAG Sie ec ORC net 0 | 34 6 4 oli aae 11 | 59 | 55 | 27 | 25 On) 67 8 7) Wee 12 Pete 10s), anette <0.) 28 3 1° ieies 13 Peale hots en | te oO, | Se 5 3/725 Total 479 | 450 | 227 | 229 O,>|.455° | 69. | 56. |e oe THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 17, NO. 1 102 T. H. MORGAN : We are concerned with the heterozygous females only. Of the two sex chromosomes of these females, the one of maternal origin is indicated by the heavier line in the diagram (fig. C), the other, of paternal origin, is indicated by the lighter line in the figure. In the present case there entered from the maternal side the three factors: normal (red) eye (W), the second lethal (1,.), and normal wings (M); there entered from the paternal side the three factors: white eye (w), the normal allelomorph of lethal (L.), and miniature wing (m). In the case of this lethal the possibilities of interchange of factors between these chromosomes are shown in figure C, at b, c, d. If the crossing-over takes place between white and lethal, the case is indicated in (b). The two gametes that re- sult from this single cross-over are WL.m and wl,M; the latter never becomes realized in the male because he carries the lethal (1,). Similarly, if the crossing over takes place between lethal and miniature the two gametes that are produced are W |. m and w L: M, of which the former carries the lethal factor. There is another possibility, viz., the double crossing-over (d). TABLE lia (From No. 4 of table 2) oa RED WHITE | RED | WHITE RED WHITE | RED | WHITE | LONG 9 MIN. 9* | MIN. 2 | LONG 2 LONG MIN. o& | MIN. Go’ LONG o ms ae ry: 4 5 é a 26 22 14 | 12 0 23 3 | 6 b 23 19>) tt] 13 1 12 3 8 c 34 39 21 18 0 28 10 8 d 30 24 12 16 2 15 5 9 Total 113 104 | 58 59 3 78 | 21 3l TABLE 15b (From No. 5 of table 2) RED WHITE | rep | ware | rep «| ware | Rep | ware NO. | LONG °) MIN. 9 | MIN. 9 LONG Q | LONG MIN. | MIN. o LONG b | Se | ila | Ga Sy ile 0 te JY oe if CAPA 34 307+} 8 So Ol ae els 6 Total | 50 AGE ae id 1° tok Og ae 6 wie SEX-LINKED LETHALS IN DROSOPHILA 103 w 1, se W1.M \s WLiem w Ee m Wi I, M wl.M 7 Ti aa W Ls m Ww lL, M WE TS a e34 31 12 12 22 39 L5ie | al eM eS 45 19 15 36 42 11 5 eg | 47 49 22 19 35 41 12 7 Total 428 407 168 166 321 420 | 120 | 78 c 61 66 26 30 0 49 rae Cee): il89 49 17 12 0 34 3 7 q 34 34 15 16 0 45 2 Sir ie s 44 45 14 10 0 49 5 ang u 48 39 18 14 1 41 5 5 w 44 29 17 10 1 44 3 9 y BL i 38 7 17, 0 59 4 it aa 47 39, || 6 11 0 AT egies 3 bh 45 Bi uN uc 2 15 0 40 10.5 |) tt ij 58 37 13 34 3 29 9 | 12 ll 28 31 12 12 Oy ae 4 2 Total | 499 425 167) {east 5 464 48 109 SEX-LINKED LETHALS IN DROSOPHILA 105 From the lot 77 (table 16) nine red-eyed females were bred to eosin vermilion males. The results show (in table 17) that the second lethal was present in the red-eyed females. In the lower part of the same table there are four cases recorded in which only vermilion and eosin vermilion offspring were produced. The sex-ratio 2:1 recurs. The results are undoubtedly due to four vermilion females (1, 5, 10, 14) from jj of table 10 having been taken instead of red females as intended. From lot c of table 16 also seven red-eyed females were bred to eosin vermilion nfales. The results are given in table 18, which is similar to the preceding tables. The sum of all the female classes that belong to the 2:1 divi- sion in tables 2 to 6 b and the sum of the corresponding males, is as follows: females red long white min. red min. white long 1964 1835 916 852 males 3 1690 247 290 The linkage shown in these results will be discussed later. TABLE 17 (From No jj of table 16) NO. RED Q manne VERM. 9 | EOSIN 2 RED ol este 3 Eat poy 2 39 39 8 | 11 0 OF tia 9 3 34 23 4 8 0 24 2 4 4 32 33 Ay, ae 0 14 2 | 16 6 28 36 Sec eo 0 19 4 9 8 24 12 Gey as 0 26 2 3 9 27 16 ll 10 0 25 8 7 12 44 29 23 10 0 30 7 8 13 32 39 14 10 0 32 2 5 15 18 Darryn) it 3 0 26 2 7 Total | 278 250 | 102 | 88 On) |) 218 33) || 968 1 40 36 | | 39 1 5 32 30 32 4 10 S0n il.) Al 24 8 14 a1 he | 3 31 2 Total tz Vi 1387) | 126 15 106 T. H. MORGAN TABLE 18 (From No. ¢ of table 16) NO. RED 9 vase VERM. 9 EOSIN Q RED Basa) ot Ass ae 1 47 25 19 14 0 34 5 10 2 53 42 13 9 0 38 8 an 3 40 35 18 17 0 28 8 11 4 48 28 13 16 1 36 7 13 6 28 31 13 10 0 By 4 2 8 19 14 16 6 0 26 3 4 9 29 36 15 12 0 25 8 5 Total 264 211 107 84 1 219 43 50 TABLE 19 (From No. e table 10) £ RED EOSIN RED EOSIN RED EOSIN RED | EOSIN aC LONG @ MIN. 2 MIN. 2 LONG 9 LONG MIN. ch’ | MIN. co’ LONG 6 35 29 16 14 31 28 11 7 8 24 22 1 14 22 19 6 10 9 19 24 12 11 25 26 10 15 10 28 20 10 12 31 24 NZ 11 14 22 15 10 6 16 743) 11 10 16 20 12 9 9 Pa 15 12 14 17 11 ul 9 4 8 12 12 1 18 24 29 12 12 20 16 17 9 19 33 21 8 11 18 26 14 14 Pei 22 22 8 12 25 23 8 BS) 24 45 38 il7¢ 21 28 23 19 13 26 1 13 6 4 5 8 6 8 27 18 26 13 10 19 Pe 12 © Total 313 282 142 140 275 265 146 127 2 25 28 16 7 2 21 5 8 3 27 23 8 2 0 23 2 he. 4 39 30 17 12 0 24 it; 8 7 26 22 13 18 0 25 5 8 11 33 24 12 18 0 29 5 10 1 33 16 17 22 0 19 7 9 20 33 31 14 16 0 27 4 By os 23 41 By 18 23 0 23 9 13 25 26 24 15 ial 0 29 1 1? Total 283 230 130 131 2 220 45 78 SEX-LINKED LETHALS IN DROSOPHILA 107 Second lethal and eosin The females of lot e, table 10, which gave a ratio of 116:46, were bred to eosin miniature males. As shown in table 19, fourteen of the females gave a 1:1 ratio, and nine a 2:1 (or higher) ratio. In the latter (2:1) class the double cross-over is the red long male. Two of these that appear must be due to a double cross-over. It is just possible, of course, that contamination may have occurred because these two males appeared in the same bottle and on the same day. a i ls wM, = 4 WimL3 Ww mH Li, WwW M LS) Wels x byMl, we m iis Ww iM ls _\ wh, = x CWinls Ww m 3 WwW M ie Wrls x x d WML We mee 3 3 Figure D The females of another lot, in table 10, viz., lot n with a ratio of 58:20, were also bred to miniature eosin males. The results shown in table 20 are strikingly different from any others so far ob- tained. The second lethal no longer appears to be present, yet in about half the cases the ratio is near 2:1. The question arises whether a new lethal has appeared, or whether we have in this set simply separated from the second lethal the other factor or factors that when combined with the second lethal gave the high 3:1 ratios. No more light is thrown on this matter by the further history of this series. One lot only, viz., No. 39, seemed to give a 3:1 108 T. H. MORGAN TABLE 20 (From No. n of table 10) oF OF rors oF ‘oD Ley % > = NO. 2 | zz z| 28 2 | Ze z | Ze z SSeS | Bs Ba ee awe | era aS i e a e a es re) re) be 1 56 39 32 31 46 53 32 12) alee 2 39 31 14 12 22 21 25 6 | 1.4 3 45 37 30 25 27 41 28 12 "| is 4 71 76 30 30 41 42 20 23 | 426 5 56 56 28 33 16 20 20 11° || 2% 6 56 50 34 19 24 24 18 10 | 2.1 7 54 42 30 25 26 32 20 3 /1.9 8 ti 28 30 31 40 39 23 27 | 0.95 9 43 40 19 21 24 27 13 18:5 10 42 34 28 27 22 25 25 5: es 11 54 39 32 29 19 36 19 13. | 28 12 56 41 35 26 63 39 23 24 | 1.0 13 61 52 30 33 34 52 27 9 |}1.4 14 61 67 36 35 39 63 25 19 | 1.4 15 43 51 25 39 35 25 24 12) As8 16 45 36 31 20 36 25 28 8. be 17 59 65 26 28 55 61 26 29 | 1.0 18 48 24 31 19 29 26 21 12. eS 19 72 61 29 26 31 29 40 1 ey 6 20 37 36 25 22 14 23 18 5 | 2.0 21 46 28 26 27 21 26 25 13 | 1.4 22 51 48 33 24 31 30 20 Te Er 23 51 33 26 20 28 29 21 16 “srs 24 40 37 32 19 40 24 19 64s 26 56 45 28 28 22 27 18 12 6°20 27 65 45 30 33 31 35 31 1S: else 28 48 50 28 28 21 37 24 tS bl 29 50 25 16 10 27 13 13 9 1.156 30 33 33 25 12 11 38 17 5 | 1.6 31 38 34 30 15 21 24 23 rere: 32 32 42 17 22 19 35 15 10: higa 33 38 32 22 31 20 25 18 5 iss 34 30 41 13 18 28 17, 18 21 it 35 29 32 1Sae aS 30 21 13 a, Waleed 36 39 20 9 9 24 20 9 8 is 37 38 30 30 20 20 35 17 10 W415 38 27 23 22 9 13 10 7 2 | 2.5 39* 49 39 16 13 22 1 0 10 | 3.3 40 33 31 5 10 21 15 11 re Be 41 20 18 18 if 11 14 7 0 1.2 42 19 21 11 6 13 10 9 5 | 1.5 Ua reso teeanarage 1874 {1612 |1025 | 918 1117 /1189 | 809 | 454 SEX-LINKED LETHALS IN DROSOPHILA 109 TABLE 21 (From No. n of table 10) Ot oH oe OF sie) Lis; 2 NO g Zz | z 2% Z me C z& 3 Be Sanaa Sales | ean) Bees wine % a { & =| P] a % a D a fe2evieat iy | ie! | 34 (|. 4%. 1. eh | aes b [ae a6 19: |) I Sh. i) Sa | or atte ¢ fates 05 Te EAE: 5 a te at d | 12 7 6 oh 3 | 6 6 Or hier e 2a }2d) | 18. | 12 28 By 1a ts aos f 5 6 6 3 5 9 6 3 |1:0.9 g | Gaal ee 5 9 5 4 5 OM ene eig9e 1351-65. | ‘72°W 2: | tap" | 66 |) ae ratio and from this lot a few females were bred to eosin ver- milion males. Their offspring are recorded in table 21. The data appear to verify the conclusion that the second lethal has gone while a disturbance of the sex-ratio remains. Can these results be interpreted to mean that the second lethal disappeared, and a third lethal, L;, that was present and gave the original high ratio of 78-16, has been retained? If such a third lethal is here affecting the sex-ratio and if it lies beyond (to the right) of miniature as shown in figure F, then the deficient class is eosin long males. “If the lethal kills then the only representative of that class will be double cross- overs and should be infrequent in proportion as L; is close to M. In fact, in a large number of cases in the table, the 1; class, is behind the other male classes. In order, however, for the double cross-over class, eosin long, to be as frequent as appears in the table, the distance of L; to M would have to be very long indeed. In order to calculate this distance it would first be necessary to find out from table 20 those classes in which the factor occurs, but it is impossible to do this, for there are too many cases whose position is uncertain. If we make a single attempt to pick out such cases on the basis of the deficiency in the eosin long class we find that the most frequent class is that with a ratio of 1.5:1 and not 2:1 or more as would be the case if a lethal like the other two lethals is here present. 110 T. H. MORGAN The meaning of the high sex-ratios that appeared in these cases must remain therefore unknown, and it is idle to speculate whether the results are or are not due to a non-sex-linked lethal, a sex-linked partial lethal, a sex-limited factor or factors, etc. The importance of following up this case was not appreciated at the time and this line was not continued except to breed seven’ individuals from No. 39° (table 21). Data from stock of Lethal II The stock that carries Lethal 1r has been maintained by breed- ing long red females to white or to eosin miniature males from the regular eosin miniature stock. This has been continued to the present time. The results are shown in the next table, where, sald Vv x xX w° Le ¥ Figure E through ten successive generations (bred in pairs), the inherit- ance of a 2:1 ratio has maintained itself. The horizontal lines in the first column separate generations; in some cases two or even three lines were tested, in each generation, in other cases only one. In all there are 13,749 flies recorded. The sex-ratio is 2.3:1. The order of the factors is shown in figure E. The percentage of cross-overs between white and lethal is 8.8, that between lethal and miniature is 16.3; and that between white and miniature is 24.7 which agree sufficiently with the per- centages already given. The females are also available for cross-overs between white and miniature, and these give 28.3. The linkage of Lethal II with white (or eosin) and with miniature The diagram (fig. C) on page 103, will serve to recall the order in which the three pairs of factors enter. The F, female has received from her mother a sex-chromosome, bearing the factor for red (W) Lethal 1 (12) and long wings (M); and a sex- 5 No. 39 is peculiar in that eosin miniature and red miniature males are too few in numbers. SBX-LINKED LETHALS IN DROSOPHILA 10g chromosome from her father, bearing the factors for eosin (w®) normal or non-lethal (L:) and miniature wings (m). The possible combination in the eggs of this F; female are repeated here: W 1M —red lethal long w* L m—eosin miniature W L m—red miniature w° 1 M —eosin lethal long INOW=CTOSS-OVEI eee ech te stele bc Goole eee { Single cross-over ¥ 1m —red lethal miniature w° L M—eosin long : L M—red long Dey pe eMOSS-OVEE o)..5 « soraeie,« ain gies, 2 = a tee 2 : ahs w° 1m —eosin lethal miniature TABLE 22 (Stock Lethal 11) | RED LONG WH MIN. RED. MIN. | WH. LONG | RED LONG WH. MIN. REDS es 9 9 9 | 9 a oF MIN. ee A 428 381 161 125 0 362 31 | 21 B 365 337 166 175 1 276 | 45 | 60. ¢ 509 502 229 232 0 368 65 | 74 D 316 329 95 115 0 305 | 26 | 63 E 318 289 104 88 1 278 31 | 962 F 312 259 95 81 0 259 19 | 61 G 7 Ste “105 24 SN AG 94 3 | 26 H 124 104 GSW epaueae ie 105 14 | 36 I 118 114 5Oe | = Abe) eal 117 17 333 J* 158 119 64 | 53 | 0 111 10 | 35 K 8 15 1S 219 1 i | wl oe i 96 107 4 | 48 1 06° “Are ez M 119 104 53) ||| 40 0 137-1 A) ees N 28 35 in |) 46 1 37 (flat O 24 14 gn) 8 0 28 4 4 P 242 O41 OWI toon | 909. | a1 | 53 Q 271 302 Ber dee al a 338 | 31 | 63 3509 3357 1388 | 1320 | 9 3135 | 359 | 672 * When Generation I was reached some of the red long daughters were bred to eosin miniature males (instead of using white miniature as before (A-I) under similar circumstances). Their offspring were as follows: Red long. Wh.min. Eos. min. Wh.long Red min. _ Eos. long 2 of of 2 of of 2 of 2 of 277 0 227 261 11 57 107 28 89 1 The long red females were again selected and bred to eosin miniature males to give generation J, K, L. Thus one generation is omitted in table 22, and from the summaries which include table 22. 112 T. H. MORGAN Of the eight possible combinations, four carry lethal and will not be represented in the male classes of the next generation frona the back-cross. The four surviving male classes represent: (1) A non-cross-over class (eosin miniature) (2) A single cross-over class (red miniature) (3) A single cross-over class (eosin long) (4) A double cross-over class (red long) W L M LE a Se a Pyar : ce m ja Ww l, M a Bx b w& LS m L, W L M I, x c ee Le m Ls Ww ly M lg Di. Least eee ee a ME a ke Preyer D Te W I, M lL, x X i w & Ly m oe WwW ly M ly : x ome w 8 a3 ae L, Ww l, MM l, x x 3 we ER m L, WwW l, M l, : SRR; A ie Gz A st aL > A OS w? Le dg ae Figure F SEX-LINKED LETHALS IN DROSOPHILA 113 TABLE 23 NO. | EOS, MIN. RED MIN. | EOS. LONG | RED LONG | Bes | ye EOS. MIN SIR idee 61 54 | (104 0 O.1° | Wye «le 26.6 Il, 200 48 | 45 0 12.5,>|,. ol Deaeee4 <2 nahh” A) Cue er ie ee ee 10.4) || 4 eeminan20:0 See isos) | es | be | O |. 2°) Se aaa OA MT cca eee | a ae eG Ve a Oa 0 | | ER 290 | AG | Te 2 13:7) || Saha abe XXII 3135 | 359 | 672 9 8.8! || 63s ees a aera | Total 4045 | 653 | 1039 Jap | GO| Tees Heeaeian In table 23 the data for the male classes available in the pre- ceding tables are brought together in the first four columns. In the four columns that follow these, the percentages of cross- ing-over between: (1) eosin and lethal; (2) lethal and minia- ture; (3) eosin and miniature, are calculated for each of the experiments taken separately and for the sum total of the data.§® There is a considerable amount of difference in the percentages from the different experiments, as an examination of the table will show. If we take the totals as the more significant it ap- pears that Lethal 1 lies nearly midway between eosin and minia- ture, but nearer to eosin. The total ‘distance,’ viz., 25 between eosin and miniature is smaller than that calculated from other data. In the preceding experiments the females as well as the males give significant figures for eosin and miniature. In the next table, 24, these are brought together, and the percentage of crossing over between eosin and miniature calculated. This is 30.7 per cent. In order to see whether the lethal factor has itself any influence on the results (which is not to be expected), the data from the sister individuals that gave a 1:1 ratio have also been utilized and are given in table 25. The percentage ‘distance’ is 32.4, which agrees with the calculation for the totals of the preceding table. Finally, the female classes alone are taken in the 2:1 count. These 6 The data in two cases, viz., 15a and 15b are not calculated separately, since the numbers are small. 114 T. H. MORGAN give a percentage of 32.3. is probably a disturbance due to viability in the males. It seems to follow that the low per- centages found when the males of the 2:1 ratio are taken alone The linkage of Lethal II with eosin and with vermilion As in the last case, the data for vermilion are to be considered. Figure C will serve to show the mode of entrance of the factors into the experiment if the factor V be used in place of M near TABLE 24 NO. SEX RED LONG WH. MIN. RED MIN. WH. LONG ae 91:1 392 344 166 163 1:1 245 332 159 132 Q 2:1 488 488 223 226 XI Q 2:1 367 327 164 159 XIII Q 2:1 467 410 230 162 SIV o ge A7Otts 4 450 227 229 Xv2 Q 2:1 113 104 58 59 Xv? 9 2:1 50 46 14 16 pd. 9 1:1 313 282 142 140 1:1 275 27. by) 146 127 o 2-1 283 220» \/ 130 131 xx 9 1:1 1874 1612 1025 913 OT 9 1:1 139 135 65 4 J 1:1 112 120 66 44 XX Q 2:1 3509 3357 1388 1320 Total 9106 8512 4103 3693 * The male class of XX is omitted because the white long males are far be- hind expectation. TABLE 25 NO. SEX RED LONG WH. MIN. RED MIN. WH. LONG XI 2 392 344 166 163 of 245 332 159 132 XIX Q 313 282 142 140 of 275 275 146 127 © 139 135 65 (2 XXI of 112 120 66 44 1476 | 1488 744 678 Total SEX-LINKED LETHALS IN DROSOPHILA 115 which it really lies. The following list will serve to recall the possible combinations in the gametes of the F, female. ! W 1. V —red lethal POM SCEGUSOUCK OM rasta np esas 35.56 Fede e's» F : aie w® Lz v —eosin vermilion {W L. v —vermilion w° ls V —eosin lethal BOTIICCENOSA=OVELI 16 caja alti isles ors» ace ogee Wl. v —lethal vermilion w° L. V —eosin Doublevenoss=Oviers yee sce os cee W i, V—red 347 w’ l, v —vermilion In table 26 the available data are given in the first four columns, and the percentage distance between eosin and lethal, and lethal and vermilion, and eosin and vermilion in the following three columns. The calculations are given for the experiments taken singly, and for the totals. There is again a considerable discrep- ancy in the separate counts. The totals show that lethal lies between eosin and vermilion, but nearer to eosin. The total percentage distance between eosin and vermilion is 27.9. As before, a calculation from the total data was made, table 27, on the basis of the male and female cases taken together, which TABLE 26 NO. | EOS. WARE VERM. | EOSIN RED | Race | robe Pd ca | 464 | 48 109 B/W y Soe 18-2) lose yt 219" 2) 33 68 Of OP 101 oe seg XVIII | 219 Pe an ete eae Ae We dad PO Ee eo = — = a | | = =— | = = ——— Total | 902 | 124 | 27 | 6 | 104 | 185 | 27.9 TABLE 27 NO. | SEX | RED LONG EOS. VERM. | VERM. | EOSIN XVI Q 1: 428 407 | 168 166 Ad 321 AOR S=\sey, 120 78 921 | 499 a ae 181 XVII Q 2:1 278 2530 | 102 88 XVIII Q 2:1 264 211 107 84 Total | 1790 1713 664 597 116 T. H. MORGAN gives 26.5 per cent. Similarly the totals based on the experi- ments with 1:1 ratios gives 25.2; and those based on the females alone of the 2:1 ratios give 27.5. The difference in these is not greater than is expected. The lower percentage for the males taken alone is again probably due to viability. Summary of linkage data of Part II In the following table all the data available in the preceding pages (Lethal 11) bearing on each linkage value have been brought together in a grand total: per cent total cross-overs cross-overs White (or eosin) lethal II......... 8,152 812 10.0 Lethal I iminiatures: ...%.. 282 h 6,407 974 15.2 Hethalei vermiliont. 4.5: oes 1,604 313 19.5 White (or eosin) miniature........ 36,022 11,048 5 S0Re Eosin (or white) vermilion........ 6,023 1,612 26.8 In the preceding table there are some discrepancies that appear when the percentages of crossing-over of the different counts of the same experiment are compared. Discrepancies like these may be due sometimes to insufficient numbers, sometimes to variations in differential viability (including the lagging behind of some of the male classes); in the case of lethals the absence of the ‘contrary’ classes (since one class dies) makes it impossible to check up viability in the male classes. Special external and internal conditions may at times affect the degree of interchange between the homologous chromosomes, which, by changing the gametic ratio, will affect the realized classes. To determine the loci of factors accurately special studies and corrections are necessary. The present data pretend no more |. than to give the approximate positions of the loci, and, in this sense, the results are as consistent as is to be expected under the conditions of the experiments that were carried out with another end in view. SEX-LINKED LETHALS IN DROSOPHILA 117 SEX-RATIOS HIGHER THAN 2:1 A glance at the last columns of the tables will show in many cases that the sex-ratio is higher than 2:1. Most of these differ- ences are obviously chance deviations. There are some, how- ever, that are so high as to suggest higher ratios. In fact, the start was made with a 5:1 ratio. It seemed at first as though those ratios higher than 2:1 might be due to two lethal factors that were simultaneously present, although this involves either two simultaneous mutations or the appearance of a new lethal in an egg that already contained one (through mutation or fertili- lization). Whether such an hypothesis is improbable will depend on how often lethal factors appear in stocks like these. Let us see how it might be shown that the high ratios could be explained as due to two lethals. If a third lethal 1; were present it would be in all probability beyond miniature, or, like the second lethal, between miniature and white. On the first supposition the pos- sibilities are shown in figure F. There are three single cross- over classes, three double cross-over classes and one triple cross- over. Of the sixteen possible classes of males only four would come through, as shown in the following list: W 1,M1; —not realized w° Lo m L; —eosin miniature W L. m L; —red miniature w* l, M1; —not realized W l,m L; —not realized w® L. MI; —not realized W 1,M L; —not realized \w* L. m1; —not realized {JW L: M1; —not realized w* l, m L; —not realized W L.m1; —not realized i w® l, M L; —not realized W1.m1;,; —not realized w* L» M L; —eosin long W L, M L; —red long w’l,m1,; —not realized Canim INON-CEGSS-OVEE: 4 aoe eek os ee (pbesimeolercross-Over:.. 4. eee fee Ke) me Sina lekCrOSS-OVEL ake a nee ne (a) ising leveross-0Vvete. eee oes (e) Double cross-over...... ee res Gs Doubleterass-over!:. 6 c4)08.- o-oo e ee (=) Doubleieross-Overn. «cease ee dh) SeDrripleicross=overs. 5 c22)2.: Tease fone One of the four that are possible is the triple cross-over or red long. A triple cross-over is a very rare occurrence and would THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 17, No. 1 118 T. H. MORGAN not be expected within the numbers that appear in the experi- ment. The double cross-over is eosin long whose expectation is also rare. In table 11 there are two lots, Nos. 16 and 20, in which no males appear in this class, but as the expectation calls for very few flies here, it is unlikely that this can be the cor- rect interpretation. If it were the correct interpretation we should get evidence of this third lethal when the red females of No. 16 were bred, as in table 12. But here there is no evidence of a third lethal, nor is there any in the following generations. It seems unlikely, therefore, that there can be any such lethal (i.e., one beyond miniature) involved in this experiment. On the other hand, if the supposititious additional lethal should be between eosin and miniature, as shown in figure G, the effect would be to lower the class red miniature or eosin long below ex- pectation for one lethal. In fact, the percentages of these classes are too low, but one would not be warranted, I think, from these data, in postulating a third lethal to account for this numerical discrepancy. It seems probable, then, that the high ratios are either extreme cases of the 2:1 ratio, or that there is present some disturbing element not yet detected. The former view does not seem probable for the numbers are large. The second interpretation is more plausible, especially when the fact is recalled that the start was made from an exceptionally high ratio (5:1), and the line maintained by selecting the offspring of the pairs that showed the highest ratios. There is another possibility that should at least be mentioned. The males hatch later than the females, and in the crowded bottles the males string out for some days after the females have ceased to hatch. In the present case there was some crowd- ing, but the bottles were run to a finish, or practically so, in order to get the full male count. That the low male ratio is not due to this condition is shown by the fact that the non-cross-over class (eosin miniature males) compares favorably with the correspond- ing classes of females. SEX-LINKED LETHALS IN DROSOPHILA 119 Sterility An occasional test was made of the males that were obtained in the high sex-ratio pairs. Thus the sex-ratio from which table 1 came was 7829 to l67¢. Every one of these sixteen males was sterile. (Of their sisters only 13 out of 78 were sterile, or at least gave no offspring). Ww jh be M a eT rE ee bare Le, i Ww le I M i ee eae we L, Le st Ww 1s is M << Cc w © Ls Les vu Ww i i M x d w © Ls lug ae Ww i Ie M x x e Ww V i M x x f w 4 ls ™M x x g w 2 ee Le A Ww - lp M x x x h Figure G 120 T. H. MORGAN In No. 39 (ratio of 115:33), seven out of fourteen females mated gave no offspring. One lot in table 1 shows a ratio of 116 9°¢ to 462%. Out of 3199 tested, four produced nothing. Out of 1474, seven were sterile. One lot (xvi1) of table 1 had a sex-ratio of 9699 to 3002. Out of thirty-nine females mated, only one gave no offspring. Six males tested were sterile. No. 12 of table 1 had a sex-ratio of 12299 to 3277. Out of forty-three females mated, twelve produced nothing. Out of five males tested two produced nothing. Summing up the results for the males, we get: males males sex ratio tested sterile SAG He OF ce re See certo Ee coe teis tene CRB eh 16 16 TG AG Ce 2c ee ee RE EME og, ie eee 14 if OG30 PAS 2: AEA eee a yeh seks eae | Lie tah tere 6 6 22S Ee 5 TN Ses en er Erg ls etnias We ca cs 2 TO tall een Been ete hes th Nee en ce nS ER 41 31 AN HYPOTHESIS TO ACCOUNT FOR THE EXCEPTIONALLY HIGH RATIOS The following very high sex-ratios have been recorded for Drosophila: i females males 1 0 135 2 108 0 3 104 0 4 73 0 5 63 0 ‘ k 6 40 0 Quackenbush - 43 0 8 33 0 9 31 0 1 68 0 2 52 1 3 30 0 7 Quackenbush, L. 8S. Unusual broods of Drosophila. Science, N.S., vol. 32, p. 183, August 5, 1910. SEX-LINKED LETHALS IN DROSOPHILA 121 females males ratio 312 3 104:1 169 5 34:1 8 Rawls 276 26 10:1 291 54 5901! In those cases where the offspring were tested (Rawls) it was found that the very high ratio disappeared. It seems to be due to a particular combination that was subsequently lost. If we assume that two linked lethals occurred, one in each of the sex. chromosomes of the mothers that gave these high ratios, an explanation of the results is apparent. Such a female herself could live, since the lethals affecting different parts of the indi- vidual have each its normal allelomorph in the other sex chromo- some. The eggs produced would be in the main of two kinds— half containing one lethal, half the other. Since all of the sons derive their single sex-chromosome from the mother, they will perish. But if crossing over in some of the eggs of the F, female should occur then in such eggs one of the chromosomes will get both lethals, and the other both the normal allelomorphs. The former type of egg, if fertilized by a male-producing sperm, would fail to produce a male; the latter type of egg would pro- duce a normal male. The few males that appear may repre- sent this class. There are two ways in which the original double lethal female might be imagined to arise. A new sex-linked lethal might appear in one of the sperm-cells. If such a spermatozoon fer- tilizes an egg already containing a lethal a double lethal female that is viable would result. The frequency with which lethal mutants have turned up would make plausible this assumption. The same end would be reached if, in a female already contain- ing one sex-linked lethal, another lethal should appear in the other sex-chromosome. For two other results no hypothesis even can be suggested. First, the remarkable sterility found in the F; males when these high sex-ratios appeared; second, the converse case described by Quackenbush where 135 males and no females appeared. 8 Rawls, E. Sex ratios in Drosophila ampelophila. Biol. Bull., vol. 24, Janu- ary, 1913. Lee T. H. MORGAN I wish to acknowledge my indebtedness to Mr. A. H. Sturtevant and Mr. C. B. Bridges with whom I have discussed the theoretical questions involved in the experiments, and also my indebted- ness to Miss E. M. Wallace and to Mr. J. 8. Dexter who have made the greater part of the 55,000 records that furnish the evidence on which the conclusions rest. CLUSTER FORMATION OF SPERMATOZOA CAUSED BY SPECIFIC SUBSTANCES FROM EGGS JACQUES LOEB From the Laboratories of the Rockefeller Institute for Medical Research, New York INTRODUCTION In several papers F. Lillie! has described a very interesting specific phenomenon of apparent sperm agglutination which oc- curs when the sperm is mixed with sea water which has been in contact for a short time with a sufficient quantity of eggs of the same species (Arbacia and Nereis): In the case of Arbacia the addition of two or three drops of egg-sea water } (i.e., one volume of eggs to four volumes of sea water) which has stood half an hour, to about 2 cc. of fresh milky sperm suspension causes formation of agglutinations 1 to 2 mm. in diameter in a few seconds. The agglutination may be so strong that the fluid between the white agglutinated masses appears perfectly clear. The masses gradually fade from view in a few minutes, but microscopic agglutinations may remain half an hour or more. The agglutination is, therefore, only transitory or reversible, as Lillie states. It is specific since e.g., the supernatant sea water of Arbacia eggs acts only on Arbacia sperm and not on other sperm. It is very natural that Lillie should have been led to the idea that such a striking specific phenomenon as this agglutination must play a role in the process of fertilization and he has recently offered a very carefully worked out hypothesis which makes this phenomenon of agglutination not only the center of the process of fertilization and of artificial parthenogenesis but he even hints that it may be involved in the phenomena of heredity. 1 Science, N.S., vol. 36, p. 527, 1912; Journ. Exper. Zodl., vol. 14, p. 515, 1918. 123 124 JACQUES LOEB Lillie’s theory of the phenomenon of agglutination is an appli- cation of Ehrlich’s side-chain theory, a fact which gives it addi- tional interest. In previous papers I have described the secretion of a substance by the ova of the sea urchin, Arbacza, in sea water, which causes aggluti- nation of the sperm of the same species. The eggs of Nerezs also secrete a substance having a similar effect upon its sperm. I therefore named these substances sperm-isoagglutinins. During the present summer I have ascertained that in the case of Arbacia, and presumably also of Nereis, the agglutinating substance is a necessary link in the fertilization process and that it acts in the manner of an amboceptor, having one side-chain for certain receptors in the sperm and another for certain receptors in the egg. As this substance represents, presumably, a new class of substances, analogous in some respects to cytolysins, and as the term agglutinin defines only its action on sperm suspensions, I have decided to name it fertilizin.* The writer had for many years observed that when the eggs of the Californian sea urchin Strongylocentrotus purpuratus, were fertilized with sperm of their own species the spermatozoa would not always scatter but would form small clusters which were often visible with the naked eye. These clusters would disappear in a few minutes. The whole phenomenon resembles strikingly the phenomenon described by Lillie under the name of sperm agglutination, and is possibly identical with it. The writer was interested to find out what the conditions of this cluster formation of the sperm and its relation to the process of fertilization were. Since he is not certain whether this cluster formation observed by him on the Californian sea urchin is identical with the observations of Lillie on the agglutination of sperm in Arbacia, he will confine himself to a discussion of his own experiments and observations, leaving it for future work to decide to what extent they harmonize with Lillie’s observations and con- clusions. 2 Science, N.S., vol. 38, no. 980, p. 524, October 10, 1913. CLUSTER FORMATION OF SPERMATOZOA 125 METHOD OF OBSERVATION AND THE SPECIFIC CHARACTER OF CLUSTER FORMATION If we put one or more drops of a very thick sperm suspension of the Californian sea urchin, 8S. purpuratus, carefully into the center of a dish containing 3 cc. of ordinary sea water, and let the drops stand for one-half to one minute, and then by gentle agitation mix the sperm with the sea water, the at first rather viscous mass of thick sperm is in a few seconds distributed equally in sea water and the result is a homogeneous sperm suspension. When the same experiment is made with the sea water which has been standing for a short time in a dish over a large mass of eggs of the same species, the result is entirely different. The thick drop of sperm seems to be less miscible and instead of a homogeneous suspension of sperm we get as a result the formation of a large number of distinct clusters which are visible to the naked eye and may possess a diameter of 1 or even 2 mm. The rest of the sea water is almost free from sperm. These clusters of spermatozoa last for from two to ten minutes and then dissolve by, the gradual detachment of the spermatozoa from the periphery of the clusters. This phenomenon is to some extent specific. The sperm of the sea urchin Strongylocentrotus purpuratus, will give the cluster formation with the supernatant sea water of the eggs of 8S. purpuratus; the sperm of the sea urchin 8. franciscanus will give the cluster formation with the supernatant sea water of eggs of its own kind as well as with the supernatant sea water of the eggs of S. purpuratus.” In the latter case the clusters dis- solve a little more quickly than if franciscanus sperm is added to the supernatant sea water of franciscanus eggs. The sperm of purpuratus will not form clusters with the supernatant sea water of the eggs of franciscanus. It is of interest that the specificity is not reciprocal in the case of these two sea urchins. The sperm of neither formed clusters with the supernatant water of starfish eggs or of mollusc eggs. The sperm of starfish (Asterias ochracea and Asterina) gave no cluster formation with the supernatant sea water of their own eggs or of the eggs of the two sea urchins. 126 JACQUES LOEB We shall have to return to these data in a later chapter when we discuss the relation between cluster formation and fertilization. The following experiments were carried on with the sperm of S. purpuratus and the supernatant sea water of the eggs of the same species, unless the contrary is stated. APPARENT SURFACE TENSION PHENOMENA AND CLUSTER FORMATION In analyzing the formation of these clusters the writer was struck with the fact that the cluster formation showed peculiari- ties which resembled the action of surface tension. The clusters were usually spherical, or had a tendency to become so. When two clusters were brought into contact with each other they fused at once into one spherical cluster with a larger radius, a behavior which would also be observed in the case of drops of substances immiscible with water under similar conditions. The formation of the clusters themselves resembled surface tension phenomena. When a drop of purpuratus sperm is gently agitated in a little dish with a few cubic centimeters of ordinary sea water streaks and cylindrical masses of sperm are formed in the water which, however, show nothing that reminds one of surface tension phenomena. The spermatozoa are gradually scattered without surface tension offering any resistance to the scattering. If the same experiment is made in the supernatant sea water from the eggs—in egg-sea water—the streaks of sperm produced by agitation behave somewhat like cylinders of a very viscous substance which is immiscible with water, e.g., a viscous oil or a calcium soap. Short streaks or cylinders contract into spherical masses, the above described clusters; and long cylinders break up into a series of small clusters. In an attempt to account for this apparent or real réle of surface tension in cluster formation the writer thought first of the possi- bility that it might be due to an agglutination of the masses of sperm under the influence of the egg-sea water. A study of the real phenomenon of sperm agglutination, however, showed that it does not lead to any formation of spherical clusters. The writer had shown eleven years ago that real sperm agglutination CLUSTER FORMATION OF SPERMATOZOA 127 can be produced if we add 2 or 3 ce. of 4; NaOH to 50 ce. of sea water.’ He found recently a good method of producing sperm agglutination with less alkali in the case of the sperm of starfish. When this sperm is put into 50 cc. sea water + 0.5 ec. X, NaOH it shows a tendency to agglutinate only after about one hour. But we can produce a real agglutination of the spermatozoa after about only twenty minutes when we put the sperm into the super- natant sea water of eggs or sperm of purpuratus. This agglutina- tion is not specific, since it can also be produced by a great many other substances, e.g., cattle serum or even white of egg. In this case the spermatozoa at first stick together to form short rows or threads; and later the threads begin to stick together and form irregular networks. At no time is there any appearance of clus- ter formation or anything suggesting the phenomena of surface tension. The writer is therefore under the impression that the cluster formation of the sperm in the supernatant sea water of its own eggs is a phenomenon of a different type from agglutination. MOTILITY OF SPERM AND CLUSTER FORMATION In observing the clusters the writer was struck with the fact that the spermatozoa at the periphery of a cluster are in free pro- gressive motion, a fact which is incompatible with the assumption of agglutination. When the clusters were small or when the sperm suspension was thin it was possible to observe the spermatozoa which are in the center of the cluster. It was seen that the sper- matozoa in the center also were in very lively motion, with the pos- sible exception of small lumps or groups of spermatozoa which may have stuck together. The clusters reminded the writer of a dense swarm of insects which move like a coherent mass through space. These clusters move like one solid body through the water, notwithstanding the fact that the individual spermatozoa are free to scatter. Under the influence of these observations the writer formed the idea that the cluster formation and possibly the apparent phenom- ena of surface tension might be the outcome of some tropistic 3 Loeb, Arch. f. d. ges. Physiol., Bd. 99, p. 323, 1903; Bd. 104, p. 325, 1904. 128 JACQUES LOEB reaction of the spermatozoa. If this were the case, we should expect that anything that diminished the motility of the sperma- tozoa would lessen the tendency of the sperm to form clusters, and if the sperm were paralyzed completely the cluster formation would also cease completely. It was easy to show that both assumptions were correct. To 3 cc. of a dense sperm suspension in ordinary sea water were added 1 or 2 drops of a 0.1 per cent solution of NaCN, and the whole thoroughly mixed. In one or two minutes the sperm lost its motility and did not regain it when put into sea water. When one or several drops of this immobilized sperm were added to the egg-sea water and when after one minute the dish was gently agitated, the sperm behaved exactly as if it had been put into normal sea water. Not a trace of cluster formation was notice- able; a slight agitation sufficed to bring about a perfectly homo- geneous mixture of the sperm in sea water. After two hours the sperm became motile again when put into sea water. When such sperm, after the recovery of its motility, was put into egg-sea water a very powerful cluster formation occurred again. These experiments were varied and always proved definitely that the whole phenomenon of cluster formation existed only when the sperm was motile. There are other ways of paralyzing the spermatozoa. When the sperm of purpuratus is heated to a temperature of 35°C. or even 36°C. the sperm remains motile and the phenomenon of cluster formation is striking when a drop of such sperm is added to 3 cc. of egg-sea water and the mass is agitated. As soon as sperm is brought to a temperature of 37.6° or above and rapidly cooled, the motility is gone and no cluster formation takes place. The same experiment was made with the sperm of Strongylo- centrotus franciscanus and the supernatant sea water of eggs of the same species. When the sperm is heated to a temperature of 36.2° its motility continues and the cluster formation is not dimin- ished. When the sperm is heated for one minute to a temperature of 37° the motility of the sperm is only diminished and only small clusters are formed. If the sperm is heated to 38° the motility of the sperm disappears and the phenomenon of cluster formation is impossible. CLUSTER FORMATION OF SPERMATOZOA 129 The same result is obtained if the motility of the sperm is dimin- ished or annihilated through the addition of KCl to sea water. These are all very striking demonstration experiments, which leave no doubt that the cluster formation and the apparent sur- face tension phenomena depend exclusively on the motility of the spermatozoa. On the other hand, .the writer convinced himself that in the true phenomena of sperm agglutination, the motility of the sperm is of no concern. We have mentioned the fact that the sperm of Asterias, when it has been in 50 cc. sea water + 0.5 cc. 3; NaOH for fifteen or twenty minutes, undergoes a real agglutination when mixed with the supernatant sea water of different kind of eggs or of cattle serum. This real agglutination takes place just as well after the spermatozoa have been completely immobilized by KCN as before. We may therefore be sure that the cluster formation is not due to an agglutination. « CLUSTER FORMATION A POSSIBLE TROPISTIC REACTION The writer’s idea of a tropism underlying the cluster formation is at present only a mere working hypothesis about which it is therefore not necessary to say much. It is briefly this, that the spermatozoa which are rendered extremely active by the egg-sea water are at the same time repelled by it, in other words, that they possess a negative chemotropism or a negative differential sensi- bility towards the egg-sea water; while any small or large mass of spermatozoa at the boundary of or in egg-sea water acts as a center to which the isolated spermatozoa are positive. This would account for the fact that the cluster formation is a function of the motility of the spermatozoa and could also account for the apparent surface tension phenomena. We should also under- stand why the cluster formation (just like Lillie’s ‘agglutination ’) lasts only a few minutes. Since the egg-sea water must gradually diffuse into the mass of spermatozoa, the boundary at which they are repelled must finally cease to exist. As soon as the concen- tration of the active substance of the egg-sea water is the same or almost the same in the cluster or the mass of spermatozoa as in the surrounding sea water there is no more force active which may induce or preserve the cluster formation. 130 JACQUES LOEB The idea of a negative reaction of the spermatozoa to the egg- sea water is in contradiction to Lillie’s statement that the sperma- tozoa are positive to the egg-sea water. The writer is not quite sure whether Lillie’s statement is based on a correct interpreta- tion of his observations. Lillie introduced a drop of Arbacia egg-sea water into a suspen- sion of Arbacia sperm under a cover glass. In this case a dense ring of spermatozoa was formed ‘“‘at the margin of the drop with a simultaneous formation of a clear external zone about 1.5 to 2 mm. wide; the ring then breaks up into small agglutinated masses and so becomes beaded” (p. 550). In the interior of the drop very few spermatozoa are found. If the spermatozoa were posi- tively chemotropic to the egg-sea water, as Lillie suggests, they should rush into the drop instead of forming aring aroundit. The writer is inclined to interpret this formation of a ring with a clear external zone around it as an indication that the spermatozoa are negatively chemotropic to the strong egg-sea water and possibly positively chemotropic to the more diluted egg-sea water or to the dense collection of spermatozoa in the ring. Those in touch with the margin of the drop are repelled by the drop and those at some distance from the drop are attracted towards the ring or towards the drop. This creates the dense ring next to the drop of egg-sea water and explains the formation of the clear space externally to the ring. When a cluster scatters it does not scatter equally but one notices that isolated microscopic lumps or beads of spermatozoa may be left in the center of the original cluster. Later these beads or lumps scatter also. It is possible that the spermatozoa constituting these lumps or beads stick temporarily together and that this is caused by a specific substance contained in the egg-sea water. This agglutination, however, cannot account for the fact that cluster formation is only possible if the sperm is very motile. The cluster formation is, aside from the increased motility, the only striking phenomenon which the sperm of the Californian sea urchin shows in the presence of egg-sea water. The writer wishes the statements of this paragraph to be taken only provisionally. CLUSTER FORMATION OF SPERMATOZOA 131 THE CONDITIONS WHICH DETERMINE THE SCATTERING OF THE CLUSTER The clusters (just like Lillie’s ‘agglutinations’) have only a short duration of from two to ten minutes, as the circumstances may be. It was of interest to find out some of the conditions which determine their duration. It was found that the stability. of the clusters depends upon the alkalinity of the sea water. Ina neutral solution the big clusters may last a considerable time, half an hour or more, while in sea water to which a sufficient amount of alkali has been added the clusters may scatter in a minute. The reader must remember that if we add HCl or NaOH to sea water part of the added acid or base will be neutralized by the carbonates and phosphates of the sea water. To5cc. supernatant sea water from purpuratus eggs were added 0, 1, 2, 3, 4 drops of 4 NaOH and 3 drops of a dense suspension of purpuratus sperm were added toeach. In all dishes a large cluster was formed. In the dishes with 4 and 3 drops of NaOH the clus- ters were dissolved almost instantly after formation, in the dish with 2 drops the resolution occurred more slowly and it lasted longest—about eight minutes—in the sea water to which no alkali was added. In a second experiment to 5 cc. of the same egg-sea water 0, 1, 2, 3 and 4 drops 4+ HCl were added, and then purpuratus sperm introduced. In 5 cc. egg-sea water + 4 drops of HCI no cluster formation occurred, probably because the motility of the sperm was too rapidly annihilated. In the dish with 3 drops HCl only a trace of cluster formation was noticeable; with 2 drops a moder- ate cluster formation occurred and only in the two dishes with 0 and 1 drop of HCl was the cluster formation very powerful, since here the motility of the spermatozoa was not impaired. In the egg-sea water without acid the clusters disappeared much more quickly than in the sea water with 1 or 2 drops of acid. Experiments in which neutral artificial sea water was substi- tuted for normal sea water showed that at the point of neutrality the cluster formation is most durable. The big clusters continued to exist as long as half an hour, while in alkaline solutions they disappeared very rapidly. In acid solutions no cluster formation hoe JACQUES LOEB was possible probably on account of the fact that the sperma- tozoa became immobile. These and other experiments prove that an increase in the alkalinity of the solution shortens the duration of the clusters; in spite of the fact that an increase in alkalinity of the sea water favors the real agglutination of sperm. The writer tried then to ascertain which salt solutions favor the formation of these clusters. To investigate this point the ovaries and testes of purpuratus were washed in an m/2 NaCl solution and then put directly into another m/2 NaCl solution without coming in contact with sea water. It was found that the super- natant solution of the eggs did never, or only exceptionally, give rise to cluster formation with the NaCl sea water; the reason for this may be partly the fact that the spermatozoa are practically inactive in a pure NaCl solution and that although the presence of the supernatant NaCl solution from the eggs stimulates the spermatozoa into activity this may not always be sufficient. The addition of KCl does not materially improve the cluster forma- tion, the addition of the chlorides of Mg, Ca, Sr and Ba and of MgSO, vastly increases the cluster formation or induces it in an otherwise inefficient NaCl solution. It is not possible to draw any conclusions from these facts upon the nature of the process underlying it. THE ORIGIN OF THE SUBSTANCE CAUSING THE CLUSTER FORMATION Lillie assumes that the substance which causes the phenomenon described by him as agglutination is given off by the egg itself though he states that the jelly which surrounds the egg—viz., the chorion—is saturated with this substance. The writer was curious to know whether the phenomenon of cluster formation depends upon a substance given off by the egg or whether it is due to a substance originating from the chorion. It could easily be shown that the latter is the case. Herbst had stated that the chorion of the sea urchin egg can be dissolved by acid. The writer therefore put a mass of eggs of purpuratus for three minutes CLUSTER FORMATION OF SPERMATOZOA 133 into 50 ec. sea water + 3 cc. 7; HCl. The eggs were constantly squirted with a pipette to prevent them from sticking to the glass and were then transferred to normal sea water. They were then washed five times in succession in normal sea water under constant squirting with a pipette and then left standing in a refrigerator with a small volume of sea water. At no time did the sea water in which these eggs were kept give any trace of a cluster formation with fresh sperm. The supernatant sea water was tested a few hours after the acid treatment and two or three times daily on four consecutive days. These eggs which had apparently lost their chorion had permanently lost the power of giving off to the sea water a substance which causes the cluster formation of the spermatozoa of the same species. If the substance were con- stantly given off by the egg it should have been found after some time in the supernatant sea water. The experiment was repeated a number of times with the same negative result. On the other hand, it was easy to show that the acid sea water (50 ce. sea water + 3 ec. 4 HCl) in which the eggs had been washed contained the substance which is responsible for the cluster formation in large quantities. This acid sea water was filtered and the filtrate neutralized with NaOH (with neutral red as an indicator). The neutralized sed water gave with sperm of the Same species a very powerful cluster formation. This neutralized sea water kept the power of inducing cluster formation for about three days (during which time it stood in the refrigerator) but had lost it the fourth day. This seems to indicate that the substance causing cluster forma- tion is derived from the jelly-like substance surrounding the egg (the chorion) but does not emanate from the egg itself. If this substance which causes the cluster formation should be identical with the substance which Lillie calls ‘fertilizin,’ which is very probable, it is obvious that his conclusion that the substance comes from the egg is untenable. This would also make it im- possible to attribute to this substance a réle in the process of artificial parthenogenesis. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 17, No. 1 134 JACQUES LOEB CLUSTER FORMATION AND FERTILIZATION Lillie’s application of the side-chain theory to the problem of fertilization rests on the assumption that the substance which causes the phenomenon he describes as agglutination is indis- pensable for fertilization. This ‘fertilizin’ is in his theory an amboceptor which must combine at one end with the spermato- zoon at another end with the egg; the ‘fertilizin’ when in com- bination with the spermatozo6n undergoes a change and then fertilizes the egg. It is therefore a substance given off by the egg which in Lillie’s opinion causes its fertilization, and not, as we all had hitherto assumed, one or more substances contained in the spermatozoon. We have just seen that the substance which causes the cluster formation does not come from the egg but is given off by the chorion—or possibly is the chorion itself which is slowly soluble in sea water. ; We can show in a number of different ways that eggs which have lost or do not possess the power of giving off a substance which induces cluster formation may possess the normal power of being fertilized. If we treat eggs of purpuratus for three minutes in 50 ec. sea water + 3 ec. % HCl and wash them about five times in sea water they have lost the power of causing the cluster formation of the sperm of purpuratus. Such eggs can be fertil- ized immediately after the washing or at any time during the next two or three days if they are kept in the refrigerator. Their power of being fertilized is not in the least impaired. One hun- dred per cent of the eggs were invariably fertilized and the fertil- ization took place instantly after the addition of sperm. Prac- tically all the eggs developed. The membrane was slightly abnormal which was an after-effect of the acid treatment. The power of the eggs of being fertilized remains unimpaired while their power of giving off substances which cause cluster formation is completely and permanently lost. When we treat the eggs with a fatty acid instead of with a mineral acid they form, when transferred to normal sea water, a fertilization membrane. The fatty acid also dissolves the chorion and such eggs when washed afterwards lose their power of induc- CLUSTER FORMATION OF SPERMATOZOA 135 ing cluster formation of sperm. It is interesting that Lillie states that such eggs lost also their power of causing agglutination, which seems to suggest that Lillie’s ‘agglutination’ and the writer’s ‘cluster formation’ may be the same thing. Lillie states that such eggs which have formed a fertilization membrane have lost their power of being fertilized and he sees in this a support of his theory that without ‘fertilizin’ the egg can no longer be fertilized. Eggs which have formed a fertilization membrane under the influence of butyrte acid can easily be fertilized with sperm if the membrane is torn by shaking. The writer has repeated this experiment this winter and confirmed the earlier observations by Kupelwieser and himself to the same effect. He alsomade sure that the eggs which after the artificial membrane formation can be fertilized with sperm have completely lost the power of causing cluster formation. We thus see that complete loss of the. power of inducing cluster formation can be combined with maximal power of the eggs of being fertilized. HYBRIDIZATION AND CLUSTER FORMATION The best test for a possible connection between fertilization and cluster formation is afforded in the phenomena of hybridization. If the phenomenon of cluster formation were inseparably associ- ated with the power of the eggs of being fertilized, we should expect that sperm should only be able to fertilize the eggs of a spe- cies if the egg-sea water of the same species caused the cluster formation of the sperm. It is easy to show that no connection of this type exists. It is impossible to call forth cluster formation of the spermatozoa of the starfish Asterias ochracea with the egg-sea water of purpuratus and yet 100 per cent of the eggs of purpuratus can be fertilized with the sperm of ochracea and as many as 80 per cent of these eggs ; may develop. The writer showed that this hybridization takes place only in hyper-alkaline sea water and it was therefore nec- essary to test the possibility of cluster formation in both neutral and alkaline sea water; these tests gave always absolutely negative results. 136 JACQUES LOEB The sperm of purpuratus shows no trace of cluster formation with the egg-sea water of franciscanus and yet the eggs of fran- ciscanus are readily fertilized with the sperm of purpuratus. If the cluster formation were caused by a substance which was necessary for fertilization in the sense of Lillie’s theory these and probably many other hybridizations which occur should be impossible. It often happens that in hybridization less than 100 per cent of the eggs are fertilized. The writer tried whether the yield of fertilized eggs could be increased if the egg-sea water from the species which furnishes the sperm be added to the mixture. This would furnish the sperm with the specific ‘amboceptor.’ It was found that hybridizations occur just as well if not better in normal sea water than if the egg-sea water from the species from which the sperm is taken be added. All these facts contradict the assumption that the substance which induces the cluster formation of the spermatozo6n is nec- essary for fertilization. ARTIFICIAL PARTHENOGENESIS AND CLUSTER FORMATION The writer has expressed the idea that the causation of the de- velopment of the egg either by a spermatozo6n or by the agencies of artificial parthenogenesis is due to an alteration of the cortical layer of the egg which may or may not be accompanied by a membrane formation; and he has shown that all hemolytic sub- stances are able to bring about this alteration. Lillie agrees with this idea but differs in regard to the origin of the agent which causes this change in the cortical layer of the egg in the case of fertilization by a spermatozo6n. According to the writer, this change is caused by a substance contained in the spermatozoon while Lillie assumes that it is a substance contained in the egg which must, however, be activated by the spermatozo6n in order to cause the alteration of the surface of the egg. It seems to the writer that Lillie’s assumption is unnecessarily complicated. Moreover, if it should turn out that the substance which is re- sponsible for the cluster formation is identical with the substance which Lillie calls ‘fertilizin’—which is very likely the case— CLUSTER FORMATION OF SPERMATOZOA 137 Lillie’s theory becomes untenable, since this substance does not, in all probability, originate from the egg but from the chorion and since there is, as we have seen, no connection between the presence of this substance and the power of the eggs of being fertilized. A second difficulty which Lillie has not considered lies in the fact that the writer has shown that in addition to the membrane form- ing substance still another, namely a corrective agency, is neces- sary for the causation of the development of the egg. The cor- tical change induces development but the egg as a rule perishes if the second factor of artificial parthenogenesis is not applied (hypertonic solution or lack of oxygen). The writer is suspicious that even a third factor may be implied. It is under these circum- stances difficult to see how the assumption that the ‘fertilizin’ causes development—leaving aside all other objections—can act as an adequate substitute for the known facts of artificial parthenogenesis. Lillie sees a proof for his idea in the statement of Glaser‘ that the filtrate from eggs of Arbacia ground up with an equal volume of sea water will cause normal unfertilized eggs of the same species to undergo one or more cell divisions if they are transferred from the filtrate after one or two hours to normal sea water. Lillie concludes from this that the egg contains its own fertilizing substance, the ‘fertilizin,’ and that it was this ‘fertilizin,’ liber- ated from the eggs when they were ground up and contained in the filtrate, which induced the cell division of the Arbacia eggs in Glaser’s experiments. Waiving the question of how this ‘fertil- izin’ was ‘activated’ without the presence of sperm, the writer sees no reason to assume that the egg extract acted through the ‘fertilizin,’ or any other specific substance, since he has shown that a large number of non-specific substances are able to induce the first cell divisions (without membrane formation) in the egg of Arbacia, e.g., traces of any weak base like HN:, or protamine.® A slight increase of osmotic pressure could also have such an effect. It even suffices to let the eggs of certain females lie for 4 Science, N.S., vol. 38, no. 978, p. 446, 1913. ' 5 Jour. Exp. Zo6l., vol. 18, p. 577, 1912; Arch. f. Entwceklngsmechn. d. Organ., Bd. 38, p. 409, 1914. 138 JACQUES LOEB some time in sea water.® The writer fails to see any reason for assuming that the cell division in Glaser’s experiments was induced by ‘fertilizin.’ The fact that the composition of the mixture which he used was unknown is not sufficient proof that ‘fertilizin’ was the active agency. In conclusion, the writer would like to call attention to the fact that the cluster formation described in this paper or the agglutina- tion of sperm described by Lillie inhibit the fertilizing effect of the spermatozoa instead of enhancing it, since the cluster forma- tion prevents the spermatozoa from reaching the egg. Even from a teleological viewpoint it is difficult to understand why a sub- stance which only prevents the fertilizing action of the sperm should be a necessary link in such action. The writer is inclined to believe that the cluster formation or agglutination of sperm does not occur when fertilization takes place under natural conditions. The writer is not quite sure whether his interpretation of the cluster formation of the spermatozoa as a tropistic reaction will have to be modified or not. If it should turn out to be essentially or partly correct, it might be worth while to point out that we have many examples of tropistic reactions which are of no possible use to the species, e.g., all the phenomena of galvanotropism which are only laboratory products. The writer takes pleasure in expressing his thanks to Prof. S. S. Maxwell of the University of California for his kindness in putting the Herzstein Laboratory at New Monterey at the writer’s disposal. SUMMARY 1. The writer describes the formation of clusters of spermatozoa which is observed when the sperm of a sea urchin is put into the supernatant sea water of eggs of the same species. This specific phenomenon of cluster formation may possibly be identical with the specific phenomenon described under the name of sperm agglu- tination in recent publications of F. Lillie. * Arch. f. EntwekIngsmechn. d. Organ., Bd. 36, p. 626, 1913. CLUSTER FORMATION OF SPERMATOZOA 139 2. The cluster formation resembles the phenomena of surface tension in various respects, e.g., inasmuch as the clusters are spherical or tend to assume a spherical shape, and inasmuch as the fusion of two clusters results in the formation of a larger spherical cluster. When sperm is put into ordinary sea water or the super- natant water of foreign eggs these apparent surface tension phenomena are not observed. In real sperm agglutination neither cluster formation nor the above-mentioned surface tension phenomena are noticeable. 3. It was found that the cluster formation is a direct function of the motility of the spermatozoa. As soon as the spermatozoa. are immobilized by NaCN or by high temperature or by KCI the: cluster formation ceases; as soon as the motility of the sperma- tozoa returns the cluster formation occurs again when the sperm is put into the supernatant sea water of eggs of the same species. The real agglutination of sperm occurs just as well when the sperm is immobilized as when it is motile. The cluster formation is therefore not a form of agglutination. 4. The clusters last only a few minutes, like Lillie’s sperm ‘agglutinations.’ The writer has found that in a neutral solution they last much longer than in an alkaline solution and that they scatter the more rapidly the more the alkalinity of the sea water is raised by the addition of NaOH. 5. The writer offers tentatively a working hypothesis which assumes that the phenomenon of cluster formation is essentially or partly due to a negative chemotropism of the spermatozoa to the egg-sea water. 6. It is shown that eggs which have been treated with acid! sea water lose permanently their power of producing a sub-— stance which causes the cluster formation of the spermatozoa of their own species; while the acid sea water in which the eggs were: treated when filtered and neutralized with NaOH induces a very powerful cluster formation. If it is true that the acid dissolves; ‘the chorion (the jelly-like substance surrounding the egg) this; experiment would prove that the substance which causes the: cluster formation is not formed in the egg but in the chorion. If this substance is identical with the substance which Lilhecalls 140 JACQUES LOEB ‘fertilizin,’» his theory concerning the rédle which this substance plays in the fertilization and development of the egg will meet with serious difficulties. 7. It is shown that eggs which have been treated with a mineral acid like HCl and which have permanently lost the power of causing a cluster formation of the spermatozoa can nevertheless all be fertilized with sperm of the same species and that the rapidity with which the sperm fertilizes these eggs is equal to that with which normal eggs are fertilized. When the acid used was a fatty acid and when membrane formation occurred the eggs also lost permanently their power of inducing cluster formation but retained their power of being fertilized by sperm, provided that the membrane was first torn. 8. The supernatant sea water of the eggs of Strongylocentrotus franciscanus will not induce cluster formation of the sperm of Strongylocentrotus purpuratus; yet the latter sperm fertilizes the eggs of franciscanus. The sperm of Asterias ochracea undergoes no cluster formation in the supernatant sea water of Strongylo- centrotus purpuratus, no matter whether the sea water is normal or hyperalkaline although the starfish sperm readily fertilizes most or all the eggs of Strongylocentrotus purpuratus in hyper- alkaline sea water. 9. The facts mentioned under paragraphs 7 and 8 prove that the substance which is responsible for the cluster formation is not necessary for the process of fertilization. FERTILITY AND STERILITY IN DROSOPHILA AMPELOPHILA I. STERILITY IN DROSOPHILA WITH ESPECIAL REFERENCE TO A DEFECT IN THE FEMALE AND ITS BEHAVIOR IN HEREDITY ROSCOE R. HYDE Department of Zoélogy, Columbia University CONTENTS MERE MOU CCID Were ira? ck ck: «| kl oz 3h SRM 3 170 Rar ARGH TAENAY Sehr as cracks aS ad ors iat che a honoree: ais ae Ageia cee oe 171 INTRODUCTION The following series of papers embodies the results of my work on ‘Fertility and Sterility in Drosophila ampelophila.’”’? The first paper deals primarily with a case of sterility in which the female is affected, together with the effects of selection upon female sterility and its behavior in transmission. The second paper is concerned with the low fertility of a mutant, and the behavior of this low fertility in heredity. The third study deals with the effect on fertility of crossing different races. The fourth study treats of the effect on fertility of crossing within and with- out an inconstant race. The work has been in continual progress since the fall of 1911 when it was undertaken at the suggestion of ero, Ll. He Morgan. Two investigators have studied sterility in Drosophila. Castle and his students found sterile individuals of both sexes in their 141 142 ROSCOE R. HYDE cultures. Moenkhaus found in his strains that sterility was practically confined to the males. He attempted to harmonize Castle’s results with his own by the assumption that each had used a different measure of productiveness. Both investigators are in agreement that inbreeding can not be the cause of sterility, and that sterility is amenable to selection. From Castle’s paper I gather that he considers low productivity and complete sterility in the female as related and that sterility in this case is related to egg structure. Moenkhaus seems to imply that sterility is due to some condition of the sperm. Evidence that bears on these questions is given in the following pages. I wish to state in the beginning that my results warrant me in making a sharp distinction between fertility and sterility as it actually exists in the strains that I have used. Fertility, as I have found it, does not grade into complete sterility. There may however be gradations in fertility. The total sterility that ap- pears in my cultures bears no relation to low fertility. Sterility as it affects the flies in my strains is due to a different condition from that operating to reduce the fertility. This statement is certainly true of the sterility as it affects the female. The dis- tinction paves the way for bringing the results of other investi- gators under a common point of view. Failure to distinguish between sterility and fertility has I suspect led to confusion, cer- tainly in the work of Castle, and probably in the results of Moenk- haus, because as I shall show the two things may relate to quite different phenomena. The inheritance of sterility and fertility in Drosophila must be separately treated if any progress is to be made. In a culture of Drosophila, which I had been inbreeding and to which I shall refer as the inbred stock, there appeared an increas- ing number of sterile pairs. The sterility affected primarily the females as was evident by testing them with other individuals. In the sixth generation which had descended from a single pair of grandparents of the fourth generation, 51 pairs in a total of 105 pairs were sterile. There were 47 sterile females, 3 sterile males and 1 questionable case. FERTILITY AND STERILITY IN DROSOPHILA 143 It seemed probable that the defect was due to something in the hereditary mechanism rather than to some factor in the environ- ment, since the same defect reappeared among the grandchildren which hatched at different times during a period of two months. Since the effects of inbreeding were under observation I was at first inclined to attribute this sterility to inbreeding. STERILITY AND INBREEDING It is a popular belief that one of the effects of inbreeding is to induce sterility. Moenkhaus and Castle, however, after many generations of inbreeding with Drosophila could find no evidence in favor of such a view. Nevertheless, since sterility appeared in my ‘inbred’ culture to such a high degree it was tempting to attribute it to inbreeding. As Moenkhaus has clearly pointed out, if ina sterile inbred stock, the sterility can be eliminated by continued inbreeding, then inbreeding can not be held to be the causative agent. This is the test to which my own case has been put, and by selecting from those families that showed the sterility in the least degree a fertile race was produced. Tables 1 to 7 show the manner in which the sterility appeared in the inbred stock and its elimination by selection. 1. Sterility of the inbred stock TABLE 1 TABLE 2 Sterility of the first five inbred Sterility in the Fs generation generations | NO. NO. No. | GENERATION | PAIRS sree 8 | as a oe eee ieee (———= — = —— —— +: EF, | 16 1? 25 1 iii in LA a F, (6 0 | 27 11 03 30 0 F; o 3 28 5 Ssh ct 0 F, 12 3 30 32 US EZ 1 F;* | 10 2 2 31 31 16 | 14 2 | | 32 10 Gall)" 1? | | hg enn Mane 0 otal... | 61 8 Be otal...) 105. |.051. a7 ASeeLe * The ten pairs of F; were from a sin- * Miscellaneous paire. gle pair of F,4 144 TABLE 3 ROSCOE R. HYDE Sterility in the Fz generation TABLE 4 Sterility in the Fs generation ancus-| No. | NO. STERILE STERILE ANCES- BF ee NO. |srERILESTERILE TRY NO. "0" nee aera : TRY No. “O° eee Sree e 27 134 1 17 a 4 127 | 149) 26 11 10 1 30 127 5 2 127) 15) 21 5 5 0 32 129 | 12 6 WE ANS US 3 3 0 129 164 2 2 0 127 | 165 3 3 0 127 | 166 2 1 1 | NB) UGYE 1 0 il 1384 | 172) 28 12 10 2 Total. 34 | 15 | 4 Total 110 | 39. | 34. liera TABLE 5 TABLE 6 Sterility in the Fy generation Sterility in the Fo generation NO. NO. : he NO. NO. eae NO. PAIRS Pare Pvorey TERILE oa NO. PAIRS _PAIBS rte ie 172 | 228) 28 6 5 1 226 | 232 0 0 0 151 | 226 8 0 0 0 227 | 229°) LO 0 0 0 15k.| (227 4 0 0 0 Total 40 6 5 1 Total hil 0 0 0 TABLE 7 Sterility in the Fi, generation ANCES- SOE sO STERILE|STERILE TRY NO.| “O- ae ards ? cS 229 | 236 4 0 0 0 229 | 237) 10 2 1 1 229 | 238 7 2 0 2 229 | 239 2 1 1 0 229 | 240| 14 2 1 i 232 | 241) 61 6 0 6 229 | 242 | 22 4 1 3 229 | 243) 33 5 1 4 229 | 244) 18 2 1 1 229..|..245 |) 11 4 0 4 Total 182 28 6 22 FERTILITY AND STERILITY IN DROSOPHILA 145 Tables 1 to 7 show that in the beginning of the experiment only a few sterile pairs were found (table 1). Moreover, the fertile pairs were high producers (Part II, table 2) yet sterile individuals appeared in their offspring and these were largely females. At times fully half the females were affected, although there were undoubted cases in which the males were also affected. It is to be noted in the eleventh generation, as shown in table 7, that out of 182 pairs only 6 sterile females appear. The steril- ity as it affected the female had been practically eliminated, for while it had been affecting 50 per cent of the females it affected at the end of the experiment less than 4 per cent, It might seem as though the character had been transferred to a certain extent to the male as 22 males out of 182 are recorded in table 7 as sterile. But the figures, as given here for the male, do not in all probability represent the actual facts in the case. Many of these males were very small, and after they had been paired for a few days prac- tically all the males that proved sterile were crawling over the food with their wings drooping at their sides. This may or may not have prevented them from mating with the females. I was not able to follow the case further, but I have found no evidence in any of the other experiments that sterility, as it affects the fe- males, can be shifted in heredity to the males. That sterility is not due to inbreeding, and that selection is an effective agent in controlling it, is shown in the history of another strain, to which I shall refer as the truncate stock. This stock had been inbred for forty-two generations by Mr. Alten- burg, a graduate student in the department. A great many sterile pairs were appearing in the strain and the broods were so small that it was somewhat difficult, as Altenburg told me, to keep the stock from dying out. Table 8 gives the history of the sterility as it appeared in this stock. When I took charge of the stock fully half of the pairs were sterile and yet on continued in- breeding a large percentage of the sterility disappeared. Table 11 shows that in F,, there were 3 sterile pairs in 21 pairs. That there were but few sterile individuals in the truncate stock at this time is also borne out by crosses that were made with fertile races. Sterility had been largely eliminated, and moreover this “ 146 ROSCOE R. HYDE had been brought about by selecting from those families that showed the least sterility (no. 17, table 9, and 17a, table 10). It seemed to me at first that the large number of sterile indi- viduals that appeared in this stock might bear a causal relation to the low productivity of the fertile pairs, but at the point in the experiment when the sterile individuals were eliminated there was no rise in productivity. I shall return to this question in the second part of the paper. 2. Sterility of the truncate stock TABLE 8 Sterility of the truncate stock as it appeared in successive generations of inbreeding GENERATION Fa) Ful Fos | Fas Faz| Fas| Fag| Fso| For} Fsz| Fsg 53| 46| 19) 1 14) 21) 24) 2)| 8 15 27.12) 60 | 22 3) 0) 0; eae 2h 107 Number pairs tessed) (43)o5.0... 2285 50) sas 31 Number pairs atemile. . ws <0 el e se biel 22* * Some doubtful cases are included. TABLE 9 Sterility of the truncate stock in F4; generation ANCESTRY NUMBER TOTAL SFI || ea} 9 | 15 17 | 21 | 23 25 | 26 33 | 37 | 52 | 53 4/3/4/4/4]/5/3/1] 46 No. pairs tested .. .| 1/8 O0/O0)1)1)/2)/1)1)0)1)0 12 | feat | 1 No. pairs sterile ...| 1 | 0 | 1 2 2 TABLE 10 Sterility of the truncate stock in F'45 generation NUMBER AND ANCESTRY TOTAL 5a | 8a | 17a | 2la 23a | 33a 52a | | | Number pairs tested.... 1 ree 1 1 2 2 19 Number pairs sterile... . if 3 | 0 1 0 0 | 1 6 For the forty-seventh generation one pair was chosen from 17a. It is to be noted that no sterile individuals had appeared in the FERTILITY AND STERILITY IN DROSOPHILA 147 TABLE 11 Sterility of the truncate stock as it appeared in the F'49 generation ANCESTRY NO. NO. PAIRS TESTED | NO. PAIRS STERILE | STERILE 2 STERILE o' 26 3 2 0 2 33 1 0 0 0 35 16 1 1 0 39 il 0 0 0 Motaly 2o6 | 21 | 3 1 2 last two generations of this family. Fourteen pairs were chosen from the Fis generation, two of which were sterile. From this generation pairs were made up as shown in table 11. It should be recalled that the truncate stock was bred at the same time and under the same conditions as the inbred stock, and while the former developed into a fertile stock the latter developed into a sterile female stock. Attention is called here to the fact that the individuals in table 2 are the grandchildren from a single pair of the F, generation. Sterility appears in the different families and affects about 50 per cent of the females. Family 27 is an exception in that 11 pairs show no sterility although descended from the same grand- parents. This small amount of evidence is in accord with the - assumption that some individuals come through without the defect in their germ-plasm, and shows how selection has brought about its results. That one of the grandchildren in this combination had some factor which prevented the appearance in her of the defect is shown by the fact that the defect reappears among her daughters (table 3, no. 134). These numbers are too small to base a safe conclusion upon; but attention is called to them here, since the asumption is borne out in the experiments which follow. BEHAVIOR OF STERILITY IN HEREDITY In a study of sterility one meets at the outset a peculiar diffi- culty in that he cannot breed the animals that show the very defect which he wishes to study. Again there is no way by which sterile animals can be recognized except by pairing them with different mates which involves an additional amount of labor. Moreover 148 ROSCOE R. HYDE the same strain on inbreeding may behave differently in successive generations as has been shown in the history of the inbred and truncate stocks. An additional difficulty is met in finding a highly fertile strain against which to test a known sterile strain. In the face of these difficulties a number of detailed experiments have been required in order to determine (1) the method of transmission of the female sterility, and (2) the relation between low production and sterility in the truncate stock. I shall present first the evi- dence that bears directly on the method of transmission of female sterility. TABLE 12 Sterility as it appeared in the Woods Hole stock LOT DATE | geet: pee STERILE | ogee oT A. >) Moca toh. aco nfs 2% 68. Sepa 0 1 By i duneeeeigioe eee aD 4 2 2 Atul sige. ee 63 3 1? 2 A wild stock to which I shall refer as the Woods Hole stock was characterized by the presence of but few sterile individuals. Consequently this was good material against which to test the sterility of the inbred stock. An entirely fertile stock would be the ideal one to use but it is the experience of all who have bred . these flies that a sterile pair is occasionally met without any assign- able cause. Table 12 gives the history of the fertile Woods Hole stock. At the time the following cross was made there appeared in the Woods Hole stock that was used as a control by placing under similar conditions, 4 out of 70 pairs that were sterile. There were 2 sterile males and 2 sterile females (table 12, Lot B). In order to determine the behavior of the sterility in the F, and F, generations, the inbred females were paired with the Woods Hole males. The reciprocal cross was also made. The offspring that resulted from this cross were paired together, brothers and sisters, for the F, generation. The grandchildren were paired for the F, generation. The offspring from three families of the inbred FERTILITY AND STERILITY IN DROSOPHILA 149 stock were chosen to cross into the Woods Hole stock (see table 4, Nos. 149, 151, and 172). The pairs that were tested from these different families as a control show the sterility to have been present in different intensities in these families; No. 149; 11 pairs in 26 were sterile; No. 151, 5 pairs in 21 were sterile; No. 172, 12 pairs in 28 were sterile. That the controls give a fair measure of the sterility is shown by the crosses made into the fertile Woods Hole stock. ‘These crosses are given, together with the sterility as it appeared in the F, and F; generations, in table 13 a to 13 f. In these tables the first number gives the serial number of the cross (e.g., inbred male by Woods Hole female) and is designated as P;. The second column gives the number of pairs of offspring that were tested from the given family, for the F; generation, while the next column gives the corresponding number that were sterile. The fourth column indicates the number of pairs selected to be- come the parents of the F, generation. The succeeding columns give the corresponding number of pairs tested and sterility as it affected the different sexes. Crosses between inbred & and Woods Hole TABLE 13a Crosses between the inbred stock and the Woods Hole stock showing the behavior of the sterility in the F; and F2 generations. Inbred & No. 149 X Woods Hole 9? P, No. PAIRS Fj PAIRS F; | ANCESTRY | PAIRS F2 PAIRS Fz | STERILE STERILE TESTED STERILE NO. TESTED STERILE 9 1 a 1 la 12 1 0 1 ibe etd 2 2 0 2 i 0 2a 13 3 3 0 2b 7 1 0 1 2c 8 4 1 3 5 | 0 3a 12 0 0 0 3b 20 Rel) iO 1 4 2 te 4a 26 9 9 0 | 4b 16 10 9 1 5 5 Uwe? Ba 10 1 1 0 | 5b 20 1 1 0 5¢ 12 0 0 0 otal sy an. Lae leet 166 33 26 4 THE JOURNAL Ok EXPERIMENTAL ZOOLOGY, VOL. 17, No. 1 150 ROSCOE R. HYDE TABLE 13b Inbred & No. 172 X Woods Hole 2 Eine PAIRS Fy PAIRS F; | ANCESTRY | pares F2 PAIRS F2 STERILE STERILE TESTED STERILE NO. TESTED STERILE 6 8 0 6a | 20 0) 0 0 6b 14 4 4 0 6c 13 0 0 0 6d 32 0 0 0 6e 21 2? 7 2 Keg 7a | 4 Lo% 1 0 “Ao | 28 3? fe 20 ile 8 5 0 8a 9 4 4 0 9 4 0 9a GZ! || 2 2 0 10 3 19 10a! 18 9 9 0 11 3 0 12 6 20 12a | 19 2 1 1 12b 2 0 0 0 13 9 0 13a 16 2 2 0 13b 15 1 0 1 14 6 0 14a 2 0 0 0 14b 3 0 0 0 15* 16* | ile Totalkeeae 46 4 252 31 23 2 * Sterile TABLE 13c¢ Inbred & No. 151 X Woods Hole @ Pot PAIRS PAIRS ANCESTRY PAIRS PAIRS STERILE STERILE TESTED F; | STERILE F) NO. TESTED F2 | STERILE F. ou 18 12) 1 0 19 17 0 19a 0 0 0 20 8 0 20a 21 0 0 0 21 8 0 21a 28 1? 22 9 0 22a, 25 2 1 1? 23 11 0 23a 15 0 0 0) 24 oe a ne 24a 5 0 0 0 25 8 0 25a 10 1 1 0 26 13 0 26a 21 Bye 3? 27 9 0 27a, £8: | 0 0 0 28 6 0 28a, 20 2 1 28b 12 il 0 1 29 0 ‘Totals 113 1 183 10 6 2. FERTILITY AND STERILITY IN DROSOPHILA tot TABLE 13d Inbred 2 No. 149 X Woods Hole &; reciprocal of 13 4 = NO PAIRS PAIRS ANCESTRY | PAIRS pate STERILE STERILE TESTED F | STERILE Fy NO. | TESTED F2 | STHRILE F2) 9 30 6 0 30a 50 a pee 2 31 a 0 3la 4 1? oe 3 0 32a 20 3 3 0 33 5 0 33a 18 9 9 0 33b 14 1 1 0 33¢ 15 3 3 0 34* | si 36* By totale es 21 0 121 21 18 2 * Sterile Reciprocal: Crosses between the inbred 2 and the Woods Hole & TABLE 18 e STERILE STERILE P, NO. PAIRS PAIRS ANCESTRY | PAIRS PAIRS TESTED F; | STERILE F; NO. TESTED F?2 | STERILE F2 ie} J 38 10 0 38a | 28) Messe my, (3 1 f | 38b 5 Oey) 0 0 | be ta8e.. Wo Borer 7 osamenlamoeo: Ye | iG 39 7 Pe 30a ula uals 6 ape 40 8 0 | 40a 15 1? | | | eaObeo ment 7Tesr i Gwen: 36h Minas 41 ee a Gas ib 4a 10p0 Lr O 0 0 42 Se pari ict FNC Aa oe 24 | S10 Os Bhi, (0 42b 35 oN dm aes eet 1 Nae 43 Pee Stewie aC) Ba hen Doe cede Sa memlemas CO 44 De Sad 44a 40 5h iliac ame 45 2 Op ah. 45a Gr etl heeds fy. wae 0 46* | | | 47* | | 48* | | | | 49* | 50* | Eyles | 52* | 53* | atal- +0 43 3. 302 76 64 2 152 ROSCOE R. HYDE TABLE 13f Inbred 9 No. 151 X Woods Hole &; reciprocal of 13 ¢ PAIRS PAIRS ANCESTRY PAIRS PAIRS STERILE STERILE no. Py TESTED F) | STERILE F NO. TESTED F?2 | STERILE F2 g 55 13 0 55a 3 0 0 0 56 9 0 56a 51 3 2 57 13 0 57a 16 1 1 0 58 9 0 58a 22 1 1 0 59 inl 0 59a 14 0 0 0 60 11 lo 60a 24 8 8 0 61 9 0 6la 6 3 2 1 62 4 1l%¢ 62a 22 1 1 0 62b 4 0 0 0 63 8 0 63a 4 2? 1? 1? 63b 4 0 0 0 64 ill 0 64a, 50 0 0 0 65 4 0 65a 32 3 1 2 65b 6 0 0 0 65¢ 5 2 66 67 68 69 70 71 2, 73 74* (or 6" Cite Motaleence 102 2 262 24 17 3 * Sterile This experiment brings out the following facts. In the sterile female stock (inbred) used as a control 39 pairs out of 110 pairs were sterile. In the Woods Hole stock 4 pairs out of 70 pairs were sterile. In the F; generation only 12 pairs in 351 proved sterile. Of those tested 5 were females, 4 were males. In the F, generation 1286 pairs were tested, 195 of which were sterile. There were 153 sterile females and 15 sterile males. The sterility of the female reappeared in both the cross and the reciprocal, indicating that both the brothers and sisters of the affected FERTILITY AND STERILITY IN DROSOPHILA 153 females were capable of transmitting the defect. The control shows that family No. 151 of the inbred stock was the least affected, and it is to be noted that sterility reappears among the grandchildren of this family in less intensity than among the grandchildren of the other two families. This fact seems to indicate that the intensity to which the sterility reappears in the F, generation bears a causal relation to the intensity in which it entered the cross. Crosses between inbred, truncate and Woods Hole stock In the sixth generation of the inbred stock, when the sterility of the females appeared in its greatest intensity (50 per cent; table 2), crosses were made with the Woods Hole stock and also with the truncate stock. The truncate stock was also crossed with the Woods Hole stock, and since the truncate stock at this time was a relatively fertile stock, it serves as an excellent control for the female sterility of the inbred stock. This experiment then consisted of three crosses, together with their reciprocals and controls: (1) The inbred stock by the Woods Hole stock; (2) The inbred stock by the truncate stock; (3) The truncate stock by the Woods Hole stock. The control on the Woods Hole stock used in this experiment shows that one sterile male occurred in 58 pairs (table 12, Lot A). TABLE l4a Crosses between the inbred stock and the Woods Hole stock, showing the behavior of sterility in the F, and F2 generations. Woods Hole 2 X inbred & No. 31 no. Pi meerap \|eemartm | ANCRETEY TESTED ae faded aa ge 1 1 . 2 2 36* 37 22 1 37a 17 0 0 0 37b 13 0 0 0 37¢ 16 0 0 0 38 4 0 38a 10 0 0 0 39 1 0 39a 16 2 1 1 40 0 | Potala... 27 1 72 2 | 1 *Sterile @ 154 ROSCOE R. HYDE The truncate stock shows only a slight degree of sterility (table 8, Fi). The crosses themselves bear out this statement, for, out of the 24 pairs, all are fertile (table 16). Two died, however, before their fertility could be established and are not recorded in the table. Of the 251 pairs tested in the F; generation there are only 2 sterile individuals; a male and a female. The behavior of the sterility is the same in the cross as in the reciprocal. If the sterility behaves like a character which segregates and recombines TABLE 14b Woods Hole * X inbred 9 No. 31; reciprocal 14a no. Pi Bi a Seana ANCESTRY coaean sree SUES SHEE 1 Fi hs Fe | F2 = cs 41 0 Dye” ae 0 0 0 0 42 r Oy.) 7425 12 vi 4 43 24 OR)! ease |S eos: sae 2 1 | 43b | 20 0 0 0 | * 48¢ 40%. ois 1 fod. hc Sah eG 3 44* | | 45* Teak. oe 31 0 129 20 | 9 2 * Sterile 9 emilee > TABLE lia Crosses between the inbred stock and the truncate stock showing the behavior of sterility in the F,; and F2 generations. Inbred 2 No. 31 X truncate & No. 35 | wo.P: | cugre | grams | Avoumer | came, | rte, | emus | omens | F, Fy 5 F2 ey 25 | 26 Ol ie Fonda. tela, MG ee bey Dab | #36 0 0 0 | | | eee | ear 28 9 4 4 | 25d 18 6 4 7 | | 25e 33 3 26* Att eee 28* 7, | Ne eM EAL Nils 9 P| ee | eo eS a 2 Totaki?.<2.] (kG as eM Onaiesy eenle 24 ul 7 * Sterile 9 FERTILITY AND STERILITY IN DROSOPHILA 155 TABLE 15b Inbred & No. 31 X truncate 2 No. 35; reciprocal 15 a No. Pi aeetee saad sk ia paeiaD ae sae ahead te Soiieas Fi F; F2 F2 31 32 33 8 0 33a 46 4 2 33b 29 15 a ily 33¢ 48 9 7 34 19 0 34a 24 1 1 0 34b 39 3 2 1 34¢ 12 5 2 3 35 11 0 35a 48 6 3 1 35b 31 3 1 35¢ 38 22 9 A Motels: 38 0 5 315 68 34 7 TABLE l6a Crosses between Woods Hole stock and truncate stock showing behavior of sterility ir F, and F2 generations. Woods Hole 9 X truncate 3 aD. a TESTED STERILE Boe | TESTED ereitcn STE Se STERILE 18 35 10 0 18a 7 1 1 0 Bebe 1 0 0 0 19 35 15 0 19a 22 2 2 0 19b 12 z 20 26 9 0 20a 15 8 4 4 20b 17 2 2 0 21 26 18 0 a a 0 0 0 aie: | 85 1? | Fie 8 0 0 0 22 35 fh >. 29a || Sead 2 1? 1 O91 |) ume 4 4 0 23 35 OP es Pog wear || 7 0 0 0 | | 23b 5 0) 0 0 24 Se Os Pie ators. |e Om: 23 oF De a0 0 | | 24b 13 2 ee 1 See een | | Total:....| 148 | 1 | | 243 25 15 6 * Pairs chosen at random from the above crosses. 156 ROSCOE R. HYDE in the germ plasm we should expect it to reappear only slightly intensified in this case among the grandchildren, A study of the 528 pairs tested in the F, generation verifies the expectation. This case shows that sterility as it reappears among the grand- children bears a more or less definite relation to the degree to which it was put into the cross. The behavior of the sterility of the truncates at this time is significant for their earlier history shows that half the pairs were sterile. As stated, it was at first my opinion that sterility was inherent in the strain, and that it bore a causal relation to low productivity. But the above results seem to show that sterility and low production of fertile brothers and sisters are two entirely different things. Such a TABLE 16b ‘ Woods Hole & X truncate 2; reciprocal 16 a a Se | esreD STERILE a TESTED sTPRtLE SEES ae 1 35 13 0 la 12 Lie 2? Ly 1b 25 1 0 1 2 35 0 0 0 0 0 3 35 10 0 3a 14 0 0 0 3b 7 1 1 0 4 Bi) 12 0 4a 16 ili 1? 0 4b 20 1 1 0 5 +35 25 0 5a 20 2 2 0 5b 19 0 0 0 6 35 Uf 35 8 33 4 ld 8a 14 1 0 ult 8b . 38 1 1 0 9 33 10 26 11 0 10a 19 2? 1 10b 6 Die 1 10¢ 17 0 0 0 11 26 i 0 lla 8 1 0 i 11b 11 0 0 0 12 33 8 0 12a 13 0 0 0 12b 26 11 0 11 13 34 14 34 5 0 17 35 8 0 Totalee 103 1 285 27 8 17 FERTILITY AND STERILITY IN DROSOPHILA 157 view paved the way for a study of the low production of the trun- cates, which is dealt with in the second part of this paper. It is to be noted in 12 b (table 16 b) of this cross that 11 pairs out of 26 are sterile. This proved to be due to the males. Twenty-eight other males from this family were tested with the hope of establishing a sterile male strain. Seventeen of the 28 were sterile. This gave promise of yielding a sterile male line; but on inbreeding a few pairs the sterility vanished as suddenly as it had appeared. This is the only case in all the experiments in which sterility affected the males to any appreciable degree. Let us turn now to the evidence which this experiment gives on the transmission of female sterility. A study of the tables where the sterile female strain (inbred stock) was tested against the truncate and Woods Hole stocks shows that the sterility reap- peared again after skipping a generation and that it affected primarily the females. The defect reappears in both the cross and the reciprocal. A great deal of weight is to be attached to this evidence since the sterility did not appear to any appreciable degree when the truncate stock was crossed to the Woods Hole stock. Moreover, the sterility as it was affecting the female was now at its maximum. In table 14 a there are three inbred males that did not transmit the defect. An apparent explanation is found on the assumption, which is borne out by all the evidence presented, that the defect behaves like a unit character and accordingly some families are not affected. Crosses between inbred and truncate stock Before proceeding to a summary of these sterile cases I wish to give the results of an experiment in which the inbred stock was crossed into the truncate stock. Sterility was present in both stocks at this time. Tables 17 a and 17 b give the result of the sterility as it appeared in the F, and F, generations. Table 1, F3, gives the control for the inbred stock while table 8, Fs, gives the control for the truncate stock. The inbred strain up to the time this cross was made had been relied upon as a fertile strain. That sterility had appeared in this stock and 158 ROSCOE R. HYDE TABLE 17a Crosses between the inbred stack and the truncate stock showing the behavior of _ sterility in the F; and F2 generations. Inbred 2 X truncate 3 xo. eee eet | no eee a ae ee Fy F; F2 | F2 18 3 0 Tsu laa 19 3 0 6 2 20 | 6 0 9 3 21 2 0 0 0 Doe ays 24 4 Os | haiti 1 2 Dee 26* 27* | 28* 29 | 8 il | 33 0 30 | 6 0 | 14 3 31 | 0 OF hag 0 32 Lathes Ore ile ke 1 33 | 1 0 0 0 34 7 0 3 1 Whabales cette es ho! #9 1 84 14 1 2 * Sterile TABLE 17b Inbred & X truncate 9 ; reciprocal 17 a os Semen ff) emcee learn |) ome | eee Fi Fi F2 F2 1 7 1 4 3 | De | 3 6 0 8 5 | 4 0 0 0 0 5 6 0 22 4 1 6 0 0 0 0 7 10 (0) 4 4 4 8* 9 11 0 11 5 10 13 2 28 3 1 | ial 12 7 0 17 G 0) 2 13 4 0 2 0 14 6 0 | 3 1 | Totals. es eee ae 70 3 99 | ey 5 3 * Sterile FERTILITY AND STERILITY IN DROSOPHILA 159 that it was also present in the truncate stock at this time is shown not only from the controls but also from the crosses themselves. It looks at first sight as if the sterility reappeared in a higher de- gree in the reciprocal than it did in the cross, since it occurs in 32 cases out of 99 in the reciprocal while in the cross the sterility occurs in 14 pairs of the 84 pairs tested. It is to be noted, how- ever, that the pairs as made up are not equally distributed, for, number 29 of the cross has 33 representatives none of which are sterile. If correction is made for this we have 14 sterile pairs in 51 or 28 per cent which corresponds fairly well with the 33 per cent of sterility as it occurred in the reciprocal cross. The fact that 33 pairs of number 29 showed no sterility is significant as has already been pointed out; for, it shows that some families do not show the defect, while in closely related families the defect may be present to a high degree. In table 18 I have compiled the results of all the experiments that deal with the transmission of sterility. A census of the whole situation shows that of 417 pairs set aside as controls 113 pairs were sterile. Twenty-seven per cent of the controls were sterile. There were 84 sterile females and 14 sterile males. Of the 149 crosses made between the sterile and fertile strains 35 pairs, or 24 per cent, were sterile. If the sterility behaves any- thing like a recessive character the expectation is that sterility will not appear in the F, generation. A total of 832 pairs tested in the F, generation verifies the expectation as only 19, or 2.3 per cent, are sterile. These seem to be chance occurrences as both sexes are equally affected—5 females and 6 males. If one could find an absolutely fertile race against which to test the sterility, it is altogether probable that no sterility at all would appear in the F, generation. The sterility as it appeared here in the F; generation apparently gives a fair measure of the degree to which the sterility entered into the experiment from the strain against which the sterility of the sterile strain was tested. In the F, generation we find that the sterility as it affects the female reappears. Of the 2644 pairs tested, 407 pairs, or 16 per cent, proved sterile. There were 237 sterile females and 60 sterile males. The rather large number of males given here in- HYDE ROSCOE R. 160 éc+¢ | HE 89 | STE 0 0 0 sé |0 0 0 |S ee ve x G T € IZ | SE “ON (8) OL | Ae (1g) OT j PE LST oh Lg te JON (8) 1 y x 2 il) wee Ter 0 | 0.) O } 8 10 vy |e ————(te) SL rte | fe | Ts | cot | tecoN G 6 0G | 6cT 0 0 0 Té 0 @ 6 |g eee LL i 5G lai OT Té | TE ON = x 1éI+8 | LP 1g SOL ye < eer (2) 11 ai 2M Fa x I OF Sn 8¢ T I G GL — == T ESN eh I ta oN 9 \éI+PT/2I+F2] Shs 0 T I HAS (== = NO) pe oe x I 0: Sit 8¢ es | (1 BA SL bak x Bo) EB ap e ES+STLE+S \48+61] $8 I 0 I COs 10" Ste. | 2 | t | wie | —| —| rt] w& Elo wt ——é1 | | LT | = S.L | I A x aw = ai OF € g cé | 66 = = ts Oa == a6 si |——o7, | 9yt9}8 poyeur ayes | poyeur |] 3 | 4B | 2 | 2 | Beane ame aoe le es aria ae be ear Pal i 4 aTaas ae ae BO 5 (BO) SERRE A ATUGES VUAALS Bete ee -Iuaddxa@ dO ‘ON NOILVUANGD FT NOILVUANGD Ly a | 0 5 | 8 -Noo 919H spooy ‘M “paique ‘7 faypoun.y saznowpur J, ‘sjuamisadxa uasaf{ip ay} Ur paanyag 72 sp Aizyrsags fo hanwuns Buinoys 8I ATAVL 161 ‘HOF ‘S}UoUITIedxe 9a1Y} 9S0y} UI pojse} sured Jo JoquINU [B4OJ, ‘PLE ‘SJUSWITIOdxs 9914} VS9Y} UI Sesvd d[I1048 JO Laquinu [BJO], FERTILITY AND STERILITY IN DROSOPHILA ¥G | SST | 8he | 9SPT L IP 1 gaat b=) a Rd || a [wood toad UT [BIO T, | 9¢ | 62 | 6ST | S8IT ZI LOPS IELHSO™|) eam RN ar eee: SSO10 UT [BJO], 9T 'G BG UG "* "+990 10g 09 | 28% | LOF | F792 61 ZEs |Z |9 (ce \6rT FI $8 STL 2 | aS eA éT+2 [41 +91/22+22| 292 z ZO | — Ir (eo W-——(IST) 61 0 ¢ G IZ | ISt “ON Be x ne) ey I G re 6g | OTT oti A M x z ie P OL 2I+T [48+% 86+9 | SST if SIT 0 |0 0 (er Pare a G +9 94 | 20E g ee |i— 8 Of M———(221) oI z OT ral 8% | S21 ON he x NS ae LT G 23 68 OIT aq Ut G M WA Pao ST G P OL Z €Z |49+92| zgz 7 9 |—|I—le lt —C——é6M J z ST [gI+02) 121 0 2 Nt es ee I OT II 9% | 6FT ‘ON x a (6F1) MT ¢ tS 68 OI x ® III VA ae wr De z re P OL F 9% | ee | 99T i1+11.9% 0 0 0 ¢@ ———oM poenuryuoh—sT AIAV.L 162 ROSCOE R. HYDE cludes 15 sterile ones that occurred in one family and was probably due to a chance combination. A glance at the summary given in table 18 will show that the sterility as it affected the female is practically the same whether inherited through the male or female. I think that the 574 cases of sterility recorded in this summary of 4042 pairs tested gives a faithful picture of the actual facts in the case as seen on a minor scale in each of the experiments and in the history of the inbred stock. The evidence is conclusive that sterility as it affects the female may be present in different intensities in the different families and that female sterility is transmitted by the fertile brothers and sisters of the affected females to the granddaughters but not to the grandsons. This character behaves like a recessive Mendelian character in that it disappears in F,; and reappears in Be: CROSSES INVOLVING A SEX-LINKED FACTOR In this experiment, which unfortunately was not carried out on a very extensive scale, the white-eyed stock was used. This stock arose from a mutant of the inbred stock. Females were TABLE 19 Behavior of sterility in F2 generation of No. 12 b; control of the supposed sterile male strain NUMBER 1 | 2 Sod Number pairs mated... Je Oe | 8 | 12 3 11 3 | 43 Number pairs sterile..... 0 | 0 1 | 0 0 2 0 | 3 paired with the fertile males from No. 12b, table 15. The recip- rocal cross was not made. Family No. 12b arose as a cross be- tween the Woods Hole and truncate stocks. One-half of the males from this particular combination were sterile. It was my hope that this would develop into a male-producing strain; but the sterility vanished suddenly. A number of flies from No. 12b were allowed to breed in order to obtain the next generation. From this eighteen pairs were FERTILITY AND STERILITY IN DROSOPHILA 163 drawn off and tested in F, generation. Much to my surprise, all proved fertile. From 5 of these, seven pairs were made up in the F,, as shown in table 19. As a control of the white-eyed stock eight pairs were tested all of which were fertile. Fifteen males, brothers of the sterile males whose fertility had been established, were crossed with 15 white-eyed females; all the pairs proved fruitful. Offspring from five of these crosses were paired as shown in table 20. All the males represented in the 62 pairs in the table have white eyes like the mother, while all the females have red eyes like the father, since this is a case of sex-linked inheritance. On inbreeding, four classes of offspring will be produced in the next generation: (1) white-eyed males; (2) red-eyed males; (3) white-eyed females; (4) red-eyed females. The four classes were about equally represented. The virgin fe- males were placed with a number of males; the males with a number of virgin females. The result is given in table 21. TABLE 20 Sterility as it appeared among the children of the crosses; F' generation | NUMBER | | | TOTAL 1 2 3h kl A aad od Number pairs tested . . 16 21 5 | 6 13 62 Number pairs sterile . . | 3 9 0 | 1 | 0 13 TABLE 21 Sterility as it appeared in the F2 generation. Offspring from two of the 21 pairs of No. 2, table 20 Family No. 2 a Family No. 26 CLASS | CLASS | Red | Red | White | White "| Red | Red | White | White Cee Gh We | ot 2 fou Number | Number | | tested.-..| 18 19 | 10 10 tested....) 14 215) tS 20 Number | Number | sterile....| 3 ie 0 sterile....| 2 |10+1?| 1? 13+2? 164 ' ROSCOE R. HYDE The figures given in this experiment are very small and the source of the sterility doubtful since the controls throw no light upon it. It is to be noted, however, that both classes of females are affected in family No. 2b. Im the case of the white-eyed females more than 50 per cent are affected. To analyze the evidence:from the foregoing experiments let it be assumed that a gene responsible for the functioning of the oviduct is carried by the X-chromosome and that in the sterile individuals this gene has changed so that the oviduct fails to function. In analogous cases when the female is affected both X’s are affected but she remains normal so long as only one X is affected. When the affected X goes into the males they should not be sterile because the male has no oviduct. But such a male will be able to transmit the defect. To follow the argument, let us make three assumptions. a. If the defect in this experiment came from ie white-eyed grandmother then her germinal make-up would be expressed by the formula wwXx, in which the x represents the affected cromo- some. Since the red-eyed male did not contain the defect his formula is RXO. Crossing these two individuals gives: wx—-wxX RX-— O- Gametes wRXx—wRXX—wxO-wX0O F;, Red-eyed females and white- eyed males, none of which show sterility. If by chance the flies chosen have affected genes, we get: wx-—-RX a ek! O Gametes wwxx— wR Xx—Wwx ore RX oO Flies of the F2 generation; only white- eyed females are sterile But this does not agree with the facts of the experiment. b. If the defect was brought in by the red-eyed grandfather and the female is free from it, then it follows: wX —-wX Tage mest Gametes wRXx—wxX0O. Ff, Red-eyed females, white-eyed males: all fertile. FERTILITY AND STERILITY IN DROSOPHILA 165 On mating the F, the following results: wX—Rx Gametes wX-— O wwXX —-wRXx-wXO —-RxO F, White-eyed females, red-eyed females, white-eyed males, red-eyed males; no sterile individuals. This does not accord with the sterility as shown in this experiment. c. If the defect is brought in the X-chromosome of both grand- parents, it follows: wx—-wxXr eG Gametes wRxx—-wRXx-—-wxO-—-wxXO Fi Red-eyed females, white-eyed males. Half of the females sterile. On mating the F, the following results: Pees Gametes wx-— O wwXx—-wRxx—wXO-RxO F, White-eyed females and red-eyed females; white-eyed males and red-eyed males; red-eyed females sterile. This accords with the appearance of the eye colors but not with the appearance of the sterility, since there was practically none in F,. These three assumptions exhaust the possibilities on the hypothesis that the defect is carried by the X-chromosome and yet none of them meets the facts in the case of this experiment. I conclude, therefore, that sterility as it affected the female in this experiment was not transmitted by the X-chromosome. There is a possibility, however, that the factor involved is in the sex chromosome but so far from W that it crosses over freely with it. DISCUSSION : The foregoing series of experiments brings out the following facts. (1) Sterility as it appeared in the inbred stock affected primarily the females. The males may also be sterile but it seems probable that the sterility of the male bears no causal relation to the sterility that affects the female. (2) The defect is germinal THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 17, No. 1 166 ROSCOE R. HYDE in character in that it is strongly transmissible through inheri- tance, behaving like a character that segregates. The evidence goes to show that the character in all probability is not carried by the X-chromosome. (3) The sterility as it affects the female can be eliminated by selection. Inbreeding, as such, does not appear to be the causative agent in producing this kind of sterility. A close study of the females revealed the fact that they were not laying eggs. The abdomen swelled until it was very large and in many cases the ovipositor protruded some distance beyond the body. The female could not deposit her eggs. A score or more females were sectioned and these sections showed eggs present in abundance. In fact, a mere examination of the fly after it had been mashed on a glass slide showed in all cases, ap- parently well-formed eggs in large numbers. I have examined scores in this manner after they had been tested with a number of males and except only one doubtful case I have found that all had eggs. The exact nature of the defect cannot be stated until more attention is given to a study of sections, but the defect is of such a kind that it prevents her from laying her eggs. The possi- bility that parasites were the source of the trouble suggested itself. Sections show, however, that these flies are remarkably free from parasites in so far as one can make out with ordinary stains, and the behavior of sterility in transmission, moreover, shows the futility of such an hypothesis. In the case of the fertile female one has merely to bring pressure to bear upon the abdomen to cause her to expel anegg. The sterile females will not respond to such treatment. On the con- trary, the wall of the abdomen will burst, and allow the eggs to roll out, while none will pass through the oviduct. Whether or not the eggs are ever fertilized I cannot say. I have frequently seen the sterile female copulate and I have planted scores of eggs (obtained by opening the female) on banana but never has a single larva emerged.! All attempts at artificial fertilization proved a failure. 1 A fertile female placed on a very poor grade of banana may refuse to lay her eggs for some time. She holds them as long.as she can but after a few days will lay them. FERTILITY AND STERILITY IN DROSOPHILA “Let It is to be recalled that Castle in his cultures found both males and females sterile. Moenkhaus found in his strain, in 64 cases tested, that the males alone were sterile.2, Moenkhaus in attempt- ing to harmonize Castle’s results with his own points out that the two cases can not with certainty be compared since Castle took no account of the emergence of larvae but merely the production of pupae. In other words, Castle’s measure of productiveness was eggs that gave rise to pupae; Moenkhaus’s measure, the emergence of larvae. From my studies this difference seems rather apparent than real since in only one or two very questionable cases have I found that larvae after’ emerging from the egg were unable to develop to the adult stage. I do not wish to be misunderstood on this point. Not every zygotic combination, as is evident from my later studies, results in the production of a well formed fly. In fact, the percentage, as I have been able to show, may be very low. Whether the sperm enters the egg, or having entered the combi- nation dies, is another question. The fact holds in mystrains that if the larvae emerge some of them develop to the hatching stage. I wish to point out that my results and those of Castle are more likely to be harmonized on the asumption that a distinction exists between fertility and sterility. Castle and Moenkhaus seemed to consider that the defect related to sperm and egg structure. I infer from Castle’s paper that he supposed that sterile females occurred more frequently in a strain inclined to low productivity ;. that low fertility and complete sterility as it affects the female are causally related. I interpret him to mean that there is a range in the capacity of the eggs for fertilization by good sperm, extend- ing from zero (or complete sterility) to high fertility. In my strain no such condition is involved; but it is conceivable that the egg-laying power ranges in degree from zero to complete productivity. If so some of the sisters of the sterile females 2 He mentions one exception which from the present standpoint seems signifi- cant. ‘‘In one instance I found among a brood, beside a sterile male, two females that failed to deposit eggs although eggs were evidently present in the oviducts.’’ It seems likely that this is the same defect that I have found so prevalent in my cultures. 168 ROSCOE R. HYDE should be low producers. This would seem to correspond to Castle’s results but in my case we meet with a difficulty at the outset on such an assumption for it appears that the egg-laying power of the affected female is totally abrogated. My evidence will not deny the possibility of transitional stages but it is not clear how an egg-laying mechanism should be able to expel a few eggs and not all. Critical evidence is hard to obtain on this point because low production may be due to a variety of causes. Attention is called to the truncates, in which there was no rise in productivity as the sterile individuals were practically eliminated. The history of the inbred stock allows us to form a more reliable opinion, for the total output of offspring at first was much greater. In table 2, Part 2, we get results that look as though the produc- tivity is running down as the sterile females appear. The table, however, does not convey all the facts of the experiment, for in the F,; generation where 65 offspring is recorded as the average per pair, some of the flies failed to emerge from the pupae. The generations which precede and follow show more nearly the actual facts in the case, namely, a gradual decrease in productivity. That this was a high producing strain in the beginning of the ex- periment there can be no doubt; for in the second generation the average for sixteen pairs is 365 offspring. The low production in the first generation is probably due to inexperience in handling, as half of the females were dead at the end of two weeks. The pro- duction gradually fell despite the fact that facility in handling became more and more perfected. In the F,, generation the pro- duction had dropped to 159 per pair despite the fact that the sterile females had been practically eliminated. This led to an investi- gation of the defect. It was found by isolating the eggs of the females of the F,, generation, which had been placed with a num- ber of their own males, that only 32 per cent hatched. And yet in the same experiment under identical conditions 58 per cent of _the eggs of the F,, female hatched when paired with the males from the truncate stock. Moreover, the males of the F., genera- tion were able to fertilize 52 per cent of the eggs of the females of the truncate stock. FERTILITY AND STERILITY IN DROSOPHILA 169 These facts make it evident that incompatibility of some kind had arisen between the egg and sperm of the inbred stock. It is not evident that this is in any way related to the appearance of the affected females in this strain. In regard to the method of transmission of the sterility as it affects the female, Mast, one of Castle’s students, came to the conclusion that “‘ An an ier fertile male may transmit partial or complete sterility of the female sex as a racial character to his granddaughters though not apparently to his daughters.”’ Since Castle and Mast’s paper deals primarily with the effects of in- breeding, the data upon which these conclusions are founded are very small. As will be seen from my tables, this method of trans- mission has been verified many times, the exception to be taken here is that in all likelihood the partial fertility bears no relation to the complete sterility. The reciprocal cross that led Mast to the above conclusion resulted in one case in the production of females half of which were sterile and the other half were of low productiveness. This case shows that a female of a race inclined to sterility may trans- mit that character directly to her cross-bred offspring. . . . . This difference in heredity through the two sexes would seem to indicate that sterility of the female is dependent upon egg structure, the eggs produced by mothers of a fertile race always yielding fertile daughters. But the eggs of cross-bred females, whose father was of an infertile race, produce some of them fertile, some infertile females. Such a method of transmission is not borne out in my cases. The females transmit sterility to the granddaughters in the same way as the males and it is certain that the defect skips a genera- tion. Castle and Mast’s paper gives us no certain proof as to the defect in the female. It seems to be assumed that she is laying eggs but that these eggs are incapable of fertilization. If this is true her sterility cannot be compared at all to that of the females in my case. In all my experiments, however, I have never found but one or two questionable cases in which the female having laid her eggs proved sterile. In fact, toward the end of the experi- ments it was easy to tell which sex was at fault by observing whether or not the female was laying her eggs. While this was never made the final test it always held good that when the female 170 ROSCOE R. HYDE laid eggs which did not develop the male was found to be at fault. Some of the eggs of such a female always hatched on placing her with fertile males. It may or may not be that Castle’s case and mine are similar since they behave somewhat differently in heredity. Since Castle’s paper deals primarily with another problem, the data upon which the above conclusions in regard to the method of transmission are based are too small to base any definite conclusions upon, since the influence that may have entered into the experiment from the other sex is not with cer- tainty under control. A glance at Mast’s last conclusion will show that the male also may have been at fault. CONCLUSIONS 1. One kind of sterility in Drosophila ampelophila is due to some defect, probably in the oviduct of the female, so that she is unable to deposit her eggs. 2. The defect is transmissible through inheritance by at least some of the brothers and sisters of the affected females when mated to a fertile race, to the granddaughters, but apparently not to the sons or daughters or grandsons. It is therefore reces- sive and affects females only. 3. The process of inbreeding brothers and sisters cannot be held to be responsible for this condition, but probably serves to bring it out When latent in a strain by making the necessary combinations. 4. The character seems amenable to selection and can be made to affect fully 50 per cent of the females or can be practically eliminated by making the proper selections. 5. It seems very probable that sterility as it affects the male bears no causal relation with sterility as it appears in the female. 6. The defect in the female behaves after the manner of a Mendelian character in that it reappears after skipping a genera- FERTILITY AND STERILITY IN DROSOPHILA L711 tion. The normal function is dominant to the negative condition. An unaffected male can transmit something as a dominant char- acter which causes the normal egg deposition of his daughters. 7. The defect is sex-limited in the sense that it affects the fe- male only but it is probably not sex-linked in the sense that it is carried by the X-chromosome. If so, it is such a distance from W that it crosses over more or less freely with it. I wish to express to Professor Morgan my appreciation of his interest throughout the present study. I am also grateful for the generous gift of the Dyckman Fund which the Department voted me for 1912-1913. BIBLIOGRAPHY Cast Le, W. E., CARPENTER, T. W., Cuark, A. H., Mast, S.O. anp Barrows, W.M. 1906 The effects of inbreeding, cross breeding and selection upon the fertility and variability of Drosophila. Proc. Amer. Acad. Arts and Sciences, vol. 41. Moenxuats, W.J. 1911 The effects of inbreeding and selection on the fertility, vigor and sex-ratio of Drosophila ampelophila. Jour. Morph., vol. 22, no. 1, pp. 123-154. Morean, T.H. 1911 The origin of nine wing-mutations in Drosophila. Science vol. 33. PY 7 iy 4 ‘ - : a 7 ; 7" My i. i 4 p =) i ‘ de" ’ SOPRA OS t Ce par, Lima hee sh ge ae ; ihe | a ' it 7 ‘ Tt ee a . . . FERTILITY AND STERILITY IN DROSOPHILA AMPELOPHILA II. FERTILITY IN DROSOPHILA AND ITS BEHAVIOR IN HEREDITY q ROSCOE R. HYDE From the Zoélogical Laboratory, Columbia University NINE FIGURES CONTENTS VL BY TiS DOW g GY OCS a Aca ger at ane AD ge Sn Bar ew ee cb ae 173 - History of the low-producing truncate and the high-producing inbred stocks 175 Crosses between the low-producing truncate stock and the high-producing “PCL STR EE ST ga ee PN AN a oe MARRY Ne”. >. 176 Fertility of the F; and F2 generations from the crosses between the low-pro- ducing truncates and the high-producing wild stocks as determined by breeding the animals together in pairs............ SORE STS te RRR SRS, 182 Crosses between the low-producing truncates and the high-producing wild stocks in which an exact measure of fertility is employed................ 185 Back crosses between the hybrids and the recessive low-producing truncates 194 Hertality and high productivity of the hybrids... .... 0. ....-..::..s25---.-. 203 Fertility of the long-winged brothers and sisters of the truncates........... 206 Bebavior of the truncate wing im heredity:......204.0405.. ease eee ee 209 TTD ORM a aan er eRe Tt COR ECOL OP LR? Vb Bedh otaeeer her 2 Cs ee 211 He GeraGnevCIbeG aaah fet. wh ac Rhea orate eters ete eee aE on nie ae oa 212 INTRODUCTION The first study demonstrated that the sterility there dealt with is due to some defect probably in the oviduct of the female that prevents her from laying eggs. Since this defect is trans- mitted by males and heterozygous females, the defect when once in a strain constantly recurs. It is evident that this kind of sterility in the female is an entirely different thing from infertility in the sense that eggs laid are not fertilized or if fertilized do not ‘develop. , It is evident from Part I that the low fertility of the truncate stock would not be explained by the presence of the sterile indi- viduals for after sterility had been practically eliminated the 173 THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 17, No. 2 Aucust, 1914 174 ROSCOE R. HYDE -truncates produced no more offspring per fertile pair than at the beginning of the experiment when half of the pairs were sterile. Neither can the low productivity of the truncates be explained entirely because of the shortened length of life, although this is one factor in the result. The truncates differ in many ways from the wild stock: 1. The flies give rise to a small number of offspring per pair. Table 1 gives the number of offspring produced through several generations of inbreeding. Despite all the care in breeding that could be exercised no way was found to increase the output. 2. The truncate flies do not live as long as the wild fly. The length of life is in general about half that of the inbred stock which was used in the control. The two strains were bred under identical conditions (compare tables 1 and 2). 3. Truncates, as the name implies, have the ends of the wing squared instead of rounded. The wings extend only to the tip of the abdomen. The flies will not breed true, for they produce TABLE 1 History of truncate stock through seventeen generations of inbreeding _ GENERATION | (Pee ioy Sey F43 | Fas Fis | Fac ie hae Fs Fs | . Faz Fig = Rape aoe ee No pairs tested.. 5-27 --2.| (“Sly)% ad) (546 19 TaD? 2 i) No pairs sterile..........| 227] 27| 12 6} 0 2 30/0 i ae No pairs from which off- | | | | Ye oer spring were counted.... 9 | 26) 34, 13) 9 9 8 Me illge ie Druncate cieks shee! 173 | 855| 684 | 266 | —| 79|182 | 182] 218 Truncate 9... 02.22... 212 | 911} 696) 292] —| 99202; | | 162] 228 WiGueEOL st oe ce See 41 | 123] 136} 29} — 10 37! 31| 66 Dowegor 62S ieee 43 117} 102) 32 | —| 14) 25 DA | © ae Total @...............| 214 | 978| 820} 205 | 14! 89219 | 213 | 284 Total 9...............| 255 | 1028; 798 | 324 | 14113/227/ . | | 186 | 285 Grand’ total... 220% | 469 | 2006 | 1618} 619 28 202 446 | | 399) 569 ee Sn — = l= a = = nee : Average........... t..] 52 | 77) 47| 48 | 28 22 56 | Bry eal Average life of @ in | | | | | GAYS ise t? pak ak tsk sent al eet SG =A hoa 26| 26 BOF 23 Average life of 2 in | days. {kins bc eke) et es 0G ae el gs I cal @ty> caslie FERTILITY AND STERILITY IN DROSOPHILA 175 some offspring with wings like those of the wild fly; at least this has held true through many generations of inbreeding truncate brothers and sisters. The long winged flies from truncates throw in turn some truncates. HISTORY OF THE LOW-PRODUCING TRUNCATE AND THE HIGH- PRODUCING INBRED STOCKS In order to bring into sharp contrast the difference that exists in the productivity of these two strains I have compiled in tables land 2 the data of these two strains through several generations of inbreeding. The truncates gave on an average about 50 off- spring per pair, and the inbred stock about 200 offspring per pair. It has been stated that the inbred flies live longer than the trun- cates, but this difference, as the tables show, will not account for the marked difference in production. Table 2 gives the history of the wild inbred stock described in Part I. This was bred parallel to the truncates in table 1 and was subject to exactly the same environmental conditions. TABLE 2 History of the inbred stock | GENERATION | | Ti oy ar cin, Spee | | aa eT ee | ry i. ts | Fi Fe Fs Fy Fs Fs F; | Fs | Fo ae Fu | Fi: Fu ee es oO L ge NE Sa Ea No. pairs tested| 16| 16 Ae, E20) Ce LOn MOS) = "Se 100 40 11 182) 12 16 No. pairs sterile i datge 0| 3 3 2) 51) 15 | 39) 6 0 28) 1 1 No. pairs from | | | | reas} | which off- | | | | spring were. ray | counted... ....-. 13 PS A ST oN a ee Se a ee dt Males produced § 975 2374 439, 854 | 708 336 629 1098 Females _ pro- ey Gucedierss. «4: | 1034 | 2406 395 859 766 910) 677 1142 i =a ee a | | | wale = s Total. | 2009 4780 pee 1713 | 1474 646 1306 2240 = = eS = ——|——_| — | ass Average.......| 155" 368 209 190) 184 65+ 119 | 159 Days o@' lived...) — 45 | 42! 41 36 | 45) 34 | 45 Days Q*lived...| — 37 | 35) 33 37 | 35, 30 | 26 * The ee oaneon in this case was due to inexperience in handling. Half of the females were dead at the end of the second week. + Some of the flies failed to emerge from the pupae. 176 ROSCOE R. HYDE CROSSES BETWEEN THE LOW-PRODUCING TRUNCATE STOCK AND THE HIGH-PRODUCING WILD STOCKS The following experiments were carried out in order to determine whether the egg or sperm was responsible for the low fertility of the truncates. Females from the low-producing truncate stock were paired to males from the high-producing inbred stock. The reciprocal cross was also made. Since the number of offspring bears a more or less direct relation to the length of life of the pairs, data on the length of life is also given. Tables 3, 4 and 5 give the result of the first experiment. This experiment shows that the low-producing truncates were producing an abundance of fertile sperm, since each was capable of fertilizing on the average 428 eggs of the female of the high- TABLE 3a ’ Showing crosses between the low-producing truncates and the high-producing inbred stock x Truncate ? X inbred moe al ener | ome | a a a ee eee ae 1 1 B16 27] OL |” 419") aoe | 7126 hase) ine haa Shh a | Mice alee el eee | 4 3 G1) 0} pe8e) erie Saal” Meri 15 31 4 Po a ee a ee 15 5 | 0.1 i AAC GS SSG emis soon meals se 6t Pe a a a 43 | 8 ole Bi 864° “S4al) 938K BBs eon at 7 gt i ee ee alee 7 10 | 10 6| 185| 187] 195| 193) 388 56 | (6 11g 9 |.) cea Reel eal cl meal cold aban fs oa i 12 YOR ON hes) OBI SE stat 13) 2247 as 35 14 Let) (ele SB0uh RBI. Seal 65 1b Ge ane ceeds 16 i Gal oul o7S8a|) Oss amie4a 105") A80 37 24 17 Ne We Pe es: rs) aS Re ae 5 Total. 0a: | 41| 35) 824| 884| 865| 919| 1784) 496 | 267 Average..... ; oo a = — —]| 198 38 19 * ‘Crescent’ as recorded in table 3 are longwinged flies with a small crescent- shaped piece cut out of the inner margin at the end of the wing. They are a new class that arise on crossing the truncates with wild stock. . + Sterile. FERTILITY AND STERILITY IN DROSOPHILA 177 TABLE 3b ‘ Reciprocal: Truncate ox inbred 9 2 me, piel cave LONG Fa nae wae cEAnD Live or 3 ewnion 2 18 5 Dil Soe) wolOn) sa 343 | 680 49 59 19 5 4 130 136 135 140 | 275 16 53 20 0 PA SS 335 283 337 620 | 66 66 21 0 0 191 173 191 173 364 | 33 ts De = — — == — — a — a wey = — = = _— — — 63 63 24 5 lg 185 189 190 206 | 396 26 24 25* = — = = aye Wer — 41 11 26* = = == — —| — = 33 41 28* = — — Ss — - — 65 16 29 2) | 9} 302 293 304 | 302 606 33 49 30 0 0 86 118 86 118 | 204 37 37 31 2 4) 263) 284 265 288 | 553 | 36 55 32 0 (Hel) “2py 249 222 249 | A471 36 53 33 it 20 127 126 128 146 | 274 36 67 34 3} 1 1338 131 136 132 | 268 10 le 44 Total..... _| 23] 84 | 2254 | 2350 | 2977 | 2434 | 4711| 598 | 659 Average..... — — — — — —| 428 40 44 * Sterile. a f TABLE 4 Control for the high-producing inbred stock used in Experiment I, table 3; compare with table 2. NO. tt aetna, MALES FEMALES LIFE OF o LIFE OF 2 10 45 28 7 26-+- 26-++ 11 123 59 64 26+ 26+ 1133" — — — 35 26 14* — = -— 44 26 16* —_ — —- 40 15 17 311 159 152 40 40 18 227 106 121 15 55 19 198 100 98 40 40 20 340 163 177 66 57 21 46 21 25 26 26 22 192 108 84 66-- 34 23 231 115 116 66+ 26 Mo tals wy. si: 1713 859 854 490 397 Average. ?.. 190 — — 41 33 * Sterile. 178 ROSCOE R. HYDE Truncate xTruncate Inbred x inbred 48 % cy : 190 1978 428 Diagram A Effect on productivity of crossing truncate and inbred stocks; summary from tables 3, 4 and 5. producing strain, in fact, more than twice as many as when she was fertilized by her own male. The female of the low-producing strain was evidently producing a large number of eggs that were capable of fertilization as shown by the fact that on an average 198 of her eggs were fertilized. And yet the truncates used for con- trol only produced an average of 48 to the fertile pair. The TABLE 5 Control for the low-producing truncate stock, used in Experiment I, table 3; compare _with table 1 ee ia j NO OASIS | Bic @ yee | uae, BONE | EO ae LIFE OF | LIFE oF 2 | 1 oo") | cg a eter ihn es 25+ | 25+ 2 1 1) cet igen nea nl 9 16 30+ 30+ 3 6 Wi) 29) vi 59. | Bah Mba en ane 157 30+ 30+ 4 Lo) 20°).0389 1) 40) 3a go aaa Se to 30+ 5* a | a 34 34 6 0 Oe) A). 19) aig Neat 36 - 37 14 7* | | | | au 31 10 8 0 0 Mi) 12h 7a Saale Sa 14 14 9 9 | 137} 39 | 46°) 48 | 597) 107 30+} 30 11 0 0 AS ke GS 4 3. | 7 23 23 12* yi el all ee 14 8 13 1 0 1 A oe 6 Gr eel 8 14 2 Die |) AG eel AG a Ag arenes 95 42 42 16* a ea el — 32 32 17 0 1, Yh) 428s 9) LON mee eet S20 Yi 42+ 18 0 1 2 0 2 | a i 16 19* SS A) |) a ee Pee eS 19 20 i O25) 13) a oemtene | AB. |) BT ee Ales we eee | Gh LE |, Si Total.’...),) 29 || 329), 266°-) 292) | 29517) 3245) 619 517 464 Average.. | = | = oa Bae Pars | 48 278 | 25 * Sterile FERTILITY AND STERILITY IN DROSOPHILA 179 Crosses averaged even better than the high producing strain. Both were mutually benefitted by the cross, but it may seem that the production of the truncate female was limited because she lives on the average about half as long as the inbred female. The truncate male lives on the average about ten days longer,—long enough apparently to fertilize all the eggs of the inbred female. It should be stated that I have verified Castle’s observation many times that more than one copulation is necessary to fertilize all of the output of eggs from a female. It seems that the male lives on the average long enough to fertilize this second output of eggs. It is evident that the short life of the truncate female does not explain entirely her low production, for if the results of this experi- ment may be relied upon, there was evidently incompatibility of some sort between the egg and sperm in this strain. The results of this experiment were very surprising in the light of Castle’s work. Castle had made crosses between a low-pro- ducing strain and a high-producing strain with the result: ‘‘that low productiveness (or sterility) of the female may be transmitted directly through the egg from mother to daughter, but only indi- rectly through the sperm, the character skipping a generation.” Castle’s conclusion in regard to this case lends support to the assumption which I made in Part I, viz., that he did not dis- tinguish between complete sterility of the female and the low fer- tility that is here shown. . The problem of the increased fertility of the truncate stock when crossed into a high producing strain was now. put to another test. The experiment was carried out in the same way as the previous one except that the Woods Hole stock was used for the high-producing strain, as it had been observed in other experi- ments to give rise to a rather large number of offspring. ‘Tables 6, 7 and 8 give the result of this experiment. At first sight this experiment is not as striking an experiment as Experiment I. Neither cross did as well as would be expected from the controls. Nevertheless, the same relation is evident, for while the truncate female produced an average of 56 offspring by her own male, she produced on an average of 118 by the Woods Hole male. It is to be noted, moreover, that her length of life on 180 ROSCOE R. HYDE the average was only 15 days in the cross while in the control it was 25 days. It seems, not only from this, but from the other evidence, that had she lived on the average ten days longer her output would have been much greater than 118. The truncate male on the other hand was able to fertilize on an average at least 245 eggs of the female of the Woods Hole stock. But this length TABLE 6a Crosses between the low-producing truncate stock and the high-producing Woods Hole stock. Truncate 2X Woods Hole & NO. cane ces ar on S Tae tegie eietanen LIFE OF o’ | LIFE OF 9 1 iLL 13 152 | 169 | 163 | 182 345 25 24 2 8 3 28 64 SOP Od 103 19. 13 3 1 0 49 | 61 50 61 111 15 13 4 2 OP SOS eels saelOS 221 23 14 5 3 1 20 31 23 32 55 | ? 4 6 2 0 15 25 ily 25 42 29 14 ‘if 1 0 16 19 Ifa Wel) 36 34 13 8 1 0 28 36 29 36 65 115) 28 9 2 0 42 48 44 48 92 23 17 10 3 2 43 | 62 46 64 110 38 iil Total 34 19 | 504 | 623 | 538 | 642 1180 221 151 Average..| — ao a eas | = 118 25 15 @ TABLE 6b Reciprocal: Truncate o&& Woods Hole 9 NO. nad eae ea roe One aooF rept | LIFE OF c’ | LIFE OF Q 1 in Uae late bois ioserbroealhpeonenel| Maran 29 12 3°| 8 |150 | 150 | 153 | 167 | 320 17 38 13 O ea | fer.) 124? I Sazailaen all e192 17 23 14 1) s@eisias!() 157 aon dagen! | 206 21 38 15 5 | 9 |124 | 155 | 129 | 164 | 293 17 25 16 1) MON See M71 lel Grae) 2128 21 19 17 9 | 18 | 112 | 90 | 121 | 108 | 229 18 20 Total.....| 30 | 49 | 764 | 871 °| 794 | 920 | 1714 | 198 192 | . Averages icin ii ra == SoS 245 18 28 FERTILITY AND STERILITY IN DROSOPHILA 181 of life is much shorter than in the controls. In other words, despite all of the unfavorable influences that entered into the crosses in comparison to the controls the point at issue is plainly evident that the female is producing a large percentage of eggs that are capable of fertilization, the male an abundance of good sperm, and yet when the two meet each other there is incompati- bility of some sort between egg and sperm, as inferred by the num- ber of offspring produced. It is clearly evident that the length of life isa factor, but by no means the only factor, involved in the low TABLE 7 Control of Woods Hole stock. Brothers and sisters of the high-producing strain used in the cross in table 6 No. ee MALES FEMALES nineton’c' anemone fo) 1 | 244 iil 133 30 | 30 2 | 527 263 264 AQ) 36 3 351 185 166 26 57 4 347 155 192 21 42 5 249 116 133 22 23 6 287 149 138 30 33 Th 290 139 151 34 60 8 275 126 149 44 30 ROU) eee 2570 1244 1326 247 311 Average....... 321 155 166 31 39 TABLE 8 Control of truncate stock. Brothers and sisters of the low-producing strain used in the cross in table 6. NO. eerie. io. € aa ‘cane oe eee pa LIFE OF o | LIFE OF 9 1 1 0 1 0 | 0 2 De 25 2 4 6 36 51 40 57 97 2 32 3 4 3 15 Wy 19 20 39 21 24 4 1 1 17 19 18 20 38 26 23 5 3 4 15 17 18 PAL 39 26 ily7/ 6 14 ff 39 43 53 50 103 31 25 7 8 0 19 | 21 27 21 48 17 21 8 2 4 AQ | 934 42 38 80 30 30 Total 37 | 25 | 182 | 202 | 219 | 227 | 446 | 208 197 Average..| — — a Se 27 28 56 26 25 182 ROSCOE R. HYDE Truncate x Truncate Wood'sH. x Woaod's $ ? 5b gb g 32l 18 245. Diagram B_ Effect on productivity of crossing truncate and Woods Hole stocks; summary from tables 6, 7 and 8. production of the truncates. To judge from the history of these strains and from Experiments I and ITI it would seem as though only one egg in four or five of the truncates gave rise to a mature fly when fertilized by its own sperm. I shall not analyze the data further here as I shall later bring forward better evidence that bears on this relation. FERTILITY OF THE F; AND F, GENERATIONS FROM THE CROSSES BETWEEN THE LOW-PRODUCING TRUNCATES AND THE HIGH- PRODUCING WILD STOCKS AS DETERMINED BY BREED- ING THE ANIMALS TOGETHER IN PAIRS We may next consider the evidence that bears on the behavior of the hybrids in the F; and F, generations. In order to deter- mine the productivity of the hybrids from the cross between the truncate female and the inbred male, 65 pairs of the F, brothers and sisters were made up from the crosses as given in table 3a. From the reciprocal cross 47 pairs of the F, brothers and sisters were paired. The different families were about equally repre- sented. Both F; hybrids proved to be most virile animals and excellent producers. In the thirty days from December 22, 1911, when they were first paired as virgin flies, to January 22, 1912, when the experiment was discontinued due to an accident the sixty-five pairs produced a total of 16096, an average of 248 offspring to the pair, and the 47 pairs of reciprocals produced a total of 11800, an average of 251 offspring to the pair. The differ- ent classes that appeared are given in table 9. The experiment shows that the hybrids are excellent producers and that it makes no difference whether the father or the mother was of the low-producing truncate stock. Moreover, it seems as FERTILITY AND STERILITY IN DROSOPHILA 183 TABLE 9 Showing the n number of offspring produced by 112 pairs of the hybrids 1 in 30 days | at | TRUN- | TRUN- | CRES- | CRES- | | | CATE CATE CENT CENT LONG LONG | TOTAL | TOTAL | GRAND |) AVER- ge | 19 a rot SEAL peices! g TOTAL | AGE 8 | 431 | 633 | 6881 “6857 | 828 | 8268 16096 | 248 2 57: Fi (T2 XI’)..| | i 27 | 2) 315 436 | | 4986 | 5458 | 5574 | 6226 | ee) 251 Hecho LS) .2) % = ( 5 though the hybrids are long lived as shown by the fact that in this experiment at the end of thirty days 81 per cent of the flies were still living. Twenty-four males and 31 females had died. The death rate was not selective but was about equal in both the cross and its reciprocal. Table 10 gives evidence on the same question—the output of the hybrids as tested in pairs. In this experiment the hybrids TABLE 10 Showing production of the hybrids of the crosses between the truncate stock and the Woods Hole stock TABLE 10a TABLE 10b F,; generation from truncate @ X Reciprocal cross: F; generation from Woods Hole & truncate < ox W oods Hole 9 NO | TOTAL | TOTAL pete LIFE | LIFE No. TOTAL | TOTAL ore LIFE | LIFE “Vk ae Q TOTAL | OF co | OF Q Chimes TOTAL | OF G’ | OF 9 Mey 83 l47,| 27 304 ea eeo:|) ea es | 45 iil Pee 98s) 180%" 228 20} 24+ 33 92; 124) 216 57 | 24 4 96 | 119] 215 27 | 33 34 92) 105 | 197 40 | 24? 5 103 95 | 198 5A) 27 35 70) 094.\ 9 164) 1622s 8 | 143) 169) 312 24| 35+ av | 117 | 143). 260 33 | 27 OF “90117 | 207 20 | 24 38 | 192) 198) 390 33 | 45 10 96 | 110} 206 27 | 38 40 147 | 228) 375 24) 33+ Ey LOO), 40s) 240); 45 | 24 41 | 99) 129) 228 37 | 15 ES eoSal) 67.) 120)| > $20) | 15 AD’ Nek 54s yay 28) | e206 ae he) 27 14 |) 167) 138) 305°) 40) 27 43 | 104) 113 | 217 12 | 24 LO LOL 139 , 240) 40, 33 44 |} 106) 127) 233 40 | 20? an eal gil eat Sx | visage’ 54. |) 54 46 66| 79| 145| 33+) 27 1g) |) 149'| 150)| 292) 63 | 20 AT a Me Ag Ose az ee O00 te 1a S700") 187 33 | 49 49 219 | 269 | 488 24 | 63 20 | 126 | 1463) “272 27 | 33 50 | 112) 126)» 238) 45 hp ae 22 | 238 | ue $89 )|039)| A. «herein na leeeeme mene ee eee Total | 2053 | 2348 | 4401 | 593/540 Av. | | | 2380) 32) 27 ee | | 259 Sra gD 184 ROSCOE R. HYDE between the truncates and the Woods Hole stock were used. The hybrids came from the cross given in table 6. In the experiments given in table 11 the hybrids came from crosses between inbred and truncate stock, as recorded in tables 15a, b, Part I. At this time the inbred stock was in the F; generation and was a high-producing strain. The truncate stock was in the Fis generation. The results of these three experiments make it certain that the hybrid offspring from the low-producing truncates and high pro- ducing wild stocks are high producers. © In fact, the hybrids produce equally as well, if not better, than the high producing parental stock. Moreover, it makes little if any difference whether the father or mother is from the low-producing truncate stock. The following experiments give the evidence that bears on the question as to whether the low rate of fertility of one of the parents will reappear in the F, generation. Tables 12 and 13 give the results of testing in pairs the offspring of the hybrids, derived from the crosses between the low-produc- ing truncates and the high-producing wild stocks. The pairs as TABLE 11 Hybrids of the crosses between the truncate stock and the inbred stock; the F, gen- eration TABLE lla TABLE 11b The F, generation from truncate Reciprocal: The F; generation from 2X inbred 0 truncate o' X inbred 9 wo. | mares] F®.|roran| 2% | HME wo, [aan | corns noran | UPS | LIFE 7 | 294} 295] 589 42 | 39 1 184 | 256 440 | 58 | 48 8 | 180| 224) 404 56 | 41 6 174 | 213] 387 60 | 40 16 197 | 228 | 425 23 | 55 9* 91} 119] 210 16| 16 22 164 | 176 | 340 40 | 45 12 120; 170; 290 80 | 40 32 235 | 215 | 450 39 | 33 19 121 | 187 |. 258 73 | 23 36 50 61} 111 25 | 61 = | | 10 244| 246] 490 D3 | 73 Total | 690 | 895 | 1585 | 287 | 167 11 259 | 242) 501 45 | 62 Av. | | 30 | 166] 208) 374] 58| 44 xu | 317 | 57) 38 *Died at end of sixteen days because of failure to feed on transfer ee 409 39 | 50 to new bottle. Total | 1789 | 1895 | 3684 | 351 | 453 FERTILITY AND STERILITY IN DROSOPHILA 185 given in table 12 came from the hybrids recorded in table 9. The pairs given in table 13a, b came from the hybrids recorded in table 11. Tables 12 and 13 appear to show that low production reappears after skipping a generation and that this low production is trans- mitted through both the egg and sperm to the grandchildren. I shall not attempt to analyze the data further as I have other and more exact evidence that bears on the transmission of the low fertility of the truncates. CROSSES BETWEEN THE LOW-PRODUCING TRUNCATES AND THE _ HIGH-PRODUCING WILD STOCKS IN WHICH AN EXACT MEASURE OF FERTILITY IS EMPLOYED The term ‘fertility’ is used in so many different ways by differ- ent writers that I wish to make clear at the outset the sense in which the term is used throughout the ensuing papers. By ‘fertility,’ I mean the percentage of eggs that complete develop- ment and give rise to mature flies. For example, if 100 eggs are isolated from a stock and later 50 flies come from these eggs then I speak of the fertility of that stock as 50 per cent. It is evident that fertility in this sense can only be determined by isolating the eggs, and that the number of offspring produced by a pair does not give a measure of the fertility of that pair but only a measure of the ‘productivity’ of that pair. The marked difference in the length of life between the truncate stock and the wild strains together with the increased length of life of the hybrid makes it very difficult to draw safe conclusions in regard to the role played by the egg and sperm in fertilization in so far as establishing any definite ratios are concerned. In other words the production of offspring cannot be taken as an absolute measure of the fertilizing power of the egg and sperm for as we shall see later in the case of the hybrids, although their production of offspring is very high, their fertility (combination of egg and sperm) is relatively low. Accordingly, I wish to return to the question of the low fertility of the truncates and consider the behavior of the low fertility in this race on crossing into other races. In order to insure an TABLE 12 Showing the result of mating in pairs the F2 generation of the crosses between the short-lived, low-producing truncates and the relatively long-lived high-producing inbred stock TABLE 12a TABLE 12b Grandchildren of the cross between Reciprocal cross: Grandchildren of the truneate @ and inbred & the cross between the truncate «7 cn a 7s "i and the inbred 9 bE | s ee tou lite AL ae ae | | Soe es 2 aa S & B¢ fe) | ga | SRE TTS) ma eal 2 ° i) 7S z eS am = 6 oy | ee 56120 se ye lee: Mime eller Hon | OB lo eR ies i & Z Z S| | g ee 5 eC) | | =: So ee a ies Rees ol ee eee ee eg BI oes Rl ee ain Bae yey Wanye ie as 20 | 111] 119] 74) 144) 144 S39) Aral A Aga ay 36 M1 120° —| 34) 157) 331 38+) 38+ 1i1| 140} 256] 38) 46+ 34| 161} 319| 24) 29 ttt 1490=" Ont = igeee S23 ulolGra | 59a) | Tan ees 111] 153] 150] 41] 44 33| 172| 106; 47| ? 111} 163| 152} 5 | 38 33.) 5173) 380 | 325/946 |} 111]; 465), 0), =e ii 91; .118; O| —| — 111} 178 | 215 | 35+) 35+ 91/ 122; O; —j| — 119) AS0-) 13541) aie ais 91; 125} oO; —j| — 29 | 1024 111 | 89) 40) as |} 91) 155} Of; —| — 95| 117| 153 | 244+) 24+ 12) OTs) AN LAA 24) 2 102| 124; 192} 14| 24 e107) MG eae. age) tA | 100| 127] 136| 18+] 34 | 106 y|M2E | 504 P| 14 | 102") 1128)" "98s wala | 106} 123 | 251]; 31] 31 |: 95"| 28-200: VS4nn ee eaoOpiaMzo th 2b, Lo Mas 95 | (182) |.' 200°) 26 uae 10641) 130%) 295). 9 323 wey |; 102) 433: |) 1075) » 20rleas | 50] 134] 280; 20] 39 100} 136; 6] 388] 6 PemOr 135 [1 G2n)) sO 23) Ve ) O29) 4a") 295) Soy eos 50| 138 | 164; 30) 38 | 102; ) 61429) 265) eA ae 50 | 146 | 208 | 95-4 25 102 | 144% 42) 305) a 106 | 150 224, 23 30 | 95| 152) 46 6 39 1068) 54"): 75)" 20i) 15 Si) A565 © 199938 Arcee | 106] 175 | 178| 29] 3 102) 160 | 298) 38] —? 107°) 182 0; —) — 100 | 164| 432] ° 32| 32 \-107| 184| 44| 21 bate 102 | 166) 379| 24) 35 TAP a7 4) 112.1964) 26." 4s 1G2;)) 0681" 61) 2oaieeae NeoRLe CSAS PAOLA. ee Pee 102| 169| 189] 29) 24 ie 28d) LON esGVel ano) 1b? 102.) 170'), QSS ite ere 96'|. 143° | “Ail 36 46 | 95.| 4171 | 2264 38) Oe 2%) 145; O}| —| — 102| 174] 111| 24] 26 1A 15877) OU eh 95| 176| 205; 54| 26 17g) ASO eI eb ly? |) 102;) 177 |) 43%) 19) eee 7M 2 Salary) 112 102'|\ 179 | 230) 238i e248 26} 180 | 204} 59| 46 102 | 185) 4257) 2m 26| 184) 44| 21} 12 SUN UMM GaN SPC tf 16. 8090126) DOU sa ase tn neta: 5 eal are | 80) 151) 288° 37| 32 = AV. | jE SO aaeeee A ae | 183) |) Soleo eee >) 0 alee ae Av. | | | 140| 30— 30— FERTILITY AND STERILITY IN DROSOPHILA 187 TABLE l8a Table showing result testing in pairs the grandchildren of the crosses between the short- lived low-producing truncates and the relatively long-lived, high-producing , . tnbred stock. T = truncate; C = erescent Grandchildren of the cross between the truncate 2 and inbred FAMILY NO. | J | | Op GRAND- | LP NO. oat | TOTAL pate | Tae 5 OF | | 33 7 415 20 Ft 70) tea 16 416 Seer || LESS 5 1s RAB sie nates 417 C POT | ghsGn el he 2018 "les Rees 19 | 418 0 Oa O17 Spuera2 aa 6 Se ose sare = SRO Mh Me Te 27 27 | 387 119 182 301 | 49 33 | 388 ORI A ae at Ta G2 PTs 15 low easouiee Wee HOT. O/Ote NN! 110s) Nia tG 30 390 Oe a iOieniy uly OF» Ii AGRE mamma | 91C¢ 67 72 130° ° |) | 27 eats | 3920 ¢ (6) | ete 186.) 4895 Giezs | 393 Cee esi ON eas ayer yh Oe il 394 92 102 194 60 a 395 77 121 198 61 25 436 'T SOND | 15 437 0 0 One 8 54 15 | 438 T USA), ) 20ee | mee Si eae TL see) gasp Tit Neo) 156 52 fai) 16 | ASSO mL) || AI45 Oe AUST S| 8076s) eee |, 2.29 peso. (ea97 103 143 hy 246" pea 30 398 71 81 152 53 33 | 399 | 68 72 140 ~—39 20 36 | 407 L ~ ‘60 74 ist) | 63) 2e 408 98 102-0) 200 16. 1* Pe | 414 20) aly, 2k eel 16 21 35 F- 10) > |-400 151. {> -1847" |) 985 34 56 401 S35 |e S360 ae eee? 45 31 402 T A 88 | 12 | ose 9 37 403 T 78 73 igit eee. a sor 11 497 ie nail 63 114 16 20 428 74 100 174 | 22 32 $2, )') 429 71 67 138) 5 ? 42 | 430 62 47 109 | 50 18 30 | 410 2 2 4 37 56 434 14 14 28 16 | 6 435 0 Om) 0 50 | 6 Rotalias.. = = 1959 | 2229 | 4188 | 1279 | 881 Average. | — — — — | pled 36 25 188 ROSCOE R. HYDE exact measure of the fertilizing power of the egg and sperm the experiments were carried out in the following manner. A female was placed in a bottle with four or five males to insure an abundance of sperm. A small piece of banana was introduced on a piece of paper and on this food the female laid her eggs. Each day the food was removed and the eggs picked off individually with the point of a needle. Each batch of eggs was planted on fresh food and the number of eggs recorded. After a few hours new food was given to the parents. This is a most laborious proc- ess so that one cannot carry out as many crosses at a time as might be desired, but this does not invalidate the test as such. Since the Woods Hole stock was the most fertile one at hand, six females were tested with their brothers as a control. Atten- tion is called to the fact that these females were about ten days old when the experiment began. Eleven truncate females were tested against the Woods Hole male. Fourteen truncate females were tested against their truncate brothers. These counts have led to the results given in detail in table 14, a, b and ec. TABLE 13 b Reciprocal cross: Grandchildren of the cross between the truncate @ and the inbred Q@. ‘oF GRAND- | see ORs, oe ee Gakp) | Vor) oF |r 25 | 1 420 T 74 7fil 145 32 45 421 40 59 99 19 29 422 0 1 1 20 20 423 0 0 0 52 26 6 | 424 99 104 203 20 42 BRR sae aa 39 90 52 16 426 Pipe 92 184. |< 22 45 12 411 a 0 7 55 11 412 64 68 132 34 29 413 17 20 37 ~ 200 22 | 419 0 0 OF jt 52 | 404 22 28 60 15 15 405 | 82 90 172 62 e2 | 431 21 36 57 29 52 } = Tose ue = 569 518 | 1087 472 436 Average | me —_ — — 91 34 31 FERTILITY AND STERILITY IN DROSOPHILA 189 TABLE 14 Egg count showing fertility of the truncates stock bred to the Woods Hole stock TABLE l4a TABLE 14b Woods Hole @X Woods Hole & Truncate 9 X truncate a8 a a | 8 anit a Bees 2) ic e | & 8 Bt lee No:*|| 2 ae S - a No. Or 3 SMG) es B m P = a 6 5 § | = a | a 2 A HQ a rat ; FAA 4 : a9 a) : A p | £43 a A ae = Seiiteca ieee g Br 2 ele eo lee g 1 OT 2207 0 | 227 | 148 18 30.) 271)! 120.) 158) 42 y: 28 | 302 0 | 302 | 256 197 +|y (23h. 268) e770 |) es ee Bre eo | 2 40 0| 40| 30 20 2B) D0) |. 14:1 7G 4 25 | 424 0 | 424 | 320 21 5 2 2 Oy 0 ree on) 20.) "0" | 20"! 10 990) es6 e230 0| 239! 56 Gu 243807) | 389: \1820 23 57 | 560 | e111 | 449| 76 | 3 r apa 24 |; | 42|), 435°). 32!) A0geI) 80 Total (a 1402 0 | 1402 |1084 25 39 | 304 93 981 | 87 ee re s ey 26* | 38] 360 0| 360] 86 TABLE 14 ¢ 27 20 | 295 0| 295) 69° Truneate 2 * Woods Hole 28 34 | 282 | 0 + 282 | 114 - = ; ——z 20 ee kO,) | eS 1s e eOsee 0) | 5 8 5 & 3000) 35 0 0 0; O Wise aie e | a 31¢| 42] 394 0| 394 | 150 Pee esics | come is A a = ; | 4 | = Q 2D Spa Se Za : Totall — | 3249 | 626 | 2623 | 644 Ul Blea! -O 171|, 38 8 Saf 980s 0 93] 58. 9 One Sin EA Teli 10 107" = 20)|) =20 Oy O 1 Bae Dalene Tike 0) 12 28| 226 0| 226] 148 13 14| 179 0| 179] 68 14 Glare 0 1 1 15 gi OT Gu sale =8 16* 6 6 6 CAO 17 Taga 3 15 | 10 Total) —| 614| 55] 559 | 307 * Escaped. + This was a long winged female and is not considered in the totals. Note that her fertility was somewhat higher than the average. This will be discussed later. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 17, No. 2 190 ROSCOE R. HYDE In this experiment the entire output of eggs of the trvuncates was collected. The count from the Woods Hole stock does not represent the entire output since these flies were about tendays old when the experiment began. It is true that it is practically impossible to get the entire output of eggs by relying upon an . ordinary dissecting microscope without spending an undue amount of time, but as the same method was applied throughout this source of error as it enters into a comparison of the output of eggs of the different races is about constant. I feel quite sure that the counts, as here shown, represent over 80 per cent of all the eggs laid. That many were missed is certain as the food from which the eggs were picked was often kept and later larvae were ‘seen toemerge. In regard to the conditions of the experiment as a test of viability, no objection on the grounds of personal selection can be made, as it is impossible to tell beforehand which eggs are going to hatch. Care must be taken not to allow the eggs to stand too long before picking them as the larva will emerge from the fertile ones while the sterile eggs are left behind to be picked up and considered in the counts. The question may arise in the mind of the reader as to why 626 eggs are not considered in table 14 b in determining the percentage of eggs which hatched. The explanation is given here since the question recurs in some of the other tables and in succeeding experiments. It is evident from the method used in making the test that a picture is given of the productivity of any single female day by day. The 120 recorded in column “No. not considered ”’— No. 18 (table 14 b) represents the total number of eggs laid by the female through a number of consecutive days in which none of her eggs hatched. The eggs that failed to hatch were in practically all cases those isolated during the first few days of the experiment, although in a few cases they came in the middle of the count, more frequently however toward the close of the count. This seems a fair way to treat the facts as this may be due to other causes operating other than eggs meeting sperm. In this case the benefit of the doubt is given to the truncates for if these are considered 749 in 3643, or 21 per cent, hatched. In reality the percentage is not changed much and had this been FERTILITY AND STERILITY IN DROSOPHILA 191 known in the beginning much labor could have been saved by picking the eggs from mass cultures, although this would give little insight into the total output of eggs from a single female. The former plan allows us to get much closer to the problem in hand. ‘The results of Experiment IV are given in diagram C. Truncate x Truncate. Wood's H. x Wood's H 24 2 Q11 b ry i. 2623 i 2 1402 h. 644 h. 1084 55 2.559 kh. 307 t=No. eggs isolated. e he Tio eggs hatched. Diagram C_ Effect on fertility of crossing truncate and Woods Hole stocks; from table 15. This experiment served in a preliminary way to show that the egg-picking process could be relied upon in general as it gives results in harmony with the earlier breeding experiments. The evidence seems to show conclusively that only one egg in four or five of the truncate female, when fertilized by its own male, will complete development. Moreover, it is shown that the Woods Hole male can fertilize more than twice as many of the truncate eggs than can her own male although it has been shown in earlier experiments that he produces an abundance of fertile sperm. It is interesting to note that only 77 per cent of the eggs of the Woods Hole stock hatched. It was the expectation that this stock would show a fertility of something like 100 per cent. At first I was inclined to believe that this discrepancy was due to some defect in the egg picking process itself, but the results of later experiments together with the behavior of this stock does not bear out any such conclusion. Since these experiments were completed I have obtained a fertility of 96 per cent with two new wild stocks. It seemed to me that the explanation for the low production of the truncates was not far to seek. That the length of life is a factor is not to be doubted, but more than that it is evident from 192 ROSCOE R. HYDE the experiments that only relatively few eggs were giving riseto mature flies. I wished to test the conclusion further and also determine whether or not some factor for egg production was operating, as well as to make the cross between the truncate male and the Woods Hole female which combination was not made in the previous experiment. Experiment V. This experiment was carried out in the same manner as the former one. The flies were mated as soon as they emerged and the output of eggs recorded from each female each day during her life. The facts in detail are recorded in table 15. The results of this experiment are expressed in diagram D. Truncate x Truncate Wood's H. x Wood's H. 2l e S ie e 15 . a 1901 g [622 nh. 396 A 1227 7 @ 1015 i 4065 JU al 80 » 3438 i= Noeggs isolated. h- No eqgs hatched Diagram D Effect on fertility of crossing truncate and Woods Hole stocks; from table 16. This evidence taken with all the other evidence presented shows conclusively that there is incompatibility between the egg and sperm of the truncates since only one egg in every four or five gives rise to adult flies. It is surprising to find that the truncate male should have actually fertilized more eggs of the female of the Woods Hole stock than a male from its own stock could ferti- lize, and yet a male from the Woods Hole stock can fertilize almost as many eggs of the truncate female as he can fertilize with a female of his own race. If the results given in diagram -D are compared with those obtained in Experiment I as expressed in diagram A; it will be found that the results are in general agree- ment. Experiment VI. Reference to table 2, where the history of the inbred stock is given, will show that the fertility has steadily declined on inbreeding, despite the fact that sterile females had increased to 50 per cent and again eliminated. It occurred to me that something like incompatibility between egg and sperm might be operating in this stock FERTILITY AND STERILITY IN DROSOPHILA 193 > TABLE 15 Egg count showing the fertility of the truncate stock together with the fertility on crossing into the Woods Hole stock TABLE l5a TABLE 15 b Woods Hole 9 X Woods Hole & Truncate @ < Truncate enita tre v8 ee ee ; oo | oo (2) iS) oS I fa) | ie] rou. ty | o 2 a a a nace 2 = a ae rg Wea le |e at ere | a 5 el a a 2 es a a Beles at lecie pame q i! Es J a q 2 SE Pyeeoale See 4 Need — (i 24 | 52) 463) 0 | 463| 356 1 | 32] 359/ 19] 340| 90 25* | 511 650] 194] 456 / 138 2 12 2 0 oO 26 33 | 316 0| 316 | 148 3 23 | 251] 220°} 381) 7 27 52} 529| 0| 529 | 433 4 A.| AT 14) ese re ap 23 25 | 314 0 | 314 | 290 5 4} 21 20) 70) ae a sto Tea al! ; 6 35. | 374 0| 374] 64 Total| — | 2272] 194 | 2078 /1365 7 7 99) |.” bas Wie ania a 8 Gri eee7 4 O. | 74a ot * Laid very small eggs; some looked 9 39 | 567 97 | 540 | 148 as if they might have been parasitized. 10 49 | 989 96 | 186 | 25 11 32 | 320 0| 320] 30 TABLE 15 12 Gul 22 22 0| oO Truncate 2 X Woods Hole & 13 19 45 42 Shi wel as an Sie ai pte = ies Ti ——— Ea ;:| ra | 8 Wad aa g Total| — | 2387 | 486"! 1901 | 396 A z Q & aA BIS ee e a E No. Bs Z - : | ‘ TABLE 15d s Fils “a Sul Woods Hole 9 X Truncate @ | 54 on ) 9° : : — S| mses a one 7 Z { 2 | in | a = oO } Dp Q oe 14) 88 | 18 15) 4 We aan | e 15*| 18 O)|) 40 0; 0 aa oagtaltie i e a 16 a 0) Ou AO 30 SMa ncarin|| > el PRO ea ig 19| 265| 0O| 265 | 214 sil Gaol eee ee : Sia) 5 6). 84 0| 84/ 64 Rita [oe EA ibe x Ho marlec, 4551) SOGt Wer. 63 B90 |) 72 We 7300: 0) |) 730.er7 20 ZOE 250u\) 6h 253) | 182 30) 92 | 393 0 | 323 | 276 21 S00b 235) | 53a). 182) | 162 31*| 31/1 631 0 | 631 527 22 16; 105/;. 0) 105) 42 32 78 | 536| 17 | 519 | 395 i ee ee oe 33 | 79| 552| 18 | 534 | 389 i CRG ay ali i Be 34 | 75| 798] 12] 786 | 588 RED aA cea lean ee a ar ce eo * Escaped. ey [= ee eo F: r — | 4127| 62 | 4065 [3238 { Copulated. soul) | * Escaped. 194 ROSCOE R. HYDE * Truncate x Truncate Inbred fx x Inbred Fw 24. , g ? 32 ie Dene A. 28 Lee 5 gilo t 2625 A i. el ‘ 52 fF isel 58. A /2k t=lo.eggs tsolated h= To. eggs hatched Diagram E Showing the effect on fertility of crossing the inbred and truncate stocks. and if so this could be put to a test and at the same time more light might be thrown upon the behavior of the truncate stock. Accordingly the inbred stock which was now in the F\, generation was crossed into the truncate stock. Table 16 gives the details of the experiment. Table 17 is the control on the inbred stock that was bred in pairs at the same time. The results of Experiment VI are expressed in diagram E. BACK CROSSES BETWEEN THE HYBRIDS AND THE RECESSIVE LOW-PRODUCING TRUNCATES We may next consider the evidence that bears on the fertility of the hybrids when paired with each other and the fertility of the hybrids when back crossed into the recessive low-producing truncates. This experiment was originally undertaken in order to determine whether or not there were sex-linked factors for fertility. The results in this direction have been largely negative but nevertheless the experiments give us the key by which the low productivity of the truncates may be explained on a very simple assumption. It was my expectation to find that the eggs of the hybrids when isolated would show a fertility of 100 per cent, since I had already found that their production is far in excess of their high-producing parent. Three experiments have been required in order to con- vince myself that the fertility of the hybrids is in reality low and that the percentage of fertility in the back crosses is an actual measure of the fertilizing power of the egg and sperm. There are two classes of hybrids, as follows: Offspring from the truncate ° by Woods Hole ¢& will be referred to as A 2 and Ac. TABLE 16 Egg count showing the fertility of the inbred stock together with the effect on fertility of crossing into the truncate stock TABLE l6a TABLE 16b Inbred 2 X inb ed & Truncate ? X Inbred of }3 [a . | se lie 8 NO. OF 3 + a a No. | o 8 ara | A ke ° z a be Salen ie oe 8 ose ee | [2 laa | Ba) Sea SMe ees ve lee eae lee 1 28| 340| 0| 340] 89 Gui nas] emo 0 0| 0 2 21 175 21 54 | 60 LOD ole 18 14 4 1 Seat elo: |, 325 0| 325| 68 11) "207 52215) 38"), Tse eee 4 | 386] 400 30 | 370 | 116 Le aie 18 94 | 45 5 28h) 408 0| 403 | 174 ibs pie 4 | 0 0 0 0 6+ | 35] 453] 20] 433] 140 14 6 0 0 ol emt y LO Ih 137 3} 134) 67 tee OMmELOL 3/ 98] 38 Sone ezee| Leo!) ar 43 4 16 10 130 15) |) PETS ess Te —— = 4 15 | 122 34 88 | 46 Total | — | 2413 | 211 | 2202 | 718 18 at 96 9 87 | 39 : Oe 3 0 0 0 0 * Lived much longer but ceased to 20 | 5 45 5 40 | 32 lay eggs after 15 days; preserved for 1% 17 0 0 0 0 PR iy 22 | 27 268) 0. (2281 164 7 Discontinued. 23 9 26 12 14 4 24 20 | 126 14; 112} 29 33 17 | 128 Oi 2s aes 34 13 | 139 5. | 184.) ae TABLE 16c 37 11; 106 0; 106; 43 Inbred @ x truncate o 38 19 | 253 0} £53 | 159 a a t anh 39 5 0 0 0 0 a g Qa A & g a a g 40 10 0 0 0; O Ben ee e a 5 41**| 17 | 256 0| 256 | 137 Ie: 2 4 5 z Z 2h tall Pelivil| da (0) |) AOE |) eR? a 2 a e g 8 43 15 | 126 11 SS eel | oF ABET NER a LOA 19| 165) 86 25" | 28) 291 55 | 2386 | 111 46+ 9 | 0 0 0 0 2675) 289| 166 120) 15455 +29 AT 17 72 10 62 | 39 Zia e267) 3030!) 630), 273 |v 154 48tt) 12 0 0 O:\,."6 28 2 \ 248 0| 278 | 193 49 itn. Sz (ale Keogh ay! Borel B20 LOF 47 60 | 18 = = =|= 30* 22| 375 0| 375 | 263 Total | — | 2877 | 252 | 2625 1361 Sl) BR Ake 0| 407 | 225 Wmeeapad ; eer 5 ; ¥ of Gs be | gee re Bets a8 ** Discontinued. Total | a 2954 | 144 | 2110 |1222 if Killed and preserved for study. £ sts 7+ Laid a-.few cheesy looking eggs * Discontinued. probably due to parasites. 195 196 ROSCOE R. HYDE TABLE 16d Truncate 9 X truncate o | cs mes NTRS | ceca? aol ve ae S a s io) Qa ic NO. Or Zz = a mn 5 : g a S = ei veal. a7 aoa es Sf Ena oa g g 35 2961 179| 70} 109| 26 36 8| 9 0 | Ou) 22 el ee ——_——| — eet — Total] —| 188 70] 118] 28 Offspring from the reciprocal cross will be referred to as B ? and Bo. In these experiments A @ is tested with Ac’, B? with Bo, AQ with the truncate ~, BQ with the truncate <’, the truncate 2 with Aco and the truncate 2 with Bo. In order to get an exact measure of the fertility, eggs were isolated day by day from each of the combinations made. The number of eggs isolated each day, and the number of flies that emerged is given in table 18. TABLE 17 Fis generation of the inbred stock, showing the result of testing the flies in pairs; compare with table 16a NO. | NO. OF @ NO. OF Q TOTAL LIFE OF © LIFE OF 2 1 | 59 67 126 43 26 2 | 5d 39 94 31 31 3 113 133 246 45 37 4 33 dl 64 41 10 5 0 0 0 91 13 6 89 | 90 179 77 21 di 129 | 139 268 37 37 8 70 90 160 $1 17 9 129 13% 262 41 35 10 | 28 31 59 30 24 11 104 111 215 30 28 12 46 46 92 28 24 13 128 96 224 39 28 14 71 59 130 37 25 15 44 | 77 121 28 28 ‘Rotel eee ace 1098 | 1142 2240 679 | 384 Average...... — | = 159 45 26 197 FERTILITY AND STERILITY IN DROSOPHILA . +88 oie +82 +66 06 Ope | oe eee "** 98uqUed10g | { | | | | €¢ | €SOL TF8s LG | 9COT | 686 ‘196 169 | LEGS OGE SED | CELE |\Z8Z G96 | EazE TEE | 198E [8IOL Sloe lem Elo |e le fee it- |ze Ller | eer —lor- | se ¥Z i eee iy coe be eh FOR | Zp Pr PLE 4 Sr lee 8% epee Gh, 6) PSN OSs OG. = 7k OL: =.) eG: airegT "rez sect 0z aan tte LOL ia) 66 ipa I Scopes lel G nal ano ae EGO, Lig |—|19 | get | 6T ee hee TOP ee. 1) OR hen | 48 Cece | .6Tn lea Sho ST — | #8 OLE (09 ZI joe |86 |—| 08 LOLS ineea OS, +3) ln bt OL LI eee eae OL 8 PEL FOE. 06) = /S- 106 Aaa G0 en anee 1c8 or sel eye VAIL Sie StS OOF me Zee 8S 0 al GT Owed "Ge Hara es er OcI CT aaa OS GCL =| 4S US a= |) Teal OS PS) ie OV SO a Ge Tél Fl Sal seed OST nh Ges ieee ala eee VGr a dec alien sO et geCil OS. ScSar | ral eee e0G a4 come voles eG, SGSh |" 2eameTes|= 0G: mel KORE =" LOe IkOge. | or &¢G | FS L4bT (0 | && 801 |— | \78 --|92 |29T 0 | 29 yAIe |) (0) {| 1) SFI Il 0 | ST Tel (0 | Sr | 92 OF | 6T | 99. \c9! | ST 8S OF | Fz Sil SeOn Gs SST OL O65 SS OM SS. a (Cai 2 VG a) WAR) | lt OG NO) ce | SPT lo | SP 6ST 6 0 |9 | IST 0 |9F | I PI) 02 |66 \ee/8e [Gort Zr \o0e |ozt jo) Fe | ert | 8 OR er 06a Mi Ick Pha 9 | O0T249)|0c Stel Fe tr | et | 0 er | 98t L 0 | 02 89T 0 | 22 Ask HON Hees MW SAL Ae Neh 666 68 | OF Vi) Os a88 I¥G 9 O | 4h |ect z|2e | 26 iS|te | ze jo | oF | Tee lor | ob | COT Outs. 206 | g 0 N&SP OOT 0 | 6F 16 JLT 6h | GCL |PL | br | GET | 6Z C6 0 | 8¢ Vind | v 0 | v€ 86 0 | 24S | GOT {22 | Sg SOI I VL | O12 0 | 49 | SS 0) OLT L1G € 0 | SS Iél (0 | 6F OSE SLES MV Cea Gh alee oN2OG. ale OF. Tél 0 | 66 $9G G =eP | s SE Feel eee Soe m2 Tel aS | 006 |Z. | v2 | 20c 16. | 79 StI | 0/| 28 9LT ZI, [T toquiooeq lsqynpy | ssoq | synpy | sso | synpy | sss | SyNPV | S555, | | synpy | Soo synpy sso] 19-98 | e¢-0g | ED aes FE-61 | 91-01 | el hme, Mek, Wrest ance ua vAN OVXb6V | fax ed X ob OVX bl OLX a PLXSv | NOIGVNIGWOO 5 Y90}8 aypaUuNns) Buvanposd-N0}] aaissadad AY) PUD (YIO]S ayDIUNA} buronpoad-no) ay) pup Y90]8 a]0 77 spoo 44 Buronpoud-ybry ay) woLf bursds fo) praghy ay} waanjag sassouo yovq ay} wows ‘shop ha payojny joy) sauf fo waqunu ay) puv paqn 7081 sbBa fo saqunu ay) saarb T AIAVL 198 ROSCOE R. HYDE The percentage of eggs that hatched in this experiment is not altogether reliable, due to unfavorable conditions which arose. Previous to this time the flies had hatched normally and I still believe that the foregoing experiments measure the absolute fertility of the different stocks and combinations made, within a range of 3 per cent. In this experiment, however, the food be- came very sour and the flies did little more than crawl out of the pupae and die. The dead animals were searched for and counted. The percentages as given in this table are in consequence too low and are to be taken relatively to each other and should not be compared with the percentages as given in the other tables. Tables 19 and 20 give the details of a second and third repetition of this experiment. The results express more accurately the facts in regard to fertility in these combinations. The plan of the third experiment, table 20, was modified somewhat for it seemed that fertility could be determined fairly accurately by isolating the eggs from mass cultures provided the experiment were controlled by a stock the fertility of which was known. Accordingly hybrids between the truncates whose fertility was low and the Woods Hole stock whose fertility was high were produced and a number of these hybrids were tested in mass culture. The hybrids were also back crossed into the truncates as shown in table 20. In each case a number (the actual number is indicated in the table) of virgin females was placed in a bottle and twice as many males from the stock, in which the combination was desired, added. The males and females were kept separate for four or five days before they were placed together. At the time of placing the two sexes together an epidemic of mating took place. In each of the bottles about half the pairs were to be seen mating at the same time. None of the eggs were isolated until five days later in order to give all the females a chance to mate and in order to avoid the first few eggs laid, which as I have shown, are not likely to hatch, especially in the case of the truncates. The eggs were isolated day by day in the usual manner. ‘Table 20 gives the result of this experiment. It will be noticed that it is in close agreement with the second experi- ment of this series. TABLE 19 Showing the combinations made between the hybrids and the truncates; the number of eggs isolated and the corresponding number of eggs that hatched Tox Ba Agooc Ie ge | g8 | fa | 8 ge | $8 | fe | 2 NO. a is BS aie NO. a ane a ays 4 .9 A 5 2 ; ines | iS) Lilia eae! Wine Mee ay 13 pg 0 0 0 39 18 | 284] jo99!| 9Q 14 Nove ag ga oa A 40 13,1) FO1S: gz aco 15 13 | 200 | 69 0 41 20 | 585! 368| Oo 16 2 0 0 0 42 17) ALE | a7 ig iby ec teen. || a! 0 43 | 20: 853: | coer ero 18 past al ioe 8 0 44 | 20 | 296! 125 | 136 19 ia 8G. ls 30 0 45 20 || 288 | 403) 20 Ao) 4G ale ike aaa |_| | SP RG es aedent ae 23 £45 4) 311 10 25 aa 24 13 | 144 | 38 0 25 RP vale od) Bi Or ee wena 26 Fe TOON 4) 2SG0ul veil eae |) Sickel om heen 27 Ha Ole OCs \y “9 eon Cee se | 2a | *E | | 3 E | a3 28 20 | 0 0 0 | $8 | s8 Si | 8 Total 997 | 324 | 78 ie 2 | 513| 3| 0 Be on cent 47 11 | 296| 1581, 0 ne ae 48 20 | 400| 209] 11 hea ex Be 49 20, 480 | 798 18 z : a8 : ; : a 50 AG eAgT N28 on iG NO. Nip ete als EGS ss B vi [rept roe . las 3 5 3 sa oe Total 2171 | 1099 24 i i mie re z, ea 2 ees aoe oA =e 6 9 98 99 0 50.6 per cent ‘ 16) | ASe 220 \o we D 8 Banleas sega G _Texag¢ - Oe sis | 257 65, | 0 | fa | $8 | ge | 88 HON Ge 2a 12d. |) 0 e ab | #2 | ge | 25 11 Be egal owe a Som eer is sa ee 12 20 | 437 | 277 0 a Fe ie | SR | aase 29 20) |) 27en 150.) 6 59.6 per cent at 7 TO epoca le nO | 32 13 16) 95 25 AXA? 33 3 BB eoo8 |. <0) @a | 2a | 2a | 28 34 17 tie aL 6 wo oa Be of no. Ae | a2 | Be | Ze 35 Siieeo | | Olly Se | ce | ear) Sz 36 10 SON TOs ee Ng nce NUN ea 37 20) SEAT (orga ig 1 20 | 460 | 278 Ole 38 20R ST 767 One 9 20 | 392 | 164 0 Ose eT THE Gaara : Nae Wea r Total 1049 | 458 | 25 4 3 57 30 | 0 43.4 per cent 5 20 | 382 | 257 0 Total (1435 | 808 0 56.3 per cent 199 200 ROSCOE R. HYDE Diagram F expresses the essential relation of fertility as brought out in these three experiments. I shall consider the results of the last two experiments in making up the average since it is evident that these numbers express more nearly the actual facts of the experiment. TABLE 20 Back crosses between hybrids (of Woods Hole and truncate stock) and the truncate stock. The table gives the number of females that entered into each cross. The numbers on the left side of each column show the number of eggs isolated, the corre- sponding number on the right the number of flies that emerged. cRoss | AQXTS| BEXTH| T?XAP| TE XBH| BE XB) AP X AP) WE X Wo — —___—— a _ = — 1 - — —|—— —— ~ — | ——___________ | | | | | NO. FEMALES | 12 15 | 28 28 15 | 15 | 15 | | }eges adults eggs adults eggs adults | eggs adults| eggs adults) eggs adults! eggs adults —— ee | | | Jan. 23 713 | 110-45 | 120- 65 | 150-39 | * 61— 7 | 217-130 | 180-111 | 225-130 24 60-20 | 114- 54; 70-26) 43-13 | 100- 0* | 103- 7*| 95- 45 25 | 161-89) 130- 63 | 74-7* 55-18 | 131- 72 | 160- 43*| 116- 72 26 135-41 | 142- 82 | 120-39 | 150-56 | 151- 87 | 115- 47 | 150- 97 27 | 103-47] 90- 29; 183-21 | 170-40 | 151-100 | 150- 48 | 140- 91 28 | 105-39 | 150- 55 | 200-86 160-74 | 165-104 | 166- 73 | 175-144 29 | 160-80 | 200-119 | 175-63 | 146-42 | 190-145 | 231-153 | 176-123 30 | 110-33 88-73) 125-58 | 107-46 | 112- 86 | 203-135 | 120- 81 31 | 45-25] 63-44) 61-49) 51-16| 70- 50 | 100- 65| 95- 67 Feb. 1 | 71-35} 106-55) 94-40| 45-11] 61- 45 | 125- 94 | 100- 75 2 | 90-53) 86-49/ 80-52} 30-10 | 110- 80 | 125- 89| 36- 29 3 | 40-28} 24-14] 33-18] 20-6] 59-41] 42-29| 25-17 4 | 38-20] 45-29] 53-27| 438-19) 35-25] 60-39| 44 31 5 | 20-8} 45-22) 50-18) 10-3] 35-25] 21-12| 30-21 6 -43-23 | 50-24] 34-18] 20-7] 35-31] 86-59 | 7 27-12, 78- 37| 45-15! 30-8 | 100- 62 | 130- 58| 98- 76 8 39-16 | 91- 44] 3412) 27-8 | 102- 55 | 120- 55| 50- 38 9 | 52-25 | 150- 75 | 105-40) 43-12 | 53- 34 | 100- 49 | 100- 74 10 | 160-88 | 85- 49} 100-34 | 28- 0*| 121- 63 | 150- 80 66- 50 ra | 140-73 | 114- 42) 102-36| 17-1 | 115- 70 | 115-60 | 135- 81 13 | 50-383] 51-22] 50-20) 0 55- 28 | 53- 38) 54 36 14 | 51-26] 50-18] 50-17| 0. | 55-30] 56-30) 55-43 Motals i. 1810-859,2072-1064| 1864-728, 1228-397 2122-1363 2328-1324 2085-1421 te Pe EY fe | | Per cent...) 47 Blea: (SOA mee 64 57 70 * These bottles met with an accident and are not considered in the totals. FERTILITY AND STERILITY IN DROSOPHILA 201 v R xm AS Bol) x BE # 56.8 24 : 59.6 Sey rx UTES. 3 28 0 jad 41.5 $4.1 5.4 43.4 543 35 50.6 Diagram F Summary of tables 18, 19 and 20, showing the effect on fertil- ity of back-crossing the hybrids into the recessive low-pruducing truncates. THX Ts e¢ we x we oS 80 ASK Ae Tx Te eM BER Be 6.6 61.9 8.5) 9797 50.9 97 st @ ©= Calculated Diagram G Expresses the essential relations of the behavior of fertility in heredity a3 based on the foregoing experiments. The numbers in the circle express the calculated percentage. The other numbers express the observed percentage of fertility. In diagram G a composite picture is given of all the experiments that bear on fertility in these crosses where an accurate method of measurement has been employed. If 24, 75, 55, and 80 express the actual fertilizing powers of the gametes of the different combinations then the percentages as given in the hybrid back crosses can be explained on the following assumption. 202 ROSCOE R. HYDE The hybrid A is made up of T? and W.. Let us assume that its germ cells segregate into the parental types in so far as factors for fertility are concerned. Accordingly A @ produces two kinds of eggs, T eggs and W eggs. Ac will produce two kinds of sperm, T sperm and W sperm. Diagram H expresses the relation as to what happens when the gametes are brought together. bgt = T Pe cal Sperm T E— 53E Diagram H Showing the possible combinations of the gametes of the hybrids (A) on the assumption that the germ cells have segregated into the parental types in respect to factors for fertility. When a T sperm fertilizes a T egg 24 per cent of the eggs hatch When a T sperm fertilizes a W egg 80 per cent of the eggs hatch When a W sperm fertilizes a T egg 55 per cent of the eggs hatch When a W sperm fertilizes a W egg 75 per cent of the eggs hatch The sum, 234, divided by 4 gives on this assumption the actual fertilizing power of the combination, or 58.5, which corresponds very closely with the 56.6 observed. In like manner the other combinations may be determined. I have placed in a circle in each case the calculated percentage of fertility; the number out- side the circle expresses the observed percentage. The agree- ment between the two sets of numbers lends support to the view that the germ cells of the hybrid do segregate into factors identical with those of the parents in so far as factors that are responsible for fertility are concerned. We find here, I believe, the key for the low fertility of the truncates. They have become homozygous for many factors. The corresponding rise in fertility when crossed into other races is to be explained on the assumption that the combination formed is heterozygous for more characters upon which fertility depends and consequently more likely to develop. The assumption seems a reasonable one to make in light of the fact that we are able to predict the fertilizing power of the gametes in a given combination. FERTILITY AND STERILITY IN DROSOPHILA 203 It will be recalled that over sixty generations of continous selection has failed to produce pure truncate stock. May not this result be explained on the grounds that only the heterozygous forms with respect to this structure ever reach maturity; the homozygous or pure truncate animals never come into existence? It is of interest in this connection to recall Cuénot’s failure to produce yellow mice that would breed true. The reduced fertility of the inbred stock in successive genera- tions of inbreeding may also find its explanation on the assump- tion that the gametes have reduced to a condition approaching homozygosity for as I was able to show in the Fy, generation a larger percentage of zygotes developed as a result of outbreeding. FERTILITY AND HIGH PRODUCTIVITY OF THE HYBRIDS It has been pointed out that the hybrids that resulted from the crosses between the low-producing truncates and the high-pro- ducing wild stocks give rise to a large number of offspring. They produce on an average many more young than their high-produc- ing parents. The fertilizing power of their gametes when inbred was only 56.6 per cent, while in the Woods Hole stock it was 75 per cent. In other words, the productivity of the hybrids is higher than its high-producing parent, but the fertilizing power of its gametes is lower. This apparent contradiction is accounted for by the fact that the output of eggs of the hybrid female is greatly increased and de- spite the fact that the fertilizing powers of the egg and sperm when placed together is relatively low, nevertheless the production of offspring is very high. This fact shows how easily one might be misled by using the number of offspring as a measure of the fer- tilizing power of the egg and sperm when not controlled by more accurate means. A comparison of the number of eggs laid by the hybrid, table 21, with the number laid by the Woods Hole females, tables 14a, 15a and 15d will show that the output of eggs from the hybrid is practically twice that of its high-produc- ing parent. The following controlled experiment, table 22, confirms the foregoing conclusion and also shows that it makes no difference 204 ROSCOE R. HYDE in regard to egg production whether the father or the mother of the hybrid came from the low-producing truncate stock. In this experiment conditions were as uniform as possible. The virgin females were in each case paired with three of their brothers. The,eggs were counted each day with the aid of a dissecting microscope and discarded. ‘The females in this experi- ment were given a chance to lay their eggs on dried apples that had been soaked in water and allowed to become slightly sour. This insures great accuracy in counting, as the white eggs are easily seen against a brown background. The count as given in tables 21 and 22 supports the conclusion that the hybrid female has a greatly increased ouput of eggs as compared with the Woods Hole female. Now despite the fact that only about 56 per cent of her eggs hatch when tested with her own male, yet she is able to produce as many if not more offspring than her high producing parent, the fertility of which is 75 per cent. We should expect the hybrid to lay approximately twice as many eggs. An analysis of table 22 shows this to be true. It is true that no far-reaching conclusion can be based on such a round-about treatment as this, taken alone. This evidence must be taken in connection with the other evidence presented. If the percentage evidence as given is correct, and if it is true that the hybrids are excellent producers, as judged by the number of off- spring produced, then we should expect the egg-production to be greatly increased. The count as given here verifies the expecta- tion and makes the former conclusions more certain. The anomalous condition that the fertility of this animal may be relatively low and yet its productivity high, finds its explanation in the facts presented and shows, as before stated, how one may easily be misled in regard to fertility in heredity by using the number of offspring produced as a guide, when not checked up by more accurate means. The increased egg-laying capacity on the part of the female may seem strange at first sight, especially when we recall the fact that neither parent possessed it to such a marked degree. But let it be recalled that the hybrid is more virile, as judged by the length of life, than either parent and that the same thing is probably true in regard to their reaction to light. In fact, in every way these ae TABLE 21 Number of eggs laid by the offspring of the crosses between the low-producing truncates and the high-producing Woods Hole stock. These are the same females given in table 18 (A) AQX Act (B)BeXBa NO. | DAYS COUNTED | NO. EGGS LAID NO. DAYS COUNTED | NO. EGGS LAID 1 | 34 1056 10 | 33 974 | 6 144 11* | 10 221 3 14 0 12 | 46 826 4 13 372 13 | 84 1807 nen] 7 | 170 14 | 48 1429 6* 4 116 Gia 15 351 7* 15 462 16* 28 782 8 88 2184 50 26 552 9 Fill 1613 51 33 756 56* 30 | 828 52 52 601 57* 16 371 53 44 652 58 52 1246 54 48 901 59* 14 304 55* 9 200 60* 59 | 1265 as Se we 61 | 48 1360 Total 476 10052 Total | 471 11491 Average per day 21.12 Average per day 24.4 * Escaped TABLE 22 A Showing the total output of eggs of the Woods Hole stock and their hybrid children when crossed to the truncate stock WH? X WH? A? X Ac Be xX B& NO. tare oF Xo. BaGs ae ae | mo OF Bie or NO. | ea Or Yo ee a 812 besa doa: Top ineas |e tenn 14} 51 443 2 | 14 403 8 38 1043, 15 61 9 3 34 886 9 | 55 1511 HGS Gil" 896 1 31 589 10 29 740 7s) Gl 369 5 52 1482 11 50 1204 13 16 346 6 | 52 232 12 | 20 490 1 es ae . ar nls errs. 6 | 57 .|. ogy. , Lotal... .| 234 5998 Total... .| 232 5823 21 | 54 600 foe he ae yt 22 | 61 1034 Average number of eggs Average number of eggs laid per day 25.63. laid per day 25.09. Total..| 517 | 6216 Average number of eggs laid per day 12.02. Average number of eggs laid per day, not considering number 15, 13.61. 205 THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 17, No. 2 206 ROSCOE R. HYDE animals seem to be better somatically than either parent. To be sure, their fertility inter se is much lower but it should be remem- bered that while the individuals belong to one generation the egg and sperm produced belong to the next. The following suggestion is offered as a tentative explanation: It is probable that some factor or factors, let us say, regulate egg production in the normal wild fly.!. Let us designate these as AB. Let us assume that the truncates have dropped out some factor from this mechanism, then their formula would be expressed by Ab. ‘The other stock used had been inbreeding in mass cultures in the laboratory and had probably lost some factor, let us say A. Then their formulae would be expressed by aB. When the two races are crossed, the two factors are brought together again, thus ABab, and the normal egg-laying function is restored. The explanation holds at least tentatively for other improvements to be seen in the hybrid. The number of factors concerned is not to be looked upon to be as simple as the formula would seem to indi- cate. Instead of two factors, as the formula shows, there may be many units concerned, but the principle is the same. The things lost out of the germ-plasm of one are made good by those present in the other and thus normal conditions are restored. If this is so we should expect to find members of this species in nature that live well on to one hundred days and some that have high egg produc- tion and high fertility combined. They should produce well on to two thousand flies. FERTILITY OF THE LONG-WINGED BROTHERS AND SISTERS OF THE TRUNCATES When the long-winged flies thrown by the truncates are bred together their productivity is greater than their truncate brothers and sisters. This fact is brought out in table 23 where 120 is recorded as the average per pair. Their truncate brothers and sisters used for control produced an average of 77 offspring per pair. The flies were of the same ancestry and subject to the same environmental conditions. That the long-winged forms produce 1 Whether the actual formation of germ cells, in this case eggs, or the regulatory mechanism that governs their actual maturation and discharge is involved is a problem more difficult of solution. FERTILITY AND STERILITY IN DROSOPHILA 207 more offspring than their truncate brothers and sisters is borne out by table 24, in which the F; generation of the longs as thrown by the truncates are tested in pairs. In this experiment they gave an average of 92 offspring per pair, while 34 pairs of truncates used for control gave an average of 47 offspring per pair. (Table 1, Fis.) I have pointed out that the number of offspring produced by a pair of flies is not an infallible guide to the fertilizing power of their gametes. That the fertilizing power of the egg and sperm of the long-winged forms is greater than that of the truncate flies is supported by the foHowing experiment, in which the eggs were isolated and 37 per cent hatched. Since I have never been able to obtain more than 24 per cent of fertile eggs from the truncates I conclude that this difference when taken in connection with the other evidence is significant. Attention is called to No. 31, table 14b, in which the fertility of a long-winged female is 38 per cent, or 14 per cent greater than her truncate sisters. The data recorded in table 25 are too small upon which to base any very definite conclusion in regard to this obscure phenomenon, TABLE 23 This table shows the result of testing in pairs the long-winged brothers and sisters of the truncates. For testing of truncate brothers and sisters to these, see table 1.—F 44. | | NO TRUNCATE | TRUNCATE TOTAL | LONG | LONG TOTAL GRAND barns rot | 2 | TRUNCATES | of | g | LONGS TOTAL | | 1 5 5 LOG seek ote so 80 90 Z 15 14 29 | 76 75 151 180 3 2 4 6 21 23 50 4 3 4 ii 67 | 75° 142 149 5 11 6 17 37 28 65 82 6* ~~ — — — = -- 7 12 | 17 29 53 59 1 141 8 7} 7 14 56 56 | 2 126 g* — — | = =| = — 10 15 10 2h ib OL Ree NS a AES 144 ue ———| wet opal... 70 67 137 421 | 404 | 825 962 a eee = ot a ee BE tS, | ——s = = Average... meatt a — — | aE ae 120 —_—__ —— — - = = _ ¥ * Sterile. TABLE 24 Showing the result of breeding in pairs the long-winged individuals selected from the previous experiment, the results of which are shown in table 23. ANCESTRY SERIAL! TRUN- | TRUN- TOTAL | LONG LONG TOTAL GRAND NO. NO. |caTEG| 9 TRUNCATE fof 2 LONGS TOTAL 2 Pe seal aay 20 | 66 49 115 135 2 2 0 0 0 | 33 33 66 66 2 3 0 0 0 | 6 7 13 13 2 10 ae i oe oe = = 2 5 2 5 7h 8 Fi 15 22 2 6 0 0 oO | 68 42 110 110 3 Thea Nie— |e | a) | as Ee 3 8 5 4 | 9 57 53 110 |. 19 3 9 3 2 5 By 35 72 77 4 10 7 2 9 42 56 98 | “te 4 1*| — | — — — — = — 4 12 6 8 14 59 | Thad e186 150 4 13 0 1 1 27 42 69 70 4 14 0 1 1 41 38 79 | 80 4 15 3 1 4 | 24 2, | 51 | 55 4 Lond Ss) eis 28 72 Oe ae ie 149 4 17 15s 10 25 66 53 119. |.) p14 4 18 Hi) al Sil 5 13: (ig a hes 8 tye We Se — | _ == -: a= 8 20 2 4 6 17 31 48 54 8 21 4) sitet 15 57 40 OF Nee ee 8 22 0 1%) 1 4 5 9 10 10 23 5 9 | 14 86 86 172 186 10 of || 94.) 45 | 69 50 | 52 102 171 10 O5* | | — — — = = 10 26 1 4 | 5 30 41 71 76 Motals oe | 104 | 134 | 238 858 828 | 1686 1924 Average... = — | area = — | — 92 * Sterile TABLE 25 Showing the number of eggs isolated and the corresponding number that hatched on pairing the long-winged males and females thrown by the truncates ; | Enos NO. EGGS ioe | ISOLATED | HATCHED iat 1 152 | 66 0 Bat AL 14 31 Seem Git! yee eco NY te : 73 | 22 20 5 5/7 ))|, Rasta 0 Total | 382 140 | 82 FERTILITY AND STERILITY IN DROSOPHILA 209 but that the productivity of the long-winged individuals is greater than their. truncate brothers and sisters I can testify from my experience with them in mass culture, and it would seem as though this increased productivity is due to the increased fertilizing power of the egg and sperm. I shall return to this question in Part IV. BEHAVIOR OF THE TRUNCATE WING IN HEREDITY Attention has already been called to the fact that many genera- tions of inbreeding the truncate brothers and sisters has failed to purify this stock. The truncate always throw long-winged males and females. . The long-winged flies in turn throw some truncates. Table 1 gives the number of truncates and longs that appeared in the different generations by continually selecting the truncates. From a total of 6356 there are 5469 flies with truncate wings and 887 flies with long wings; a ratio of 1 long to 6.2 truncate. Dur- ing the period under investigation this ratio remained fairly constant. When the long-winged flies thrown by the truncates were bred together there appeared in a total of 962 flies, 825 with long wings and 137 with truncate wings; a ratio of 6.2 longs to 1 truncate. When the F, generation of long-winged flies were bred together the different classes appear as shown in table 24. In a total of 1924 there are 1686 long-winged flies and 238 truncate flies. The males and females are about in equal proportions. This givesa ratio of 1 truncate to 7.1 long. This ratio is practically the oppo- site of the condition found in the truncate wing which threw 1 long to 6.2 truncates. When the truncates are paired with the long winged wild stocks such as the inbred or Woods Hole, long wings are dominant to short, but not completely so; for while no truncate wings appear a new class of wing arises. This is along wing with small crescent shaped piece cut out of the inner margin at the tip of each wing. These will be referred to as ‘crescents.’ In the cross as given in table 3 there appeared among a total of 1784 hybrids, 41 crescent males and 35 crescent females, a ratio of 1 crescent to 21.2 longs. In the reciprocal cross from a total of 4711 offspring there were 30 crescent males and 84 crescent 210 ROSCOE R. HYDE females; a ratio of 1 crescent to 40.3 longs. I shall neglect this ratio in further discussion as it appears to be aberrant., In the cross as given in table 6 a, there were 34 crescent males and 19 crescent females, a ratio of 53 crescent to 1127 long or 1 to 21.3. In the reciprocal cross, table 6b, there were 30 cres- cent males and 49 crescent females in a total of 1714. This gives a ratio of 79 crescent to 1635 long or 1 to 20.7. The different classes appeared in F, generation of the crosses as follows. From a total of 16096 as given in the cross in table 9 there are 516 truncate males, 778 truncate females, 431 crescent males, 633 crescent females, 6881 long males and 6857 long females. In the reciprocal cross there are 273 truncate males, 332 truncate females, 315 crescent males and 436 crescent females, 4986 long males and 5458 long females. Out of a total of 27896 there are 1903 truncates and 1881 crescents. This gives a ratio of 1 trun- cate to 13.2 longs and 1 crescent to 13.4 longs. The ratio is practically the same in both the cross and the reciprocal. To summarize in a general way, we may say that during the period of investigation the truncates threw one long in 7= truncates. Their long-winged brothers and sisters threw 1 truncate in 7+ longs. ‘The new type of wing, the crescent, appeared in the crosses between the truncates and the wild stocks and in the ratio of 1 crescent to 21 longs in both the cross and its reciprocal. In the F, generation crescents and truncates appear in equal numbers from both the cross and the reciprocal and in the ratio of 1 to 14. These ratios apparently do not fall under any Mendelian explanation and yet I believe that any explanation of these ratios must take into consideration the great viability of the truncate stock and my opinion is that when more facts are available this case may be found to fall under a Mendelian formula because there is evidence of segregation and because definite ratios appear according to the combinations that are made. FERTILITY AND STERILITY IN DROSOPHILA PA SEX RATIO It is not the purpose here to enter into a discussion of sex-de- termination. I merely wish to bring together the data from the foregoing experiments that bear on the question of sex-ratios. Since the viability has been so great in some of my strains it seems remarkable that disturbances in the sex-ratio have not been encountered. The tables show the ratios to be remarkably con- stant, with a slight excess of females in almost all cases. The females emerge first and if a count is made at this time there is in almost all cases a large excess of females. At times the ratio may be as high as two or three females to one male. Toward the middle of the count, however, the ratio approaches equality and toward the end of the count the males are usually in excess and this tends to equalize the sex ratio. It is not that the first eggs laid produce females but that the egg which is to develop into a female carries its development through on an average from twelve to twenty-four hours more rapidly than the male. This is evident from the hundreds of bottles from which my counts have been made. Different strains of these flies vary somewhat in their rate of development. The truncates always emerged from two to four days later than the wild stocks used in control. The hybrid fly carries its development through to the hatching stage more rapidly than either of the parent strains. I have said that the sex-ratio is remarkably constant, with the females slightly in excess. This statement applies to the inbred stock, the truncate stock and the Woods Hole stock when brothers and sisters are paired. Diagram 1 shows the ratio of males to females to be 100 to 103; 100 to 103 and 100 to 107 respectively. I think that the slight excess of females is due to the factors to which reference has already been made. When the different races are crossed into each other a rather large excess of females appears. This statement also holds for the children and grandchildren when tested together in pairs. I do not believe that the early emergence of the female from the pupae will ac- PAA ROSCOE R. HYDE 1244 3/26 326 1g : i926 {120:107 3204 100:103 3230f (00:103 asagy 1007102 Control WHoexWHS TS x TP? Te x TS: If « Te Cross * 9A, 38 865 2271 100: : = S 920 ib 100:/19 p22. 919 100:|05 Be pi00 107 F Generation 1542). 053 87. 1828 690)... 5574) 1. 105} 0let 100147 54q 1aqus} 100:106 azeafl0l0e gq.5y!00:130 pozey ouill2 Fe Generation 1959 : 1347 ¥ 56D) a. 1129 S 2229 00:'4 pore fl00l20 “51g 10092 1986 100: 15. wee Ce EE EE EE KS es Diagram I Showing the ratio of males to females. Compiled from the foregoing experiments. T = truncate; WH = Woods Hole; I = inbred. count for the rather large excess of females in these crosses. It should be stated here as a matter of fact that the sex ratio doesnot change with the age of the female.. Diagram I gives the sex-ratio as compiled from the foregoing experiments. LITERATURE CITED Correns, C. 1912 Selbsterilitit and Individualstoffe. Festschrift der Medi- zin-Naturwissensch. Gessellschaft zur 84. Versammll. deutsch. Natur- forscher u. Arzte. Miinster i. W. Reprinted in Biologisches Central- blatt., Bd 33, No. 7, Juli, 1913. : Kast, E. M., anp Havens, H. K. 1912 Heterozygosis in evolution and in plant breeding. U.S. Department of Agriculture, Bureau of Plant Industry. Bull. No. 248. Morcan, T. H. 1905 Some further experiments on self-fertilization in Ciona. Biol. Bull., vol. 8. 1912 a A Modification of the sex-ratio, and of other ratios in Droso- phila through linkage. Zeitschrift fiir induktive Abstammungs. und Vererbungslehre, Bd. 7, Heft 5. 1912 b The explanation of a new sex-ratio in Drosophila. Science, N.S., vol. 36, no. 934, November 22. Suuii, Grorce H. 1911 Hybridization methods in co n breeding. American Breeders Association, vol. 6, p. 68. CONTRIBUTIONS FROM THE ZOOLOGICAL LABORATORY OF THE MUSEUM OF COMPARATIVE ZOOLOGY AT HARVARD COLLEGE. NO. 252. A QUANTITATIVE DETERMINATION OF THE ORIENT- ING REACTION OF THE BLOWFLY LARVA (CALLIPHORA ERYTHROCEPHALA MEIGEN) BRADLEY M. PATTEN TWENTY-FOUR FIGURES CONTENTS SU TRCMITE CON cites EODLCMIR a retea vals ae reel ie ices tee oleh c S(z Mon she wi eck, oS oo ela oR 213 TAUCT ACU te rere ah | Sep Eet yw bac Rae era aie AE a NOON NE Seen 214 Is SC STETOU ITT ee oe Be cae OMe eae nee CES ta ASCO Sap CUTE Aa NE eR isle hain 220 Miatentalyarcl alle th Od Siena iecrseaieieaa peri Cicer CR eros Ahearn eee 223 ie Culburesi.saa2 6503: OR IR PES aes Jace oh i RR Ae ie eR IONE | 223 eV eet Uy l: TESPOHSE + ...3)-0: sacirmcyea,t ele olts eis Waianae’: walba nae eee 224 21 ISKEMIE SS CNG UN CSSA SESS pA EL CREME Oe ec, Oe 9 BE en REE ES 225 aaa ans GeO RITE TGS orien 52,5 ier asda ee RL RN ROe MUN ak hue eiateads 228 See VIE MERURATIS woncers crest sehen 5.5 len eet Re eC LeU AOS IA Soy 2, tL et 229 HPL aAhGn Or asymmetry enc sch. Sel sot tend eee eae oe ee 230 fea Vicasirementand tabulations snr eric cack clin eran rs ae ae oe 232 apular and graphic statement, of results. :2.22s.0.\.. ace sees oes se cs 239 i DVIS CESISTRG) (10) 9 en oe ee oA ee Ee ee ES 244 leanne scalevol reacuhveness aes ctsosee soe ase aener Nas oa, cc ce shee ae hea oe 244 DS COS Rea TIVUN LCs eee es ener Owe tee RS Re ON ee RT 244 | OU ZA OE) PGW Ne Eee EMTS ROUEN oa eae ae CRE pt RA OMB Rao > 0 248 PEC OMES CL OClembalOMes cic at i, wel meh Sve ays aes OSE oynueceaietow. « satan 251 a. Lhe relation of phototaxis to photokinesis...........01..5...-..2. 251 bb: Orientation inthe blowily Varva..)).2.: a ele os bee ee 252 e. Analyais)of factors involved ‘in: phototaxis.......0....0.0010 ends. 272 PS MMMA EN VAN CONEINSLOM BI sibs an ndeeery ace eit Let el eee. sce gay ten reed ae AG 276 | BAT} AIGA cy a) AS a CUD a hee Ae ag Mater oe een Me oe a ER ES 279 STATEMENT OF THE PROBLEM In the work done thus far in animal behavior, no successful attempt has been made to apply any form of precise measure- ment to movements produced by light of known differences in intensity. The experiments on the blowfly larva described in this paper were devised for the purpose of making exact quan- 213 Pie BRADLEY M. PATTEN titative measurement of the light reactions.in some suitable and easily available animal. The light was applied as two opposed beams, the intensity of which could be easily controlled and precisely measured. The responses to the stimulation produced by two beams of different intensity acting simultaneously on opposite sides of the same animal were measured in angular deflections from an initial path of locomotion. LITERATURE The earliest paper on the light reactions of blowfly larvae was that of Pouchet, in 1872. He described a series of experiments with daylight and artificial light, showing the negative character of the light response of the larvae of several species of the old Linnaean family of Muscidae. A considerable portion of the paper is devoted to experiments devised to locate the light recipient organs. The fact that the light response did not dis- appear when the sensory cones of the anterior end were excised, led Pouchet to conclude that the imaginal discs of the adult compound eyes were the sensitive organs. This conclusion was borne out by his observations on the increase of sensitiveness with the age of the larva, which coincides with the increase in development of the imaginal discs. ,The following quotation sets forth his interpretation of the way in which the imaginal dises serve, not only to perceive the light but also the direction of the rays (1872, p. 316): ‘‘La lumiére, frappant, sous des angles différent, les surfaces toutes différent inclinées sur horizon des yeux embryonnaires, donne A l’animal le sentiment de la direc- tion des rayons, par l’intensité relative avec laquelle ils affectent, grace 4 leur incidence, les différent yeux.” This conclusion is interesting in its foreshadowing of the ques- tion of the relative effect of ‘ray direction’ and ‘intensity differ- ence’ on orientation which grew out of the tropism controversy nearly twenty years later. The work of Pouchet, like most studies of behavior of this period was done on the basis of preference as shown by experi- menting. Conditions were so arranged that the animals could move into regions of higher or lower intensity or of particular QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 215 color, and the results were interpretated in terms of human experience. The behavior of an animal that was assumed to . have the freedom of choice was said to indicate its preference. The students of plant behavior—because the subjects of their experiments were remote from human activities—had begun to free themselves from the anthropomorphic point of view long before the zodlogists. With their earlier analytical outlook on the natural phenomena, came an earlier seeking for mechanical explanations. In the middle of the last century, when animals were still going toward light “because they liked it,” or “because it aroused their curiosity,” the botanists were seeking to deter- mine the nature of the mechanism involved in the reactions of plants to light. As early as 1831 de Candolle had used the term heliotropism to describe the bending of plants toward the sun. This he believed to be due to the direct effect of local difference of intensity, growth being retarded on the more highly illuminated side. The inadequacy of this explanation was shown by Charles Dar- win and his son Francis in their work on “The power of movement in plants;” these investigators demonstrated the difference be- tween sensitive and reacting tissues and the transmission of stimuli from one to the other. Sachs, however, did not accept the theories of the Darwins and advanced the explanation that the turning was controlled - by the direction in which the rays of light penetrated the tissues of the plant. The simplicity of the theory of Sachs was very alluring, but, like many ‘simple mechanical explanations,” it fitted only a few of the facts. | Soon after the work of Sachs appeared, Loeb took up the study of animal reactions with the purpose of analyzing the phenomena in terms of physics and chemistry in opposition to the anthro- pomorphic ‘explanations’ prevalent at that time among zoologists. His attitude is well expressed by the following quotation from one of his later works. He says (’05, p. 1x) “I consider a complete knowledge and control of these agencies (which determine be- havior) the biological solution of the metaphysical problem of ani- mal instinct and will.’”’ Whether or not future work bears out 216 BRADLEY M. PATTEN all Loeb’s conclusions is a matter of small importance beside the tremendous advance in clear thinking which has resulted from the analytical attitude he has from the first maintained. In his earliest work bearing on the orientation of animals to light (88), he cites the reactions of fly larvae in support of his explanation of orientation. The main interest of this work was theoretical and but few facts were added to the fairly complete account already published by Pouchet. Perhaps the most sig- nificant of these were his establishment of the restriction of the sensitive region to the anterior end of the larva and the balanced reaction to equal lights acting on opposite sides of the body. Loeb concluded that the factor of prime importance in orientation was the direction in which the rays of light penetrated the tissue, as Sachs had believed for plants. This view, however, he has since abandoned, as may be seen from the following quotation (’06, p. 130), We started with the assumption that the heliotropic reactions are caused by a chemical effect of light; in all such reactions, time plays a role. We assume, furthermore, that if light strikes two sides of a symmetrical organism with unequal intensity, the velocity or the char- acter of the chemical reactions in the photosensitive elements of both sides of the body is different; that in consequence of this difference the muscles, or contractile elements on one side of the organism are in a higher state of tension than their antagonists. The consequence is a curvature or bending of the head. ‘This is followed by a turning of the body, kept up until the stimulus acts equally on the bilaterally !ocated sensitive areas. When such a condition of balance is attained, the ani- mal no longer deviates toward either side, but pursues a direct path toward or away from the source of light. If it be true that the immediate effect of light in causing the helio- tropic reactions is of a chemical nature, we should expect that it must be possible by use of chemicals to control the precision and sense of helio- tropic reactions (’06, p. 131). The striking results which Loeb (’93, ’04, 06) obtained in his experiments on the chemical control of heliotropism are the strongest sort of evidence for his interpretation that light reac- tions depend fundamentally on a chemical reaction, the extent of which is dependent on the intensity of the light. Loeb thus opened a new field of tremendous interest. The mechanical explanation of orientation to light became the ob- . QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 217 jective point of work on light reactions. Interest no longer cen- tered about the cataloguing of the positive or negative response of various forms, but about the developing of a general theory of light reactions. With the possible exception of Engelmann’s work on swarm spores, most of the work on behavior up to this time had dealt with mass reactions rather than with the details of individual reactions. In 1897 Jennings began a study of the aggregations so commonly found in the cultures of Paramoecium, by observing in the minutest detail the reactions that brought a single indi- vidual into the aggregation and held it there. The work on Paramoecium was followed by many other papers on the Protozoa and some on the lower Metazoa, all characterized by the same thoroughness in the observation of individual behavior. Such methods brought out many facts of great interest, previously overlooked. One of the points on which Jennings lays great emphasis is that behavior does not depend on the external stimulus alone. He believes that there are various internal factors which modify the reactions to the same external conditions. Former stimuli and the reactions of the organisms to them, as well as the meta- bolic processes constantly going on within the animal, have their effect on the physiological state. The physiological state in turn determines to a large degree the reaction of the organism to ex- ternal stimuli. Jennings also found that in the Protozoa on which he worked there were no bilaterally located sensory areas and that the posi- tion of orientation was not one in which the median plane of structural symmetry was placed in a definite position with ref- erence to the source of stimulation. The ‘tropism theory,’ as put forward by Loeb, evidently did not apply to these organisms. Jennings observed that changes in the direction of locomotion were brought about by ‘motor reflexes’ directed toward a struc- turally definite side of the organism. He characterizes this method of orientation as one of ‘trial and error.’ The ‘varied movements’ of locomotion involve contact with varying environ- mental conditions, selection from among these conditions is 218 BRADLEY M. PATTEN brought about by the ‘motor reflex’ produced whenever a stimu- lus is encountered. This usually removes the animal from the source of stimulation, for the ‘structurally definite side’ toward which the reflex is directed is that opposite to the side on which the sensitive area is located. If the distribution of the stimulus in space is such that these reflexes hold the animal on a direct path with reference to the source of stimulation, a response resembling a ‘tropism’ results, but the method of its accomplish- ment is different from that assumed by Loeb. Jennings did not, however, believe that the orientation of all animals was brought about by the motor reflex (’06, p. 271). “In the symmetrical Metazoa we of course find many cases in which the animal turns directly toward or away from a source of stimulation without anything in the nature of preliminary trial movements.” About this time Holmes (’05) published a detailed account of the movements involved in the orientation of the individual blowfly larva. He lays great stress on the significance of the side to side swinging of the anterior end which occurs when the larva is suddenly stimulated by lateral illumination. These random movements, he believes, afford a means of selecting a favorable direction of advance. He says (’05, p. 105): There is, so far as I can discover, no forced orientation brought about by the unequal stimulation of the two sides of the body, but an orientation is produced indirectly by following up those chance move- ments which bring respite from the stimulus. I do not deny that there may be an orientating tendency of the usual kind, but if there is, it plays only a subordinate réle in directing the movements of the animal. The orientation of these forms is essentially a selection of favorable chance variations of action and following them up (p. 106). It may be said to be a form of the trial and error method minus the element of learning by experience. Herms has published two articles on the light reactions of the larvae and adults of the Sarcophagid' flies. The earlier paper (’07) is based largely on a field study of the light reactions as they affect food habits and migration. The second paper (711) con- tains an account of numerous laboratory experiments in which he has brought out a number of interesting points, such as the increased rate of crawling under increasing intensity of illumi- 1 Herms bases his classification on Girchner’s system, published in 1896. QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 219 nation, the positive reaction of aggregations of feeding larvae to a light of moderate intensity, and the circus movements per- formed by larvae with one side of the anterior end blackened. A valuable part of the paper is occupied with the confirmation by more exact experimental data, of several of the earlier conclusions. Mast in his extensive treatise on “Light and the behavior of organisms’ has collected practically all the important work of the earlier writers and added many experiments of his own. His original work is similar to that of Jennings in the careful obser- vation of the details of individual behavior. One cannot fail to be impressed by the painstaking methods of his experiments, but it is difficult to see how his results invalidate the theories of Loeb so completely as he would have us believe. Considerable space is devoted to the reactions of fly larvae. Experiments are described which bear out the conclusion of Loeb and Herms that the sensitive region is restricted to the anterior end, but none of them aid materially a precise localization of the sensitive organs. His experiments on the effect of intensity on the rate of locomotion failed to yield any very definite results, largely because they were made with horizontal lights and hence the shadow of the animal’s body prevented the direct operation of the light on the sensitive anterior end. The most interesting part of his work is the series of experiments by which he estab- lishes that, in this form as well as others on which he worked, ori- entation depends primarily on the intensity which operates on the sensitive surfaces, but depends on ‘ray direction’ only in so far as it modifies the operative intensity on the receptive areas. His analysis of orientation will be taken up in same detail in the discussion of theories of orientation. Several authors have experimented with the effect of colored light on maggots, but the work of Gross (’13) is, by reason of the refinements of his apparatus and methods, by far the most ac- curate that has been done is this field. The colors used in his experiments were practically monochromatic and were accurately measured for intensity by means of a radiomicrometer. The sequence of effectiveness which he established for the larvae— green, blue, yellow, red, decreasing in the order named—is unusual in the greater effectiveness of the green than the blue. 220) BRADLEY M. PATTEN It is interesting that in the imago the conditions are reversed, and the blue is more effective than the green. In the work which has been done on the light reactions of the blowfly larva there is essential agreement as to the general man- ner in which the animal actually behaves under various experi- mental conditions. The interest of recent work has centered about the discussion of details of behavior which have been adduced in support of one or the other of the theories of orien- tation. There has been no attempt made to obtain definite measurements of any phase of the orienting reaction. The pur- pose of the present paper is to devise a method of quantitative measurement which shall be available in work on light reactions and to apply such a method to the orienting reaction of the blowfly larva. APPARATUS The apparatus used in these experiments was constructed so that the opposite sides of the animal under observation could be subjected to opposed beams of light the actual and relative intensity of which could be varied at will. The plan of the apparatus is shown in figure 1. It consisted of a horizontal wooden frame in the form of an isosceles right triangle. As shown by the dotted lines on the diagram, this frame was so proportioned that a similar triangle measuring two meters on the base and 1.414 meters on the equal legs, could be laid out on it. Where the side bars of the frame came together to form a right angle, a horizontal platform 65 by 35 em. was attached. On this platform, at the apex of the triangle were set up five 220-volt Nernst glowers, mounted | vertically and about 5 mm. apart, each with a’switch in circuit. Their position was such that no glower interfered with the light thrown on the mirrors by any other glower. When less than the five-glower intensity was desired, symmetrically placed glowers were used. A portion of the light from these glowers passed through the horizontal rectangular apertures (3.2 by 6.1 em.) in the screens d and d’, placed 25 em. from the glowers. QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 22] Fig. 1 Diagram of apparatus used to produce differential bilataral light stimulation. G, five 220-volt Nernst glowers; @ and M’, mirrors; / and f’ central, point of mirrors; O, center of observation stage; dotted lines, central ray of beam of light from the glowers reflected to O by the mirrors; d and d’, screens with rec- tangular openings; sand s’, light shields; a and b, 2c. p. orienting light with screens. The central ray of light from the apertures fell on the centers of the vertical mirrorgM and M’, which were cut from the best French glass, and, toinsure uniformity, fromthe same piece. These mirrors were placed at such an angle that they reflected the central ray of the beams Gf and Gf’ to the observation point O directly opposite the glowers. It will be seen that the apparatus was so constructed that the courses taken by the light over the two paths from G to O were equal in length and symmetrical in position. Further- more, the central rays impinging at O were in the same straight line but came from opposite directions. The two beams of light reaching this central point should be of equal intensity, and care- ful photometer tests showed them to be so. By using a single source of light for both beams, the uncertainty due to the pos- sible fluctuations in two separate sources of light was eliminated. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 17, No. 2 222 BRADLEY M. PATTEN A horizontal stage of slate 30 by 48 em., on which the animals were observed, was placed at the same level as the center of the glowers, and in such a position that the observation point O fell at its center. At either end of this stage, a 2-candle power incandescent lamp was set up at a distance of 55 cm. from the central point O. By means of these ‘orienting lights,’ (fig. 1 a and b) a negatively phototactic animal could be started from either end of the observation plate, directly across the field of light from the mirrors. Light screens around the glowers, s, and a larger set of screens around the observation stage, s’ , shut out extraneous light and reduced reflection to a minimum. The orienting lights also were screened except for a small horizontal aperture throwing its central ray to O. With the apparatus thus set up, a maggot could be made to move on to the observation stage,.away from one of the orient- ing lights, so that two equally intense beams of light fell on its opposite sides. By reducing the intensity of one of the beams of light, differential bilateral stimulation could be set up, of any desired proportion or intensity. It was important to find a reliable method of reducing one of the beams of light. The episcotister, though very convenient, has been shown to be an unreliable means for the accurate reduction of light intensity (Parker and Patten ’12). Several forms of gratings were tried, none of which gave a perfectly uniform field. The difficulty of securing a uniform field, with diaphragms of small enough aperture to give the desired intensity differences, was very great. The most satisfactory means seemed to be to move the observa- tion stage to points on the basal arm of the apparatus where, according to the law I« < the intensities should be of the de- sired ratio. Such a method, though cumbersome, gave an ab- solutely uniform field, and the intensity could be figured with great accuracy.’ 1 * Hyde (’06) has computed very accurately from the law I «,, the amount of d2 variation when the source of light is a cylinder of the dimensions of a Nernst glower insteadof a point. For the distances used in this apparatus, the variation is from +0.08 per cent to +0.03 per cent. QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 223 The percentage differences in the two lights used in this series of experiments were 662, 50, 334, 25, 162, 84 per cent, and as a con- trol, equality. The points on the basal arm of the apparatus at which these percentage differences exist are: 662 per cent, 63.9 em.; 50 per cent, 41.4 cm.; 334 per cent, 24.6 em.; 25 per cent, 17.4 em.; 162 per cent, 11 cm.; 83 per cent, 5.4 em. to the right or left of the central point O. The percentage differences are in all cases expressed in terms of the stronger light. The actual intensities of the beams are given in table 1. With the apparatus set up for unequal opposed lights, negative animals should deflect toward the weaker light, and the de- flection should be greater as the difference between the lights is increased. This deflection measured in degrees gives the desired quantitative expression of the physiological effects of the light. MATERIAL AND METHODS 1. Cultures The raising of blowfly larvae, and the duration of the various stages of their development under natural and artificial condi- tions, has been dealt with by Herms (’07, 711). He also made observations on the differences in behavior between certain of TABLE 1 The intensities in candle meters of all the lights used in the series of experiments. The four-glower intensity is slightly lower than would be expected because the position of the glowers was such that there was a minimum of mutual heating at this intensity. The 834 per cent intensities do not fit in with the series because ‘o obtain such a large difference it was necessary to re-set the mirrors. PER | INTENSITY OF WEAKER LIGHT | INTENSITY OF STRONGER LIGHT CENTAGE | RATIO IN CA. METERS | IN CA. METERS DIFFE R- Or | = : _ NC y | LIGHTS 5 5 = 5 3 ‘ 5 ENCE IN | LIGHT 1 3 / : 4 : | 4 LIGHTS glow-| g.ow- glow- glowe 7 vers glowers glowers lower | glowers glower | glowers glowers glowers glowers| glower | glowers eas. | eas lene 24.6 31 Equality ltol 6.32 13:.9 io) 41.1 | 6.32 13.9 24.6 | 31.5 | 41.1 82% | 11 to 12 6.05 13.4 23.6 30.2 8005). enon 14.6 | 25.7 | 32.9 | 42.9 163% 5 to 6 aed 1228 ODE San 2825 30.0 6.93 15.3 | 26.9 | 34.5 | 45.0 25 % | 3 to 4 5.50 ie 21.4 27.4 Bi ft 7.34 16.2 28.5 | 36.5 | 47.6 333% 2to 3 eral 11-5 20.2 25.9 33.8 7.83 17.1 | 30.5 | 39.0) 50.9 50 % 1 to 2 4.60 10.2 17.9 22.9 29 Os) O22 20.2 | 35.8 | 45.9 | 59.9 663% 1lto3 15.4 46.4 832% | lto6 7.6 e | 46.4 | 100 © Oto x 0.0 | 24.6 | | 224 BRADLEY M. PATTEN the species, a detail which has been too often neglected in work on behavior. ‘Though the species he worked on especially, Lucilia caesar, and Calliphora vomitoria, reacted essentially in the same manner, the former was noticeably more sensitive to light than the latter. Similarly with various Planaria, Walter (’07) has shown well defined species differences in the light response. To avoid inconsistencies which might arise by the neglect of species differences, except for some preliminary experiments, the work was done with the larvae of a single species, Calliphora erythro- cephala Meigen. The larvae do not acquire their maximum sensitiveness until the end of the feeding period (Pouchet ’72; Herms ’11); as they ap- proach the pupal stage they became sluggish, and if not actually less sensitive, are certainly very tedious to record. The most fa- vorable age to make the light tests appeared to be at about the time of transition from the feeding to the migratory stage. Like many other forms, the larvae soon become acclimated to light, and are much more sensitive if kept from one to three hours in the dark. But a long isolation from food and moisture tends to make them sluggish and to hasten the pupal stage. The best procedure is to keep the whole culture in the dark and not remove the larvae until immediately before they are to be used. 2. Variability of response There proved to be a wide range of individual variability in the response to light, the behavior of a few of the larvae being so characteristic that they could be easily recognized by their re- actions. For example, one larva had a preliminary exercise that it performed with great precision in each one of the experi- ments in which it was used. When placed with its anterior end away from the light, in the small wooden groove in which the larvae were started, this larva looped its anterior end directly back toward the light underneath its own body, until the ‘head’ emerged from under the posterior end and was struck by the full intensity of the orienting light; whereupon it promptly uncurled and crawled away from the light, responding otherwise like ordinary larvae. Out of several hundred larvae observed, I QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 225 have not seen this particular feat repeated. Though any ‘ex- treme individuality’ of this sort was uncommon, it was apparent that the rate of crawling and the response to light stimulation varied so greatly with different larvae under the same conditions that it had to be taken into account in a comparative study of a large number of individuals. To compare a larva of unknown sensitiveness, tested under one set of conditions, with an equally unknown larva tested under other conditions, would be of little value as an accurate basis for comparing the effect of the con- ditions. If one were attempting to find the precise effect of humidity on the time made by long distance runners, one would not try various runners of unknown speed, each under a different degree of humidity, and coinpare their ‘times’ as an index of the humidity effect. The only logical method would be to time the same runner under various conditions. His actual time would be an individual characteristic, but the relative time of the records under different conditions would be an index of the effect of humidity. 3. Standard test. But we cannot use this method on blowfly larvae, for the short sensitive period coupled with the rapid changes in the degree of sensitiveness with age, makes it impossible to complete a series of comparative experiments on a single larva. The nearest ap- proach that can be made to such a method is to use, instead of a single larva, individuals which are as near alike as possible. For this purpose a standardization test, as it might be called, was devised and throughout the whole series of experiments only larvae testing to the uniform standard of sensitiveness were used. During the experiments, each larva was kept in a separate, numbered box. The record of the standard test, as well as the subsequent trails, was made by putting a drop of very dilute methylen blue solution on the posterior end of the larva and letting it mark its own course on a sheet of paper placed on the obser- vation stage. This method of recording the trails of larvae was used by Pouchet (’72) and Gross (13). It has been repeatedly checked by control experiments with tap-water and appears to 226 BRADLEY M. PATTEN change the reactions of the larvae in no way. The number of the box in which the larva was kept, in combination with the date of the experiment,’ made an identification number for all the trails of an individual larva.. Thus the whole series of records was made permanent and can be referred to at any time. The trails of the standard test were made in the following manner. A maggot was placed in front of the orienting light (fig. 1, a), which forced it to crawl toward the center of the stage in a line perpendicular to the ray direction of the mirror beams. When it was well on to the stage, the orienting light was turned out and at the same time the side light (fig. 1, fO) thrown on, thus subjecting the larva to strong illumination from the left. The lateral illumination caused the larva to turn to its own right. After this trail had been completed, the side light was turned off, and the larva at once placed at the further end of the observation plate, where it was again started across the path of the mirror beams, this time under the influence of the opposite orienting light (fig. 1, b). The same side light (fig. 1, fO) was then turned on, but as the larva was crawling in the opposite direction, the light now operated on the right photo-sensitive area, producing a turning again toward the observer’s right, but toward the larva’s left. Figures 2 and 3 are photographs of actual trails made in this" way, except that the methylen blue trail has been blackened with india ink to show better in the photographs. The sharpness of the bend in such a pair of trails is an index of the light sensitive- ness of the larva, the symmetry of the curves, an index of the photo-sensitive balance. Only those larvae orienting to the new direction of light so accurately that both trails come to lie ap- proximately parallel to the direction of the rays, were used in making the records compiled in tables 2 to 5. The amount of variation within the limits of accurate orientation, according to this criterion, is shown in trails a, b and c of figure 2, all of which meet the standard requirement of sensitiveness. Though there is a clearly marked response to the new direction of light 3 The method of writing dates employed in these records is the day, month, year, notation; thus 17/6/’13 is the 17th day of June, 1913. QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 227 Fig. 2. Three examples of test trails which pass the standard orientation test. Each pair of trails was made by a different larva. Though there is some differ- ence in the sharpness of orientation, each trail comes to lie parallel to the new direction of light. Fig. 3 Three examples of test trails which do not pass the standard orientation test. Each pair of trails was made by a different larva. Though there is a re- sponse to the new direction of light in all cases, the larva fails in one or both of its trails to come to complete orientation. 228 BRADLEY M. PATTEN in the trails shown in figure 3, a, b and c, these trails do not come to lie parallel to the new ray direction. These larvae were sen- sitive to the light but not sufficiently so to orient accurately and therefore were not used in making the records on which the tables are based. In testing specimens from cultures at the optimum stage, from 15 to 25 per cent of the larvae were discarded because they failed to orient sharply to the change of direction in the light. Al- though the maggots from the same cultures were all of the same age, they were not necessarily in the same stage of development. The discarded larvae probably represented individuals that had passed, or not yet reached, their period of maximum responsive- ness, rather thar those which would never reach the normal degree of sensitiveness. The larvae showing the desired accuracy of orientation were laid aside in numbered boxes and, after a rest, subjected to bilateral illumination from the mirror beams. Such a selec- tion gave uniformity to the animals used in compiling the final results, and though it raised their standard of sensitiveness somewhat above that of the general population, it did not in any way distort the relative values of the subsequent records; on the contrary it is only by thus insuring uniformity in the material used that the comparison of results from different larvae under various experimental conditions becomes of value. 4. Plan of experiments The plan of the experiments was to subject larvae of standard sensitiveness to lights of accurately determined intensity, with the conditions so arranged that the reactions could be measured in physical units. The rate of locomotion of the larva under the influence of light of various intensities has proved a difficult and unsatisfactory basis for a quantitative study (Mast 711, pp. 184-189). But orientation is a phase of the response to light which lends itself admirably to accurate measurement. According to Loeb (’05, p. 2): “If two sources of light of equal intensity and distance act simultaneously upon a heliotropic QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 229 animal, the animal puts its median plane at right angles to the line connecting the two sources of light.’”’ It follows, as a corol- lary to this hypothesis that if the sources of light are unequal, a negatively heliotropic animal would deflect from the perpendicular toward the weaker light. This has already been demonstrated by Mast (11, p. 179) for the blowfly larva. By measuring in degrees the deflection toward the weaker light, physical units of measurement may be applied to the physiologi- cal phenomenon. And since with equal lights there is no deflec- tion of the trail from the perpendicular to the line connecting the sources, whatever deflection there may be with unequal lights may be regarded as due to the effect of the difference between the lights. The starting point for a series of experiments of this sort would be at equality of the two beams of light. But it is not enough to know merely that there is no marked deflection when the lights are of equal intensity. It must be determined how many degrees from the perpendicular the aggregate re- sponse of a number of individuals falls. This necessitates not only a method of measuring each trail, but also a method of com- piling the data obtained by measuring many trails made by different individuals. The way in which we have obtained our results may be seen from a statement of (1) the method by which the trails were made, (2) the method of measuring single trails, (3) the method of tabulating the measurements. 5. Making the trails The handling of the larvae in making the records is of sufficient importance to be described in detail. At each end of the obser- vation slate was a small wooden block, not shown in the dia- gram (fig. 1), with a groove in it leading down to the level of the slate. The blocks were located just in front of the orienting lights, the grooves extending along the line connecting the two lights (fig. 1, a and b). The larva to be tested was gently rolled from its box into the groove in front of the nearer orienting light (fig. 1, a), with its head away from the light. All other 230 BRADLEY M. PATTEN lighfs were shut off; the mirror beams by slides, and the opposite orienting light by a switch. The larva, started straight by the groove, was forced to move on to the paper in the direction of the orienting light, and so directly across the field of light from the mirrors. When it was well on the paper, the slides were pulled simultaneously from the two mirror beams, and at the same time the orienting light turned off. This subjected the larva to equal bilateral stimulation by the two beams of light, for the orienting light brought it into the field with its median plane at right angles to the line connecting the sources of light. Under these condi- tions there should be no deflection toward either light, and in fact there was none, for the larva continued in the direction in which it was started by the orienting light (fig. 4, b). At the end of this trail the maggot was allowed to crawl into its box, held edge down to the paper, then without being touched in any other way, it was rolled into the groove in front of the opposite orienting light (fig. 1, b) with its anterior end toward the center of the stage, and driven back across the field of light from the mirrors. The lateral beams were then thrown on the larva again and the second trail completed in the same way as the first. After a rest, a second pair of trails (fig. 4, c) was run in the same way, except that the order of running was reversed and the trail from the farther orienting light run first. These four trails, together with the test sheet, form the complete record of an individual and the unit for compiling the tables. 6. Elimination of asymmetry But the trails do not all conform so closely to the theoretical response as those of figure 4. For example, those shown in figure 5 have a marked deflection under the same conditions of equal bilateral illumination. A careful study of this set will show the necessity of running the trails in pairs from opposite directions. In the lower trail (fig. 5, b), the larva deflected quite markedly to the left when the side lights were turned on, exactly as if the light from the right were stronger. But when its direction of crawling was reversed, it deflected to the right as if the left hand QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 25] * « E-19 ’ E-19 E-IQ | b : 4 Sah —|5° E-29 E-29 45" E-29 . +9° a | b Cc 5 Fig. 4 A set of trails made under equal opposed lights by a symmetrically responding larva; a, the test trails; b and c, trails under the influence of balanced lights. Fig. 5 A set of trails made under equal opposed lights by an asymmetrically responding larva; a, the test trails; b and c, trails under the influence of balanced light. Dae BRADLEY M. PATTEN light were stronger. The trails are nearly perfect mirror images of each other. Although this deflection is first to the observer’s left and then to his right, it is in both cases toward the larva’s left, for the direction of its crawling was reversed between trails. Such a response indicates an asymmetry in the neuro-muscular mechanism of the larva. The right side must be either more sensitive, or muscularly more active than the left for it gives a greater response to the same amount of stimulation. This asymmetry is by no means as uncommon as might be expected, in fact a certain degree of asymmetry is far more common than even an approximately perfect balance of sensitiveness. Whether the asymmetry is anatomical or physiological is a question which will be taken up later. The point we wish to bring out here is the method of eliminating the effect of such asymmetry, for it might easily be a source of serious error. Many markedly asymmetrical larvae were thrown out by the preliminary test (fig. 3, a). But many of those giving a perfectly symmetrical response in the test-trails showed marked asymmetry when run in the balanced lateral beams. The reason for asymmetry appearing in the later records, when it was not shown in the test trails, I believe to be this. In the.case of the test trails, the larva is always orienting to a single light, and when orientation is attaimed, the anterior end of the larva is in the shadow of the posterior end. If the direction of light is changed, the larva keeps turning till it again crawls with its sensitive anterior end within its own shadow. If the ‘head’ swings out of the shadow, it is strongly stimulated and swings back again into orientation. Even if there were not a perfect balance of sensitiveness on the two sides, the lack of balance, unless it amounted to almost total insensitiveness of one side, would not appear under these conditions, for if, as the ~ larva throws its anterior end from side to side in crawling, it passes beyond the boundary of the shadow, the change of in- tensity is sufficiently abrupt to produce a response on the less sensitive as well as on the more sensitive side, thus holding the animal within a course sharply limited by its own shadow. When on the other hand the larva is made to crawl in a field of balance light, both lateral surfaces are illuminated, no matter QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 233 which way it turns, and the only means of maintaining its course is by a balance in the sensitiveness of the photo-receptive areas. This is clearly shown by trails such as those of figure 5. The guiding effect of the shadow is sufficient to obscure the unbalanced response in the trails of the test sheet and in the part of the trail made under the influence of the single orienting light. These trails lie almost exactly in the direction of the rays. But when the equal side lights are thrown on, which with a balanced larva should produce a direct continuation of the trail made under the influence of the orienting light, the asymmetry becomes at once evident in the deflection to the left. This larva, which gave a symmetrical test, proved to have the right side more sensitive than the left. Owing to some bilaterally unbalanced factor within the organism itself, it behaved, under the influence of equal lights, as if the right-hand light were stronger. Hence the deflection of a single trail is of little value in estimating the light effect. But turn the larva around and make it crawl into the field of light from the opposite direction (fig. 5, 6), it still swings toward its own left side, but by reversing the direction of the crawling, the expression of asymmetry has been made to fall on opposite sides of the perpendicular to the line connecting the sources of light. If we measure deflection to the right in plus degrees and deflection to the left in minus degrees and add the two, the trails of an asymmetrical larva, run in pairs in opposite directions, are reduced to the theoretical response, and to the actual response given by perfectly symmetrical larvae. 7. Measurement and tabulation The trails are measured as shown in the diagram (fig. 6). The lines zy and x’y’ are drawn through the trails at the point where the larva was when the side lights were turned on, and perpendicu- lar to the line connecting the sources of light. A protractor is laid on the trail, with its center at the point where the lateral lights were turned on, and the deflection is measured in degrees. I have called this the angular deflection of the trail. The measurements obtained from each individual were col- lected in tables such as table 2. The deflection of each trail is 234 BRADLEY M. PATTEN entered in the appropriate column, and the sum of the total minus deflection and of the total plus deflection of the four trails is entered in the net column. The sum of the plus and minus deflections of a larva divided by the number of trails made gives x? 10/9/7138 F-16 Vig. 6 Diagram to show the method of measuring trails. The lines zy and x’y’ are drawn through the trails at the points reached—marked by the arrows—when the side lights were turned on. The angle of deflection from this line is measured by a protractor, P. The small figures near the arrows indicate the number of wig-wag movements made when the side lights were turned on; Ist and 2nd refer to the sequence in which the trails were run; No. 7 is the box number, Which, in combination with the date, gives the identification number of the larva. the average angular deflection of the larva. The mean of the average deflections of a large number of individuals forms the final quantitative expression of the response to a given set of conditions. A study of this table shows that very few trails QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 235 TABLE 2 Equality. Four-glower intensity | NUMBERS OF THE LARVAE. Ist SHEET | 2ND SHEET | AVERAGE ae So a Bs ____| Tora | ToTaL| \yq | ANGULAR DATE Ist 2nd | 3rd 4th = | ar | DEFLEC- trail trail trail | trail | DADs: No. 14, 13/5/7183 +10} —16| +15|—7 | 238 | 25 | +2 | +0.50° No. 18, 13/5/7138 0 1+6 |+2 |4+5 | O | 18 | +13 | +3.25° No. 20, 13/5/7138 | +7 |—-8 | +18) —12|) 20 | 25 | +5 | 4+1.25° No. 15, 13/5/713 Papo i mente ie 8). OD) |e OR) Galette No. 4, 13/5/7138 SG Ge eto: ae 0 43> RS) et Wee No. 22, 13/5/7183 0 |-+5 |—7 | —1 8 5 | =3 h=O.75° No. 21, 13/5/7138 —5 | —3 | -21| +17} 29 | 17 | —12} —3.00° No. 16, 13/5/13 | +17/-14] 0 |-6 | 20 | 17 | —3 | —0.75° No. 11, 13/5/7183 /—9 | -1 | +13/-6 | 16 | 13 | —3 | —0.75° No. 9, 13/5/713 =I | 45) |) =5 -}-43.)| 16) 8 | =8° 2.008 fa nae | 167° | 166° | —1° | —0.025° conform exactly to the theoretical response. Out of the forty trails, only three show no deflection whatever. But the total plus and minus of each larva comes near to cancelling, and the average of the four trails of a larva is close to zero degrees de- flection. The average response of the whole set corresponds almost exactly to the theoretical response, the average deflec- _tion from the perpendicular being only —0.025°. This is aston- ishingly accurate when one considers that a degree on the pro- tractor of 7.6 em. radius used in measuring the trails was only about 1.5 mm. The striking results of table 2 were confirmed by similar results at various intensities of the equal beams, the details of which appear later in the paper. The immediate significance of these results is two-fold. The consistent closeness of the average trail to the perpendicular, and the equal distribution of the trails on either side of it, indicate that the method of individual measure- ment has effectively eliminated asymmetry and placed the differ- ent individuals on a uniform basis for comparison. The close conformity of the aggregate response to the theoretical response, when the lights are of equal intensity, makes a well grounded point of departure for a whole series of experiments with opposed 236 BRADLEY M. PATTEN TABLE 3 Measurements of the trails of figures 7 and 8 in tabular form. No. 16, 22/2/'18, made the trails photognonled in eg 7; No. 22, OS those of figure 8 ee he Bs. ISTSHEET | 2ND SHEET | AVERAGE NUMBERS OF THE LARVAE. £ | Pe ee | aeeaae inacas — er NET | DEFLEC- DATE Ist 2nd | 3rd 4th | | eats trail trail | trail | trail | | INO, WG PHA —3 —12 —6 —1l1 | 32 | —32 , —8? No. 22, 30/1/’03 —29 | +17!) -—37 +8 GON 2D | aul —10.25° aa a ee ») i} . 4 larvae | = » é d 98> | 25> | =(34 — Jeo 8 trails | | lights of unequal intensities. For if there is no deflection toward either side following equal bilateral stimulation, the deflection appearing under unequal bilateral stimulation may be regarded as a true expression of the physiological effect of the difference between the lights. The same method of taking the records at equality was used with unequal lights. There is, however, one additional precau- tion to be observed. A given side of the larva should be subjected first to the stronger light in one pair of trails, and first to the weaker light in the second pair of trails, thus avoiding possible cumulative effects such as might result if the stronger ight acted first on the more sensitive side in both pairs of trails. Asym- metry of response, though not cancelling to zero in this case, is checked out by the same method of running pairs of trails in opposite directions. The stronger light acting on the more sensitive side of the larva gives an abnormally great response; when the direction of the crawling is reversed, the same light acting on the weaker side gives a response correspondingly below the normal. The excess of one response is equivalent to the deficiency of the other, and the two average to a viens sym- metrical response. This is clearly shown by the two sets of trails photographed in figures 7 and 8, which were made under the same percentage difference in the lights, one by a symmetrical and the other by an asymmetrical larva. Tabulating these trails side by side, by the method already described, we obtain the results shown in table 3. 25% 59 25°%o 59 eo - a ‘e -I2 25% 59 ay b V. 2OOLn 25°% 59 +17 8 QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 237 25% 59 = Sif 25% 59 Fig. 7 A set of trails made under lights of 25 per cent difference, by a sym- metrically responding larva; a, the test trails; b and c, the trails made under the influence of the opposed lights. Fig. 8 A set of trails made under light of 25 per cent difference, by an asym- metrically responding larva; a, the test trails; b and c, trails made under the influence of the opposed light. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 17, NO. 2 238 BRADLEY M. PATTEN The total response of the four trails is nearly the same for the two larvae. Dividing the total response by the number of trails made, the average trail of the individual is obtained. This is for the symmetrical larva (No. 16, 30/2/13), 8°, and for the asymmetrical larva (No. 22, 30/1/’13), 10.25°. A similar case at 50 per cent difference is illustrated in figures 19 and 20, page 260. The conformity of these averages from such apparently different trails, together with the similar case already considered at equality, is convincing as to the completeness with which the disturbing effect of the asymmetry has been eliminated. I have laid special emphasis on this matter of asymmetry and the method of dealing with it, because it is important that the final results should be free from the possible cumulative effect of in- dividual eccentricities. The average reaction of a large num- ber of individuals, tabulated in this manner, is used to establish the angle of response under each set of conditions. In spite of the precautions used in obtaining the records, ex- tremely aberrant responses occasionally appeared. In all the tables, the average deviation of the individual responses from the mean of the set was computed, and when any records showed a departure of more than three times the ‘average error,’ they were not entered in the tables. This selection was not applied to separate trails but to the records of a larva as a whole. Either all the trails of a larva were accepted, or the larva was thrown out as an extreme variant and none of its trails were used. For these discarded trails, the trails of a fresh, tested larva, run under the same conditions, were substituted. The tables were computed both with and without this correction and the difference was found to be very slight, as the records exceeding three times the ‘average error’ were very uncommon, many of thé tables not showing a single case. QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 239 TABULAR AND GRAPHIC STATEMENT OF RESULTS Having perfected a system of measuring the reaction of the larvae to light, and established a fixed point of departure by means of the experiments with equal, opposed lights, we were able to attack the problem which actuated this work, namely, a tabulation of the reactions to opposing lights of different in- tensities, acting simultaneously on opposite sides of the animal. It seemed advisable to begin at a well defined difference of intensity and work toward equality. Accordingly the apparatus was set up for a difference of 50 per cent between the opposing lights. By throwing glowers in or out of circuit, the absolute intensity of the common source of light could be varied at will, but however the source was varied, the relative intensity of the opposed beams remained the same. The reactions were all measured and tabulated as described in the preceding section, but as it would be impracticable to reproduce in detail over 30 tables comprising the measurements of 3000 trails it has been necessary to condense greatly the tabular statement of the results. : Table 4 was compiled from the reactions of larvae stimulated on opposite sides by lights of a constant difference of 50 per cent. The absolute intensities of the opposed beams (table 1) were varied by using from one to five 220-volt Nernst glowers as sources of light. Each column in the table shows the average response of 25 larvae, each run four times. The average angular deflec- tion of these 100 trails is taken as the measure of the aggregate response to the special conditions. Table 4 is a condensation from five tables such as table 2, each column in table 4 represent- ing the “Average angular deflection” column of table 2. In table 5 the statement of the measurements is still further con- densed. By comparing tables 4 and 5 it will be seen that the 50 per cent column of table 5 corresponds to the ‘Averages for 100 trails” given at the bottom of table 4. Table 5, therefore, including as it does similar summaries of the measurements at all the other intensity differences used, is a tabular statement of the final results of the whole series of experiments. 240 19 Angular deflection at 60 per cent difference BRADLEY M. PATTEN TABLE 4 46° 2 GLOWERS | | 3 GLOWERS 4 GLOWERS 5 GLOWERS Averages from four trails of a single larva | 19.25° | 13.25° 20:25°° 0 |. Raa | 98./50°° >) 4 19.25° 19.25° 15.50° | Dt 15° (ie 20 .50° 24 .25° 15.75° 20.00° 34.00° 17 .25° 29 25° 13.50° 26 .50° 19 .50° 11.50° 24 25° 2o750° — 18:50° * 4 16.75° 32.00° 17 .25° 19600?) “| 21 25° 23 .75° 18 .50° JOB eh 19.50° 23 .25° 29 50° 22:50" | | aSsae? 19.50° 16.00° 15.25° 21 .00° 13.25° 25.75° 11.75° 17 .25° 29 .00° 31.50° 25.00° 24 .00° 21.25° 15 .25° 17 .00° 10.75° 15 .25° 22.25° 16.50° 23 .00° 17 .25° 15.50° 28 .50° 23 .00° | 15.50° HT B02!" y 17 .25° 9.00° 23 .00° 17. 25°." — 8.75° 16.75° 31.75° Dees 14.75° 14.50° 27 .25° 18.50° 16.00° | 22.75° 25.75° 17 .00° 20350°° > 4 21.00° 20.00° 33 .50° L7AT5o, a 20.50° 34.75° 18 .50° 725° 22.75° 24 50° 10y75°4 . | 20 .25° 26 .00° 12.25° 14002 "4 29 .50° 24.75° 20 .00° Et a 27 .25° 22 25° Averages for 100 trails 22.989 20.52° 19.88° 1925" TABLE 5 Based on the measurements of 3000 trails, showing the average angular deflection at five different absolute intensities and nine relative differences of intensity 4 Glowers 5 Glowers Average | | EQUAL-| 8} PER ITY —0.55° —O7107 +0.45° —0.025° —0.225° —0.09° 163 PER | 25 PER | 334 PER| 50 PER CENT CENT | CENT | CENT CENT —2.32° | —5.97° | —9.04° | —11.86°| —19.46°| —3.05° | —6.12° | —8.55° | —11.92°| —22.28° —2.60° | —5.65° | —8.73° | —13.15°| —20.52° —2.98° | —6.60° | —9.66° | —11.76°} —19.88° —2.925°) 5.125" —8.30° | —10.92°, —19.25° a Ss = a = —2.77° | —5.75° | —8.86° | —11.92°| —20.28° | 662 PER | 833 PER | 100 PER CENT | CENT CENT —30.90°, —46.81° —77.56° | —30.90°| —46.81° 1 QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 241 E-I9 83% 49 E19 Ewe is ~ b € 9 & oe 832% Bi% 49 49 b c 10 Fig. 9 The complete record of an individual larva run at equality, selected to represent the average deflection of all the trails‘made under the influence of equal lights. Fig. 10 An average set at 8} per cent difference Figures 9 to 14. show graphically the amount of deflection occurring at the various intensity differences. They are photo- graphs of actual trails, selected because they represent the aver- age deflection of all the trails run at the stated intensity difference. 242 BRADLEY M. PATTEN —=4 163 163 49 49 16% a 49 7 . & | | 11 =| 25% 25% 25% 59 25 59 ~3 a | teak) c 12 Fig 11 An average set at 162 per cent difference 12 An average set at 25 per cent difference oto are QUANTITATIVE DETERMINATION OF LIGHT cog ae oe oe 22 ae (5) Ist 13 -23 50% 39 -17 14 REACTIONS 334% 59 50% _94 39 Fig. 13 An average set at 331 per cent difference Fig. 14 An average set at 50 per cent difference 243 Ist 244 ; BRADLEY M. PATTEN In the subsequent theoretical discussion, the data have been somewhat rearranged according to the point under consideration. They afford (1) the possibility of constructing a curve of reactive- ness to differential bilateral stimulation; and (2) a means of arriving at several facts of interest in connection with theories. of orientation, because the method of light control gave an op- portunity to measure the simultaneous responses of opposite sides of the same animal to equal stimuli or to stimuli of known differences of intensity. The constancy of the angular deflection at a given intensity difference (table 5) even though the absolute intensity of the lights was varied from one to five glowers, strongly suggests that the response follows the principle of the Weber-Fechner Law. A consideration of this phase of the reactions, however, has been reserved until further experiments can be made. DISCUSSION 1. The scale of reactiveness a. Determination. From the data already presented a scale — of reactiveness covering the special conditions of these experi- ments may be determined for the blowfly larva. It has been shown: that under equal opposed lights, there is no deflection toward either side in the aggregate response, and that if the intensities on opposite sides of the animal are made unequal, a deflection appears which becomes progressively greater as the inequality of the lights is increased. Apparently the relative difference of intensity determines the amount of the deflection, for it will be seen by referring to table 5 that the deflection re- mains practically constant for the five different absolute intensity determinations made at each of the percentage differences from equality to 50 per cent. The average of the five absolute in- tensity determinations at a fixed relative difference of intensity would therefore represent very accurately the aggregate response at the given ratio and the value of the average as determined for each percentage difference between the lights may be used to locate points on a curve which may be regarded as the blowfly QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 245 3 i ° LS fe f rad Ww ae (a) HHS uw [oe fo) es, o Ww WwW a o uw a Hietinat a i 4 ina ae PERCENTAGE DIFFERENCE IN LIGHTS Fig. 15 A curve representing the angular deflection under opposed lights of graded differences of intensity, based on the measurement of 2600 trails; see also table 5. larva’s scale of reactiveness to differential bilateral stimulation by light. Such a curve has been constructed in figure 15, plotting angular deflection along the axis of ordinates and percentage differences along the axis of abscissas. Figure 16 shows nine pairs of trails made at the nine intensity differences used to plot the curve of figure 15. These particular trails were selected because they coincided with the average deflection of all the trails run at the stated differences of intensity and therefore give a very clear picture of the experimental results on which the curve was based. +2 E-39 peg 2 163 O% 49 aY 12 —19 —10 49 100% 25°%o 663% 50% 39 -Ist —15 e —2I a Ki € f Ist x 831% 5 100% | . Fig. 16 A series of trails showing the progressive increase in angular deflec- tion with increasing intensity differences between the opposing beams of light. Each pair of trails coincides with the average deflection of all the trails obtained at the stated differences of intensity; a, equality; b, 84 per cent difference; c, 163 per cent difference; d, 25 per cent difference; e, 33% per cent difference; f, 50 per cent difference; g, 663 per cent difference; h, 83} per cent difference; 7, 100 per cent difference. 246 QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 247 The consistency with which the deflection remained constant at various intensities of a fixed ratio made it seem unnecessary to run the whole series of intensities to establish the last three points on the curve, so the three-glower intensity alone was used at 662, 834 and 100 per cent difference. oa c — a Fig. 18 Trails made by an asymmetrically sensitive larva under non-directive light; a, the test trails to horizontal lights, showing the asymmetry of response; b and c, trails of the same larva started in the orienting lights and subjected to vertical illuminating at the points marked by the arrows. A definite orientation by the method of ‘trial and error’ where the ‘trials’ all result in the same amount of stimulation is hardly possible. Apparently the only explanation available is some form of the hypothesis advanced by Loeb, that equal intensity operating on symmetrically placed photosensitive areas produce a symmetrically distributed response. There is considerable experi- mental evidence that may be advanced in support of this view. Herms (’11, p. 207) blackened one side of the sensitive anterior region and found that when the animals so treated were subjected to non-directive light from above, typical circus movements were QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 2059 produced with the blackened side toward the center of the circle. I tried a similar experiment with larvae that showed a markedly asymmetrical response to equal horizontal beams of light by sub- jecting them to light from above. The tendency to turn toward the less sensitive side was very marked. The turning, under non- directive light, of larvae that have been made artificially asym- metrical in their sensitiveness to light, and the similar turning of larvae naturally asymmetrical in their sensitiveness is hardly to be explained on any other basis than the assumption that unequal stimulation of opposite photosensitive areas is proportionately transmitted to the muscles concerned in locomotion. A significant fact in this connection is that the balance of reactiveness in many of the larvae was found to fluctuate; a larva, which in one experi- ment gave a symmetrical response, often giving soon afterwards under the same experimental conditions, an asymmetrical one. Figure 20 (p. 260) isa photograph of the test trails, and six pairs of trails made by the same larvae under a constant intensity differ- ence. The response is first ‘left-handed,’ D, (i.e., stronger to the left than to the right, implying that the right photosensitive area is more sensitive), then right handed, c, then symmetrical, d and e, then right-handed again, f, and finally symmetrical, g. Walter (’07, p. 59) suggested that the asymmetry of response to light that he observed in planarians was due to internal irregularities. Undoubtedly asymmetry of response depends in many cases on anatomical differences in the opposite sides of an animal. In such cases the asymmetry probably does not change markedly in repeated reactions. But in cases like the one cited, where the stronger response is first on one side and then on the other, the balance which is disturbed must be a physiological one. We may assume that the processes of metab- olism, or the previous reactions, or ia short any of the factors which Jennings regards as effective in altering the ‘physiological states,’ have produced a change in the relative abundance or instability of the photosensitive chemicals of the receptive areas. Such an assumption is not unreasonable in view of the well known ex- periments of Loeb on the modifiability of light reactions by the use of chemicals; and the inducing by similar means of a light PATTEN BRADLEY M. ‘P1098 9YF JO OUIOD LOMO] 4J9] BUEI4x9 oY} 4B pooeld oe Splovoed ojvivdos Sulzwusisap ‘099 ‘gq ‘p ‘919449] 2YT—' ALON ‘s}q3I] posoddo oY} JO soUaNpur oy} JopuN opeUT spresy ‘6 07 q ‘s]rw17 480} 04} ‘vy ‘ ¢°ez UoTpPBep IB[NSuB ssv19AB ‘,/0E UOTJIYep [RIO], ‘esuodser Jo souvjequn Surdava YAM pepoBor YOY VAIv] B Aq ddUoIOYTp quso sod YG yw opwUt S]IBI} DATOMT, OZ ‘SIT *s7qS1] posoddo oy} Jo souongur 044 TopunN ospeul s[ivsy ‘6 04 g ‘s]res4 4899 oY} ‘v $F WOTPOOPOp LV[Nduv osvs19awe ggg UoTyoyep T8}0T, “ynoysnoiy} AT[vor4outurss papuodsor yorum vVAIv] @ Aq aoUoIOYIp yuad 19d Qe 48 apeu S]IV1} DATOMT, GT “BIW 02g 6 p 2 q D ay = 2z- oy él 805 6G S| 6e 6€ él a 9 ISI-6E 0g ee %oOS %o0S\ % %00! fez % eas 9008 46001 0S -{oz- Rae oe = - Zo- ) 91 2 A 1s| 4 1S] iss - 8o- 6 fo) 3 %OS %O08 %0S \8E 195 le+ 6l- tJ] ; 1S} eats QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 261 response in an animal which was normally indifferent to light, as recorded by Moore (’138). The cases of natural and artificial asymmetry of suscepti- bility to light are not the only indications that the light recep- tive system operates on the musculature bilaterally. The deflection of symmetrically responding larvae toward the weaker light is in every way comparable with the deflection of asym- metrical larvae toward their less sensitive side when the illumi- nation of both sides is equal. In one case the unbalanced factor of the response is external and in the other case it is internal. Moreover, it is possible to balance the internal asymmetry of the photosensitive areas by subjecting them to a corresponding in- equality of illumination from without. For example, suppose, when a larva is subjected to equal lights from opposite sides, it deflects 5° toward the left. By referring to the curve of figure 15, it will be seen that, with asymmetry balanced out, a deflection of 5° is produced by a difference of 15 per cent between the opposed lights. With the left-hand light 15 per cent stronger than the right, the same larva crawls straight down between the lights like a symmetrically responsive larva crawling under the influence of equal lights. The physiological asymmetry has been corrected by a difference of 15 per cent in the bilateral stimulus. Thus we can not only demonstrate that certain individuals do not have a perfect bi- lateral balance of sensitiveness but we can measure the amount of the asymmetry and correct it by applying bilateral light stimuli of a corresponding difference in intensity. The evidence that symmetrically located sensitive areas operate bilaterally on the musculature may be summarized as follows: (1) When the lights acting on the opposite sides of the larva are equal, the larva orients so that its median plane is at right angles to the line connecting the sources of light. (2) When the opposing lights are unequal, a deflection toward the weaker light appears. (3) Certain larvae are asymmetrical in their response, deflecting toward the less sensitive side when subjected to equal bilateral illumination. (4) The blackening of one side of the sensitive region produces a deflection toward 262 BRADLEY M. PATTEN the side artificially made less sensitive. (5) Asymmetry of sen- sitiveness may be balanced by a corresponding inequality of the stimuli acting on opposite sides of the animal. There seems to be no explanation for the response of the blowfly larva to opposed lights other than the assumption that symmetrical sensitive areas operate on the musculature of the two sides of the animal in proportion to the stimulation received. We do not know precisely what mechanism is concerned in the re- action, nor even what it is that is ‘balanced’ in the receptors. There are, however, certain general lines on which such a mechan- ism must be based. If the angle of orientation under opposed beams of light is such that the stimulation of the opposite sides is equalized, the receptive mechanism must be of such a nature that varying the axial position of the animal produces changes in the relative amount of stimulation received on opposite sensitive areas. Otherwise there would be no cause for the animal to as- sume a definite angle of orientation for each intensity differ- ence between opposed beams of light. This equalization cannot be accomplished by a median sensitive area unless we assume that the area operates differentially on opposite sides of the median line, an assumption which throws us back again to bilaterality. Nor can 1t be accomplished by bilaterally located sensitive areas that are parallel to each other, nor by fixed eyes so placed that the tangents to the eyes at the optical axes are parallel. This can be made clear by a diagram such as figure 21. The heavy black lines represent light-sensitive areas of, let us say, 1 sq. cm. of surface. In figure 21, a, the sensitive surfaces, being perpendicular to the rays of light, intercept 1 sq. em. of the light as represented by their projection on a plane at right angles to the rays of light (double lines of fig. 21, a). In figure 21, 6, the planes have been rotated through an angle of about 45 degrees. This cuts down the amount of light falling on each sensitive surface. But since the surfaces are parallel, their projection on a plane at right angles to the rays of light will be the same, and the amount of light falling on the two sensitive surfaces will still.be equal. The case illustrated by the diagram, figure 21, c, is essentially like that of figure 21, b. The retinas will receive light proportional to that which falls QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 263 eaeeas SE te rad Hie i i es | mA fe cuene name a Hiss id a auras o qapens sue faces eugespangeae pgecanacase itvsencene Lh - Ett ities aii sestecses cerseesszea] ps & co By Feay green feat so ea qe & Scammed essielss (ee aaeee bay Eres Doo He as 7 Bg em, Se Heo aH ds, fess ine oe i ee Say taets Bhi ress Bitten — EY: & : aa Fig. 21 A diagram showing that changes in the axial position of an organism with parallel sensory areas, or eyes with tangents at the optical axes parallel, do not change the relative amount of light intercepted by the sensory areas. In this figure, the lights, indicated by arrows, are assumed to be equal in intensity and opposite in direction. on a plane paratangential to their optical axes at the focal dis- tance of the dioptric’ apparatus. If these paratangents are parallel, rotation of the axes of the system will not change the relative amount of light intercepted by these planes. If, how- ever, we assume that the sensitive surfaces are inclined at an angle to each other, the case is different. Figure 22 is a diagram constructed to show the conditions set up by rotation when the 264 BRADLEY M. PATTEN ie eda e Seritias PAs Be, Gini Seacees| prerreth 5 sopra A Fig. 22 A diagram showing that changes in the axial position of an organism with non-parallel sensory areas, or eyes with tangents at the optical axes non- parallel, change the relative amount of light intercepted by the sensory areas of opposite sides. Lights same as in figure 21. sensitive surfaces are not parallel. In figure 22, a, the opposite surfaces intercept equal amounts of light because their angles of inclination are equal. In figure 22, b, the left area lies nearly at right angles to the rays of light and its projection is equal to its entire area, while the right side is inclined so that its projection is much less than its area. Consequently the amount of lght acting on the bilateral photosensitive areas is different. In figure 22, c, the case is similar to that of 22, 6. The amount of light received on the retina may not be exactly that operating on the paratangential plane at the focal distance of the eye, but it will be proportional to it. QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 269 Fig. 23 Geometrical construction so show that when the sensitive surfaces are inclined to each other, a change in the axial position of the animal may equalize the light energy operating on them, although the intensity in the opposite sides of the field is unequal. The arrows indicate the direction, the numbers, the intensity of the lights; for construction and proof, see text. We may assume, therefore, that the power to balance unequal bilateral stimulation, by a change in the position of the axis of the body depends on the inelination to each other of the photo- sensitive surfaces. The mechanism is simple and effective, and the explanation given fits the facts so far as they are known. If this assumption be true, it should be possible to compute the angle of sensitive surfaces for an animal when we know its angle of orientation under varying inequalities of illumination. ‘This angle has been computed for the blowfly larva, using the ‘angular 266 BRADLEY M. PATTEN deflections’ already ascertained. The magnitude of the angle may bear no direct relation to the actual angle at which the sen- sitive areas are located in the body of the animal, because of the many factors which may modify the direction of the rays before they fall on the sensitive surfaces. The significant test of the hypothesis would be the constancy of the angle when computed from experimental data obtained under varying conditions. The method of constructing such an angle is shown in figure 23, in which the opposing lights are assumed to be of a two-to-one ratio of intensity. The line AB is drawn perpendicular to the direction of the rays of light. On the line AB, construct angle BOC equal to the actual average angular deflection of the larvae at a two-to-one ratio of lights (p. 245). The problem now re- solves itself into the construction of an angle about OC as a bisector, which shall be of such a magnitude that equal distances on its opposite sides shall have projections on the line AB of the ratio of two to one. Construction: From a point D on the line OC draw Dh perpen- dicular to AB. Lay off on AB distances hx and hy, such that hy =2hxz. From x and y erect lines perpendicular to AB, they will intersect OC at f and e respectively. Bisect the line éf, and at its middle point, g, construct a line kl perpendicular to OC. From the point of intersection of kl and yy’ (M), draw a — line to D. From the intersection of kl and xv’ (N), draw a line to D. The angle MDN is the desired angle Proof: eg =gf (construction) Angle egM = angle fgN (construction) Angle Meg = angle Nfg (alternate int. angles of parallel lines, yy’ and xx’ being parallel by construction) ~* Therefore triangle Meg = triangle. Ngf (side and two adjacent angles being equal) Ng = gM (similar sides of equal triangles) gD = gD (identical). Therefore triangle NgD = triangle Mgd (rt. triangles, altitude and base equal) Therefore angle gDM = angle gDN and side DM = side DN. QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 267 Now by construction hz is the projection of DN on AB and /y the projection of MD on AB, and by construction hy = 2hx This fulfils all the conditions of construction _ The equal lines MD and DN represent equal bilateral sensitive areas inclined to each other at such an angle, MDN, that the surface represented by MD intercepts an area of light twice as great as the surface represented by DN, its projection on the perpendicular to the ight rays being twice as great (hy = 2hz). But the light falling on DN is of twice the intensity of the light falling on DM, so that the total amount of light received by each of the equal areas is the same. By this method of construction, the average angle of sensitive- ness was computed for four intensity differences, using as a basis the angular deflection of the larvae asdetermined by experiment. The angles are shown in figure 24. The magnitude of the angles is almost identical in all four cases. This angle, I would emphasize again, probably does not represent the actual position of the sen- sitive surfaces in the larva.? There are too many modifying factors intervening between the direction of the rays of light in the field and the angle of incidence of the light on the sensitive areas to permit of basing any conclusions directly on the magnitude of the computed angle, for the structural peculiarities of the dioptric apparatus, or of the tissue overlying the sensitive area would change its value to a large extent. However, the computed angle would be constant in a given animal, and the angle of the sensi- tive surfaces would be equal to K times the computed angle where K may be defined as the ‘structural constant’ of the animal. An interesting problem in connection with theories of orienta- tion is whether light operates as ‘‘a constant directive stimulus’”’ or whether changes in the intensity are the main cause of stimu- lation. Without entering into a detailed discussion, some ex- perimental evidence bearing on this question may be presented. 9 The organs concerned with light reception in the blowfly larva have not as yet been identified. Pouchet (’72) concluded that the two pairs of cones on the maxillal segment were not the light sensitive organs, and suggests that the imag- inal discs of the adult eyes may function in the larva. Certain unpublished ex- periments of my own confirm his exclusion of the anterior cones. I have not been able, however, to obtain any positive evidence as to the organs concerned. PATTEN BRADLEY M. 268 FY HE Hata ey seuss: mee Rien ached aa erred Pepad enw use eens Ur) Pia ot oe oe oF Buu \ seas AS a H+ | fea 7 ae Ey a Ba Hatt ; feet ie ed cbt isd Fe che ay ae pene oe a rH [ee - agent ae ot Horr (J itt} ea tenen HE ad a ~~ 24 @ By nal A nee: f actass HH Ht rtd eth eee NUADeaeteaaa stay rT Rete te fy Bi] fees peek rH hip a aa Bentee ies ae mates 1 fF se Hees hes HE i Sibetbei tees ath Hint HEH ate 4 a a its ine Bina eooet ces eed rer bpp beta a, for b, for b ; ive surfaces = 82 degrees 83 degrees tructed after the method shown in figure 23 for 50 per cent d constructed angle of sensit constructed angle ’ iagrams cons fference i ig. 24 D per cent d 25 per cent difference F 163 per cent differ- fference, constructed for 333 b] c ’ ? i d, 83 degrees; angle constructed b] ence 82 degrees. angle QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 269: Recently Ewald (13) has described experiments which prove that the Bunsen-Roscoe law holds for the response to light in the eye of Daphina, as Loeb has maintained it should hold in animal light reactions (Ewald 713, p. 236): This law states that in a light reaction the effect is proportional to the simple product of intensity and time. It was first proved to be true for the formation of hydrochloric acid from chlorine and hydrogen and for the blackening of silver chloride under the influence of light. Later it was found to apply to the phototropic curvature (Fréschel, Blaauw) of plants, as well as the human eye, though within rather nar- row limits (Bloch, Charpentier). If the law holds for the light reactions of photosensitive ani- mals, intensity must of course operate as a constant stimulus. Ewald found that the eye of Daphnia assumes ‘‘a definite normal position with regard to light’’ and that if while the animal remains fixed, the eye is subjected to lights from two sources, it takes up a definite axial position depending on the relative in- tensity of the lights. “In order to test the energy law, it is necessary to combine different light intensities with different times of exposure. If the product of time and intensity . . is the same, the eye will always give the same reaction” (p. 236). This was proven experimentally by observing the position of the eye when subjected to one constant light and one light that could be varied at will. The variable light could be taken either from a constant low intensity source or through the apertures of a rotat- ing sector wheel. An instantaneous shift from a slow, steady light to an intense intermittent light, delivering the same amount of light energy per second, caused no change in the position of the eye. If a sector wheel is used, giving too long or too short exposures to equalize the light, a change in the axial position of the eye appeared when the difference was greater than 10 per cent.!° “These observations prove that for the eye movements of Daphnia the energy law holds within the limits of accuracy characteristic of the reaction” (Ewald ’13, p. 237). 10 In these experiments, the speed of the sector wheel was about 30 revolutions per second. With the reduction of the speed to 10 revolutions per second, a re- action was in some cases obtained when the change was made from steady to intermittent light. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 17, NO. 2 270 BRADLEY M. PATTEN Before the appearance of this paper, I had made a somewhat similar experiment on the blowfly larva, with the same results that Ewald obtained for Daphnia. Using the apparatus described on page 220, one of the beams of light was cut down by a diaphragm and the other by an episcotister, so that the light coming from one side was a steady beam of low intensity, and that from the oppo- site side an intermittent beam in which bright flashes alternated with darkness. The apertures in the sector wheel were adjusted so that the amount of light from each source was equal for a unit time. It has already been established that when the larvae are subjected to equal steady beams of light from opposite directions, the aggregate response is almost precisely at right angles to the line connecting the sources of light. The average angular de- flection of 200 trails at equality (p. 240) was only 0.09°, when the degrees represented a distance of but 1.5 mm. If the Bunsen- Roscoe law holds for the phototactic response of the larvae, they should orient perpendicularly to the rays of light when subjected to the action of steady and intermittent lights of equal energy per second. The experimental results based on 136 trails made under these conditions show an average angular deflection of but 0.07° from the perpendicular.!! These results seem to show that in the blowfly larva the phototactic reaction follows the Bunsen-Roscoe energy law. Mast (’11, p. 234) says: There is no conclusive evidence, except perhaps in animals with image forming eyes, showing that light acts continuously as a directive stimulus, that symmetrically located sides are continuously stimulated. (p-235). Light no doubt acts on o-ganisms without a change of intensity much as constant temperature does, making them more or less active and inducing changes in the sense of orientation; but there is no conclusive evidence showing that light acting thus ever "functions in the process of orientation. ' Abrupt changes of intensity no doubt cause stimulation and are very effective in orientation under certain conditions” but they are not, as Mast maintains, the only ways in which light Tn these experiments the number of light impulses was 115 per second. 12 See orientation of ‘blowfly larva to single light, pp. 252 to 272. QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 271 operates to produce orientation, for our experiments on the blow- fly, to test the Bunsen-Roscoe law, clearly show that light of a constant intensity both stimulates and plays an important part in orientation. Holmes and McGraw (13) in experiments on insects have come to the same conclusion in regard to the effect of constant intensity (713, p. 373): It is not possible we believe to construe phototaxis entire'y in terms of differential sensibility. Responses to the shock of transition, whether in the direction of an increase or a decrease of stimulation, may play a part in the orientation of many forms but the continuous stimulating influence of light appears to be in several cases at least the factor of major importance. Bancroft’s (13) work, in which he showed not only that there was a distinct reaction to constant intensity present in Euglena but that it was largely the reaction to constant intensity which determined its orientation, shows the untenability of Mast’s sweeping statement in one of the forms on which Mast himself worked. The facts established concerning orientation in the larva of the blowfly may be summarized as follows: 1. When a larva is subjected to a single light, the changes of position, as it swings its head from side to side in the manner char- acteristic of locomotion and orientation, produce changes in the intensity of the light acting on the sensitive anterior end of the animal, due in a large measure to the shadow cast by its own body. 2. An abrupt change in the intensity of the light acting on the sensitive surfaces produces a reflex toward a “physiologically definite side’’—the side on which the muscles are passively stretched. 3. Repetition of this reflex automatically checks motion toward the light. 4. In orientation to lights from two sources, the side to side swinging of the head does not produce changes in the effective intensity of the light on the anterior end as a whole. 272 BRADLEY M. PATTEN 5. Orientation to light from two sources depends on the relative amount of stimulation received by symmetrically located sensitive areas. This isshown by: perpendicular orientation to equal lights; deflection toward the weaker of two unequal lights; circus move- ments when one side of the sensitive areas is blackened; the presence of a natural asymmetry of response which may be coun- terbalanced by a corresponding inequality in bilateral stimulation. 6. An arrangement of bilateral sensitive areas may be postu- lated whereby stimulation on opposite sides of the animal may be equalized by a change of axial position. This arrangement ac- — cords with the facts so far as they are known. 7. The phototactic response of the blowfly larva depends to a large extent on the stimulating effect of constant intensity. The reaction to light of constant intensity follows the Bunsen- Roscoe law. c. Analysis of factors involved in phototaxis. In the preliminary discussion of the relation of directive to non-directive light re- actions, the interpretation was advanced that phototaxis had been evolved from photokinesis by the development, in connec- tion with the latter, of certain factors which modify the action of the light on the organism, or indirectly distribute its effects. In the following section, an attempt has been made to analyze more closely certain of the factors involved in photokinesis and phototaxis, and to ascertain, as far as possible, their relative effectiveness. | The distribution of the stimulus in the field and on the sensitive surfaces of the animal, to a large extent determines the nature of the response. In a field uniformly illuminated from above there is, so far as we know, but one type of reaction, undirected activity which is maintained till muscular fatigue ensues or ‘ac- climatization’ results. But if the field is not of uniform intensity and there are regions in it where stimulation does not take place, the organisms will sooner or later gather in those regions simply because they are not stimulated enough to move away from them. Under either mode of illumination, the reaction of the animal is of the photokinetic type; the condition which deter- mines whether or not the aggregation takes place, lies not in the QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 273 reacting animal but in the distribution of the stimulus in the field. The term ‘trial and error,’ or better ‘method of trial’ may, I believe, be more appropriately applied to the method of aggregation by the kinetic reaction than to a method of orien- tation. The ‘varied movements’ of the locomotion of forms like Stentor or the blowfly larva take place whether or not the animals are stimulated by light. Though they undoubtedly play a part in orientation under some circumstances, it is by no means certain that they are essential to the attainment of orientation. In cases like that of a blowfly larva subjected to opposed beams of light, they probably play a very insignificant part in orienting the animal. It does not seem logical, therefore, to characterize the orientation of such forms as orientation by the method of trial, for even though there are some aspects of orientation that might be ‘explained’ on this basis, there are others which certainly cannot be so explained. An interpretation of orienta- tion, to be acceptable, must accord not only with the details of a special case, but with all the varying details appearing under all the different manifestations of the phenomenon. In the case of aggregation by the kinetic reaction, light causes the undirected locomotion which carries the animal into various environmental conditions. The movements initiated by light persist until ‘selection’ is accomplished with the cessation of stimulation, in other words, until the animal ‘happens’ to move into a non-photokinetic area. Such a reaction may be termed ‘automatic distribution.’ A significant fact in connection with the operation of a uni- formly distributed stimulus is that animals, such as the blowfly larva, which respond phototactically to horizontal light show a simple kinetic response when subjected to uniform illumination from above. The reason for this is the fact that under such conditions changes of axial position do not produce any differ- ence in the relative amount of stimulation received by the sen- sitive areas. There is no basis for orientation. This establishes, as one of the critical factors in a directive response, a distribution of the stimulus such that a change in axial position-on the part of the animal involves a change in the distribution or intensity 274. BRADLEY M. PATTEN of the stimulus on the sensitive surfaces. For brevity, we may call a stimulus which fulfils these conditions a directive stimulus and one which does not, a non-directive stimulus. Not all animals that are sensitive to light respond by orien- tation when they are subjected to the action of directive light. There must, therefore, be a second critical factor, a factor inherent in the responding organism. In the case of the blowfly larva, it was pointed out that orientation might be explained as a re- sult of differential stimulation of bilateral sensitive surfaces inclined: to each other at an angle, and the proportional trans- mission of that stimulation to the bilateral musculature. The explanation in this form will not apply to the case of an animal, like Stentor, which has but a single sensitive area. The two organisms, however, have in common a factor which I believe is a fundamental one in all directive reactions, a definite response to stimulation which is proportional to the intensity of the stimulus. In the blowfly larva there are bilateral sensitive areas, the stimulation of which produces a reaction bilaterally proportional to the amount of stimulation. In Stentor the spiral path of locomotion makes the single sensitive area perform the functions of two. There is a definite response to stimulation in the form of a swerving toward the aboral pole, and the amount of swerving is proportional to the stimulus. The operation of the response first on one side and then on the other brings about the orientation of the animal. Each animal has a method of react- ing which is dependent on its structural peculiarities, but the animals respond in a definite and consistent way to stimulation by light, and both give a reaction proportional to the intensity of the stimulus. When an animal does not respond to a directive light by orientation but merely by undirected activity, it is because the second critical factor of phototaxis is not present, the animal has no mechanism for producing a definitely directed response proportional to the intensity of the stimulus. We may assume that even in forms which do not respond directly, stimulation is proportional, to the intensity of the stimulus, since the experi- mental evidence clearly indicates that stimulation, in its final QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 275 analysis is a chemical phenomenon and would follow the laws governing chemical reactions. The failure to respond directly is probably due, therefore, either to the presence of receptors of such a nature that changes of axial position do not affect the relative amount of stimulation received on opposite sides of the animal (fig. 21), or to the indefinite transmission of the stimulus from the end organs to the locomotor apparatus. There are then two factors which must be present to produce phototaxis, one of these is resident in the stimulating agent, the other is resident in the organism. A directive stimulant may be ineffectual in producing orientation because of the ab- sence in the reacting organism of a mechanism capable of direc- tive response; similarly an animal with a perfectly developed directive mechanism may fail to orient because of the absence of directive distribution of the stimulating agent. The matter may be summarized thus: Non-directive stimulant + indefinite response = photokinesis Non-directive stimulant + definite response proportional to the intensity of the stimulant = photokinesis Directive stimulant + indefinite response = photokinesis Directive stimulant + definite response pro- portional to stimulus = phototaxis The idea of a constantly acting stimulus which produces a reaction proportional to its intensity was one of the fundamental conceptions on which Loeb based his tropism theory. The fact that the theory as he advanced it postulated bilaterality in the responding organism has led to its abandonment by many authors who have studied the asymmetrically sensitive Protozoa. There is no doubt that in its original form the tropism theory does not apply to the orientation of asymmetrical organisms. The pro- portionality of the reaction to the intensity of the stimulant, within the range of normal physiological response, is however in my opinion the essential basis of any reaction involving definite orientation. Its method of expression varies in accordance with the structural peculiarities of the reacting animal, but the under- lying phenomena are nevertheless in all cases essentially the same. If we include under phototaxis any reaction which involves a 276 BRADLEY M. PATTEN definite axial orientation with reference to light, the tropism may be regarded as a special case of phototaxis in which the response depends on bilaterally placed sensitive areas. The analysis of the factors involved in orientation as it has been presented here has been stripped of many details and limi- tations because I believe it has more value when stated broadly enough to allow details to be added as our knowledge of the sub- ject increases. I wish to express my indebtedness to Prof. G. H. Parker, under whose supervision this work was done, for his interest and helpful criticism, and to Dr. E. L. Mark for the facilities of the Zoological Laboratory and for the many courtesies shown me during my work there. I also wish to express the deepest grati- tude to my father, who first aroused my interest in biology and whose criticism and encouragement have been a constant help. SUMMARY AND CONCLUSIONS 1. The wide range of individual variability in the response of the blowfly larva to light renders a study based on untested ani- mals of little value as a basis for comparing the effects of different experimental conditions. 2. To obtain reliable data for a comparative study of light re- actions, it is necessary to establish a ‘standardization test’ and to make use of those individuals only, that show a uniform degree of sensitiveness. . 3. To measure accurately the reaction to light in terms of definite physical units, the larvae have been subjected to the simultaneous action of opposed horizontal beams of light of known intensity, and the response measured, in degrees, on the resulting angular deflection of the trail. 4. When the opposing lights were of equal intensity, the aver- age trail of the standardized larvae was within 0.09° of the per- pendicular to the line connecting the sources of light. 5. When the opposing lights were unequal, the ‘average trail’ showed a deflection toward the weaker light. The amount of the QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 277 deflection was definite and constant, within the limits of experi- mental error, for a given intensity difference between the lights. 6. Using the average deflections obtained under each one of a graded series of intensity differences, a curve of the response to differential bilateral stimulations may be constructed. The deflection increases regularly with the increase of intensity differences. 7. The curve of response may be applied as a means of measure- ment to other experimental data obtained under similar con- ditions. 8. In orientation to a single horizontal light: a. The changes of position due to the side to side swinging of the head characteristic of locomotion produce changes in the intensity of the light acting on the sensitive anterior end of the larva, due in a large measure to the shadow cast by the animal’s own body. b. An abrupt change in the intensity of the light acting on the sensitive surfaces produces a reflex toward a ‘physiologically ‘definite side,’ the side on which the muscles are passively stretched. ; c. Repetition of this reflex automatically checks motion toward the light. 9. In orientation to horizontal beams of light from two sources: a. The side-to-side swinging of the head does not produce changes in the effective intensity of the light on the anterior end as a whole. b. The attainment of orientation depends on the relative amount of stimulation received by symmetrically located sensory areas. | c. An arrangement of bilateral sensitive areas may be postu- lated whereby bilateral stimulation may be equalized by a change of axial position. This arrangement accords with the facts, so far as they are known. 10. The phototactic response of the blowfly larva depends, to a large extent, on the stimulating effect of constant light in- 278 BRADLEY M. PATTEN tensity. The reaction to light of constant mtensity follows the Bunsen-Roscoe Law. 11. The evolution of phototaxis is the result of the develop- ment, in connection with photokinesis, of certain factors which modify the action of light on the organism, or indirectly dis- tribute its effects. 12. The critical factors of phototaxis are: a. A distribution of the stimulant in the field such that a change in axial position on the-part of the animal involves a change in the distribution, or intensity, of the stimulant acting on the animal or on its sensitive surfaces. b. The presence, within the organism of a mechanism adapted to the reception of differential stimulation and a transmitting and motor apparatus that produces definite locomotor movements proportional to the intensity of the stimulation. 13. The ‘response factor’ may be present in the form of a bilat- eral mechanism, or in the form of a unilateral mechanism that reacts to both sides of the environment because of a rotational method of locomotion. 14. If we include under phototaxis any reaction which involves a definite axial orientation with reference to light, the tropism may be regarded as a special form of phototaxis, in which the response depends on the bilateral structure of the mechanism of response. QUANTITATIVE DETERMINATION OF LIGHT REACTIONS 279 BIBLIOGRAPHY Apams, G. P: 1903 On the negative and positive phototropism of the earth- worm Allolobophora foetida (Sav.) as determined by light of different intensities. Amer. Jour. Physiol., vol. 9, pp. 26-34. BancrorT, F. W. 1913 Heliotropism, differential sensibility, and galvano- tropism in Euglena. Jour. Exp. Zo35l., vol. 15, pp. 383-428. CanpDo.L_E, A. P. pE 1832 Physiologie végétal. Paris, 8vo,3 tomes, 1579 +xxx11 pages. Davenport, C. B. 1897 Experimental morphology. New York, 8vo, Part I. 280-++XIv pages, 74 text figures. ; Day, E. C. 1911 The effect of colored light on pigment migration in the eye of the crayfish. Bull. Mus. Comp. Zodél., Harvard University, vol. 53, pp. 305-348, pls. 1-5. Darwin, C., and Darwin, Francis. 1880 The power of movement in plants. London, 8vo, 592+ x pages. ENGELMANN, T. W. 1882 a Uber Sauerstoffausscheidung von Pflanzenzellen im Mikrospektrum. Arch. f. d. ges. Physiol., Bd. 27, pp. 485-490. 1882b Uber Licht-und Farbenperception niederster Organismen. Arch. f. d. ges. Physiol., Bd. 29, pp. 387-400. 1883 Bacterium photometricium. Arch. f. d. ges. Physiol., Bd. 30, pp. 95-124. Ewaup, W. F. 1913 The applicability of the photochemical energy-law to light reactions in animals. Science, New Ser., vol. 38, pp. 236-237. Gross, A. O. 1913 The reactions of arthropods to monochromatic lights of equal intensities. Jour. Exp. Zodl., vol. 14, pp. 467-514. Herms, W. B. 1907 An ecological and experimental study of sarcophagidae with relation to lake beach debris. Jour. Exp. Zoél., vol. 4, pp. 45-83. 1911 The photic reactions of sarcophagid flies, especially Lucilia caesar Linn. and Calliphora vomitoria Linn. Jour. Exp. Zodl., vol. 10, pp. 167-226. Hotmes, 8. J. 1905 The selection of random movements as a factor in photo- taxis. Jour. Comp. Neur., vol. 15, pp. 98-112. Homes, 8. J., and McGraw, K. W. 1913 Some experiments on the method of orientation to light. Jour. Animal Behavior, vol. 3, pp. 367-373. Hott, E. B., and Les, F. 8S. 1901 The theory of phototactic response. Amer. Jour. Physiol., vol. 4, pp. 460-481. Howe ti, W. H. 1912 Text-book of physiology. Philadelphia, fourth edition, 8vo, 1018 pages. Hype, K.P. 1906 Talbot’s law as applied to the rotating sectored disk. Bull. Bureau Standards, Washington, vol. 2, no. 1, 32 pages. Jennines, H. S. 1904 Contributions to the study of the behavior of lower organisms. Carnegie Inst. Washington, Pub. no. 16, 256 pages. 1906 Behavior of the lower organisms. New York, 8vo, 366+xIv pages. Loss, J. 1888 Die Orientierung der Thiere gegen das Licht. (Thierischer Helio- tropismus.) Sb. d. phys. med. Ges., Wiirzburg, 1888-1891, pp. 1-5. 280 BRADLEY M. PATTEN Loss, J. 1890 Der Heliotropismus der Thiere und seine Ubereinstimmung mit dem Heliotropismus der Pflanzen. Wiirzburg, 8vo, 118 pages. 1893 Uber kiinstliche Umwandlung positiv heliotropischer Thiere in negativ heliotropische und umgekehrt. Arch. f. d. ges. Physiol., Bd. 54, pp. 81-107. 1904 The control of heliotropic reactions in fresh water crustaceans by chemicals, especially CO,. Univ. Calif. Pub., Physiol., vol. 2, pp. 1-3. 1905 Studies in general physiology. Chicago, 8vo, Part I. 423+ XIII pages. 1906 a The dynamics of living matter. New York, 8vo, 233+x1 pages. 1906 b Uber die Erregung von positivem Heliotropismus durch Saure, insbesondere Kohlensiure, und von negativem Heliotropismus durch ultraviolette Strahlen. Arch. f. d. ges. Physiol., Bd. 115, pp. 564-581. Mast, S. O. 1911 Light and the behavior of organisms. New York, 8vo, 410+xI pages. Moorg, A. R. 19138 The negative phototropism of Diaptomus through the agency of eaffein, strychnin and atropin. Science, n. ser., vol. 38, p. 131-133. Parker, G. H., and Patren, B. M. 1912 The physiological effect of intermit- tent and of continuous lights of equal intensities. Amer. Jour. Physiol., vol. 31, pp. 22-29. PoucueT, G. 1872 Del’influence de la lumiére sur les larves de diptéres priviées _ d’organes extérieurs dela vision. Rev. et Mag. de Zool., sér., 2 tom. 23, pp. 110-117; 129-138; 183-186; 225-231; 261-264; 312-316. Sacus, J. von 1887 Physiology of plants. Translated by H. M. Ward. Oxford, Clarendon Press, 8vo, 836+xIv pages. UrxktLu, J. von 1909 Umwelt und Innenwelt der Tiere. Berlin, 8vo, 259 pages.. ; Verworn, M. 1899 General physiology. Translated by F. S. Lee. New York, 8vo, 615+XvI pages. Water, H. E. 1907 The reactions of planarians to light. Jour. Exp. Zodl., vol. 5, pp. 35-162. THE BEHAVIOR OF THE EPIDERMIS OF AMPHIBIANS WHEN CULTIVATED OUTSIDE THE BODY S. J. HOLMES From the Department of Zoélogy, University of California SEVEN FIGURES (ONE PLATE) In a previous paper! I have described the formation of strands and sheets ectodermic epithelium from pieces of amphibian larvae which were cultivated outside the body in lymph or plasma. It was shown that these extensions of epithelium were due to an amoeboid activity on the part of the cells at the edge of the extending mass of tissue. This year when opportunity for working on amphibian embryos and larvae was again pre- sented, further observations and experiments on the behavior of the epithelial cells led to the discovery of several additional facts, as well as to the confirmation of the conclusions already drawn. Where not otherwise stated the observations here described were made on the larvae of Diemyctylus torosus. The behavior of the epidermis from tadpoles of Rana and Hyla proved to be essentially the same as in the species mentioned. Studies were also made on the epithelial outgrowths from tissues of the adult frog. : Pieces of embryos were cultivated as before in lymph or plasma from the adult. It is very difficult to employ plasma on account of the rapid coagulation of the blood. Lymph has the practical drawback that in most amphibians a very small quantity is available while in the frog it varies so greatly that no uniform results could be obtained with it. Lymph tends to become too watery for use a short time after the frog is dead, and only a few drops of a usable quality can be obtained from any one animal. Epithelial cells do not wander out to any considerable ‘Holmes, S. J. Behavior of ectodermic epithelium of tadpoles when culti- vated in plasma. Univ. of Calif. Pubs. in Zool., vol. 11, 155, 1913. 281 282 Ss. J. HOLMES extent unless they are in a medium of more or less firm consist- ency. When the lymph does not form a coagulation about the tissue no outgrowth occurs, although the cells may remain alive and the cilia upon them keep beating for weeks. In Ringer’s solution epithelium from both embryos and adults will remain alive for a long time, but no outgrowth as a rule takes place. As so many cultures put up in frog’s lymph proved un- successful the attempt was made to find a substitute for coagu- lum of blood and lymph which would elicit the same sort of thigmotactic reaction from the epithelial cells. After several experiments which proved to be failures, it was found that a mixture of Griibler’s nutrient gelatin and blood serum afforded a culture medium that was very suitable for the purpose in hand. The proportions of the substances finally decided upon were about equal parts of serum and a two per cent solution of the gelatin. The gelatin solution was made and sterilized by boil- ing. In obtaining the serum, blood was drawn from the heart into a small glass tube and the clot removed. Then the gelatin warmed sufficiently to become fluid was mixed with an equal quantity of serum and centrifuged to remove the blood cor- puscles, and the clear liquid was then transferred to another receptacle. The mixture thus formed becomes quite fluid when but slightly warmed and remains fluid for an hour or more at ordinary temperatures. One can work with it therefore with- out undue haste, and should the supply become solidified in the course of making preparations of the tissues it can be made fluid again by a slight degree of warmth. The same fluid may be preserved for several days if kept free from infection. I have found tissue to thrive in it after it had been prepared for several days as well as in mixtures freshly made. The ease with which this medium can be made and used greatly facilitates making preparation of tissues, and the results obtained by its use are also more uniform than those obtained by the use of plasma or lymph. The ectoderm from pieces of young tadpoles grows out in this medium with great rapidity. In twenty-four hours the new growth may exceed the area of the introduced tissue and in EPIDERMIS WHEN CULTIVATED OUTSIDE BODY 283 forty-eight hours it may be five or six times as great. Usually the ectoderm extends in the form of broad sheets, but some- times narrow strands are seen which often branch, and occa- sionally some of the branches may meet and fuse with those of other strands. In many cases most of the epithelium had migrated away from the implanted piece, but the latter was in all cases covered by a thin layer of this tissue. It is common for outgrowths to appear both in contact with the cover slip - and with the lower surface of the hanging drop, and they are about as frequent in the one situation as in the other (figs. 1 and 2). During the period of active extension the advancing epi- thelium is always furnished with an amoeboid border of very clear protoplasm as has been described by Harrison.? This border varies much in width and is often so exceedingly thin and transparent that it is very difficult to follow its course. The very fine processes that are sent forth are mostly in close contact with the substratum, and a number of observations make it evident that they possess considerable adhesiveness. The con- clusion reached in my previous paper that the epithelial exten- sions are due to the amoeboid activity of their hyaline margin is confirmed by the discovery of several additional facts. It was found that the epithelial membranes possess a re- markable degree of contractility. With the application of a stimulus a large epithelial extension may shrivel up to about one-tenth its original area. Often a very broad extension sev- eral times the area of the tissue from which it came may con- tract to a very narrow fringe, giving one the impression that to a large extent it went back to its original situation. The contraction starting at any one point may be seen to spread to surrounding areas until finally the whole mass may be involved. At other times only a small part of the sheet of ectoderm may draw back. In either case the margin of the retracted ectoderm becomes much thicker and rounded, the amoeboid processes disappearing. If the amoeboid margin is watched carefully in an epithelial extension that is just beginning to contract the ? Harrison, R. G. The outgrowth of nerve fiber as a mode of protoplasmic movement. Jour. Exp. Zool., vol. 9, 787-848, 1910. 284 S. J. HOLMES . amoeboid processes may be seen to give way suddenly as if they could no longer resist the tension of the cells behind. The free margin then becomes thickened and rounded, and the cells take on a quite different form. When one part of the hyaline border gives way the adjacent parts follow; the retraction, however, may soon stop, or it may spread widely according to various cireum- stances. At the beginning of the contraction the pseudopods may be stretched out considerably, as the cells behind tend to draw away, before they become loosened from their attachment. A characteristic of these epithelial outgrowths quite as striking as their extreme contractility is their extraordinary sensitive- ness to slight stimulations. Bringing the slides from a cool place to the stage of the microscope where the temperature is a few degrees higher usually causes a retraction of some part of the margin of an epithelial outgrowth. The process of con- traction was observed a great many times, and in fact it usually happens to a certain extent whenever slides are brought from a cool place for examination. It is probably the transition to a warmer environment rather than the higher temperature per se that causes the contraction, because preparations kept at room temperature form equally great extensions of epithelium, and in many cases the sheets of ectoderm which become contracted when exposed to a higher temperature subsequently extended again while the higher temperature remained unchanged. Con- traction of any part of the sheet of ectoderm may be iniated by placing the point of a warm needle above the region in question. That the cells are responsive to a mechanical stimulus was shown by touching the margin of the extension with a fine glass rod. Local contractions uniformly followed. Whenever a prepa- ration was washed in Ringer’s solution preparatory to furnishing it with a new supply of nutrient gelatine, the epithelial extension shrivelled up to a small fraction of its previous dimensions. By placing a drop of Ringer’s solution upon the preparation the epithelium is caused to contract as soon as the solution diffuses into contact with it. Whether the contraction is due to a slight osmotic effect of the Ringer’s solution, or to the stimulating influences of the salts is uncertain. EPIDERMIS WHEN CULTIVATED OUTSIDE BODY 285 In several cases a part of the epithelial outgrowth of a piece of tissue was cut off and transferred to another drop of the culture medium. The act of cutting causes a violent contraction of the epithelial outgrowths, and the operation usually has to be done quickly to be successful. A broad and very thin sheet of cells may shrivel up, after it is cut off, into a small rounded mass that has very little resemblance to its previous condition. Sub- sequently these isolated pieces may spread out as widely as before. The application of an unfavorable degree of heat causes a contraction of the epithelium quite aside from any stimulating influence of a change of temperature. Often the areas of epi- thelium may be broken up in this way into isolated masses of cells or even into individual cells. The influence of heat on epithelium is very similar therefore to its effect on the blasto- meres of a dividing egg. The cells or cell masses tend to become rounded up and inert. Light has very little effect on the epithe- lial cells. If the heat rays are filtered out, epithelium may be exposed for hours to the most intense light without manifesting any evident reaction. In general one may say that epithelial cells, like so many free organisms, respond to various unfavorable influences by con- traction. When kept for several days in the same culture medium the epithelial cells show a tendency toward rounding up. This often results in the rupture of strands of cells, or the breaking up of sheets of cells into isolated masses. Under favorable conditions epithelial cells rarely isolate themselves from the general mass. It is probably the accumulation of products of excretion that causes the contraction of cell masses which are kept too long in the same medium. The breaking up occurs much more quickly in cultures kept at room temperature than in those kept in a cooler place, owing probably to the more rapid metabolism and the greater accumulation of toxic products. That it is not the injurious influence of room temperature alone that causes the contraction of the cells is shown by the fact that by washing the contracted masses in Ringer’s solution and giving them a fresh supply of the culture medium, the epithelial cells may again form extensions at ordinary room temperature. 286 S. J. HOLMES Cultures may be kept alive for a long time at room temperature, but they require more frequent washing and change of medium. Extensions of cells occur more rapidly at room temperature than when the cultures are kept in an ice box. While it is clear that extensions of ectoderm are mainly due to the outwandering of cells, a certain amount of cell di- vision is also found to occur. Division figures were not infre- quently observed several days after the preparations were made. In two cases a division figure first seen in the prophases was watched continuously through the entire process of mitosis. The individual chromosomes could be distinctly seen and as many as fifteen could be readily counted, although the actual number is greater. The arrangement of the chromosomes in the equatorial plate, their pulling asunder, and the formation of the two daughter nuclei could be easily followed. The chromosomes appeared as V-shaped rods, and during the ana- phase the open end of the V’s were directed towards the middle of the division figure. Cleavage of the cytoplasm was completed soon after the telophase of the nuclear division. The whole process of mitosis was completed in less than three hours. Mitoses become rarer the longer the tissue is kept in a given supply of culture medium. In one preparation from the larva of Diemyctylus mitotic figures were common four days after the tissue was isolated, and some were found on the seventh day, but none later. Changing the tissue to a fresh supply of culture medium, however, may cause cell division to be resumed. One preparation which had been subcultured several times showed numerous division figures fifty days after it had been removed from the body of the animal. Two days after its last transfer into fresh culture medium it showed over twenty-five mitotic figures. One of these was watched through into the telophase when the cell body could be distinctly seen to constrict into two separate cells. I have never seen mitotic figures so abundant as in this piece of epithelium which had been kept for fifty days in an artificial medium. The washing in Ringer’s solution and its transfer to a fresh supply of fluid had apparently given the tissue a new lease of life. EPIDERMIS WHEN CULTIVATED OUTSIDE BODY 287 As stated in my previous paper, extensions of epithelium often give strong evidence of amitotic nuclear division. Further observation not only furnished additional evidence of the same phenomenon, but revealed some of the conditions by which ami- totic division is induced. Newly extended sheets of ectoderm present little or no indication of amitotic division. The nuclei of the cells are round or oval and rarely present any indenta- tion of their outline. With epithelial outgrowths which have been kept for a week or more in the same culture medium indi- cations of amitosis are more frequent. Instances of two, three or four nuclei in a cell are common (fig. 3), and in some cases as many as eight nuclei were seen in a single cell. No clear indications of cell division following the division of the nucleus were observed. In many cases the amount of nuclear material in relation to the cytoplasm was obviously increased to a very considerable extent. In all cases where numerous nuclei were seen in a cell there was a considerable amount of yolk present, although cells with very little yolk remaining in them frequently had two nuclei, but rarely more. While evidence of amitotic nuclear division occurred abundantly in many preparations kept without a change of medium, other preparations made of the same material but changed every few days to a fresh medium showed no indication of amitosis, al- though they were kept much longer than the preparations in which the medium was unchanged. The appearance of indica- tions of amitosis very frequently goes along with signs of dimin- ished activity, such as the rounding up of cells, the disintegration of certain cells in the culture, and the general inactivity of the epithelial tissue. It is a commonly received doctrine that ami- tosis occurs most frequently in cells of declining vitality. Its association in cultures of epithelium with life under unfavorable conditions lends a certain support to this view. Most of the epithelial outgrowths observed showed cells of somewhat different types. The ordinary cells of pavement epithelium (fig. C), as soon as most of the yolk disappears, con- sist mainly of a clear, rather homogeneous protoplasm, more or less granular in the vicinity of the nucleus. A small amount of 288 Ss. J. HOLMES yolk in the form of tiny spherules may persist until a late larval stage, even in the ectoderm of the tail and dorsal side of the body, while it is fairly abundant in the cells of the ventral side of the abdomen. As the cells become older and the yolk is gradually used up they tend to become relatively broad and thin, and to increase in transparency. Certain cells in early larval develop- ment become distinctly alveolar in structure (fig. 5). Such cells occur either singly or in small groups scattered about among the cells with homogeneous protoplasm. The alveoli are of various sizes in the same cell. These cells resemble and prob- ably correspond to the Leydig’s cells which have been described from the epidermis of several amphibians. I have observed the same type of cells scattered about in the epidermis of the young larvae of Diemyctylus in much the same way as they commonly appear in epithelial extensions in vitro. In the young larvae of Triton, Salamandra and Siredon, Maurer’ de- scribes the Leydig’s cells as scattered about singly among other cells of the deeper epidermis. On account of their vacuolated protoplasm these cells are considered as glandular in function, but they take no part in the formation of the cutaneous glands of later development. It is the appearance of the alveolar con- tents of the Leydig’s cells that has given rise to the common opinion that they are mucous glands, but the cells have no direct connection with the exterior, and are usually covered with one or more layers of ordinary epithelium. In the preparations of epithelium in vitro various intergradations were found between ordinary epithelial cells and cells of the vacuolated type. Occasionally also cells with granular contents occur in the epithelial outgrowths (fig. 4). The material forming the granules has a different appearance from the yolk, and probably repre- sents the accumulated product of some sort of secretion. As we have mentioned before, preparations may be kept alive for a considerable period if they are changed to a fresh culture medium. No especial attempt was made to find how long tissues could be maintained alive, but some of my preparations of epi- thelium were living over three months after implantation. One 3 Maurer, F. Die Epidermis und ihre Abkémmlinge. EPIDERMIS WHEN CULTIVATED OUTSIDE BODY 289 cross section taken from the tail of a young Diemyctylus larva before it hatched from the jelly was put up in the middle of January. In a few days the epithelium showed extensive out- growths along the cover slip and lower surface of the hanging drop. On January 27 after it had begun to show signs of de- terioration it was washed and transferred to fresh gelatin and serum. Growth was resumed, and the cells showed a healthy appearance. The piece was transferred to fresh medium on February 2, and again on February 18. On February 24 it was divided into two parts. Both pieces, although their epithelial extensions became much shrivelled up after the transfer into fresh culture medium, soon put out extensive sheets of epithelium. Both pieces were transferred on March 3, when the same con- traction and subsequent extension were again observed. One piece was transferred once more on March 10, and lived about a month longer. It is represented in figure 10. The second piece which showed signs of diminished vitality in the rounding up and separation of many cells and cell masses was re-transferred on April 1, after which it became active and sent out extensive sheets of epithelial cells. The histories of many other preparations were essentially the same as the foregoing. In all cases the tissue was washed in Ringer’s solution for a half hour or more, and the old gelatin mass around it removed before being put into a new supply of nutrient medium. , ne OQ oD il pa son A coe oe oon ih oon ian onl Soe ih oes Mh nn oe =n et onl = = : a i NAA eS Senn NNN So os oe ech ollie ts Ilo Wee] te) N ies es) | (ete al VS | | | | ON oO | Ss OO OO N OD sH sH OD OD | OD OS OD OD ess rmsN es (o) | | { | Se) NO ay | St St SL N OI OD YD ms NAN | © OO rt N OD A) | | ON AN ay MN OD oD OD On OD OD OD Ga ON ON YD Ye i) Gel t=) (3) (a oD roy | sai | a pone oP | |- = oD NN = oD 6 MI OD | sH cH 10 1D | SH OD NI CO | = ANAS] = ——- —_ | —_— —$$$_$___— EX Kerker © rea 1 19 29D i t- ~~ sH ost co 19 (S) “No = Sen ee ieee hol Se) _ me & = =< Se here: Fae | ° QL A GA. & ANIA & ~ wiSie aa NAAN] + OO OD OO IN AN Pe ANNAN a [resiad ; | | ms ex] Ge) (ok NI || Ge) Sal Soil Sanh | te Sy al | a SSI alah) oD ON OD 4 | =) aN Ao} SN | HH oo | No = Dlesb ao N = — a | LS } | & Were } oOo loo Sp St ay PS rest a Onr es tS coornrr N ¢ ae | oo — fe) = | SS Ww = = = | fe — i ° ° ° awe e S&S o O oD BAAN AN a NAAN | OD aA a a | | | | | | | 4 mA om | 2 SN OH) ON OW |) HNO SH ra ee re ys io BEER) | et aea This should not be taken to mean that any particular gene mutates as often as any other; it is defiinitely known that, both in Drosophila and in other forms (corn, Marabilis, ete.), some genes are more likely to mutate than others. 336 HERMANN J. MULLER LITERATURE CITED Bripaes, C. B. 1913 Non-disjunction of the sex-chromosomes of Drosophila. Jour. Exp. Zo6l., vol. 15, no. 4. 1914 Proof through non-disjunction, ete. Science, vol. 40, no. 1020. Bripcss, C. B., and Sturtevant, A.H. 1914 A new gene in the second chromo- some, etc. Biol. Bull., vol. 26. Merz, C.W. 1914 Chromosome Studies in the Diptera. I. Jour. Exp. Zeodél., xolle aly/, saVo}s It. Morean, T. H. 191la Random segregation versus coupling in Mendelian inheritance. Science, vol. 34, no. 873. 1911b An attempt to analyze the constitution of the chromosomes on the basis of sex-linked inheritance in Drosophila. Jour. Exp. Zoél., vol. 11, no. 4. 1914a No crossing-over in the male of Drosophila, etc. Biol. Bull., vol. 26. 1914b Mosaics and gynandromorphs in Drosophila. Proc. Soc. Exp. Biol. and Med., vol. II. Morean, T. H., and Lyncu, C.J. 1912 The linkage of two factors in Drosophila that are not sex-linked. Biol. Bull., vol. 23, no. 3. Srevens,N.M. 1908 Astudy of the germ cells of certain Diptera, with reference to the heterochromosomes and the phenomena of synapsis. Jour. Exp. Zo6l., vol. 5, no. 3. Srurtevant, A. H. 19138a The linear arrangement of six sex-linked factors in Drosophila, as shown by their mode of association. Jour. Exp. Zodl., vol. 14, no. 1. 1913b especially with the fertility of two mutations and two wild stocks that originally came from two different sources. It will facilitate treatment of this subject to give a brief résumé of the history of these stocks. FERTILITY AND STERILITY IN DROSOPHILA 345 HISTORY OF THE STOCKS USED IN CROSSING 1. The Woods Hole stock This is a wild stock that originally came from Woods Hole, Massachusetts. How long it had been in captivity I do not know, but when I took charge of the stock it showed a fertility of 77.3 per cent. The stock was tested at various times throughout the year 1912-1913. The behavior of the fertility of this stock is given in table 1. The stock was cultured en masse. TABLE 1 Showing the fertility of the Woods Hole stock roe nese eases | he ee eee September 11 to September —, 1912................| 1402 | 1084 | 77.3 September 16 to October —, 1912.................. | 1622 122 eae O January 20 to February 14, 1913................... He2085 | 1420s anOK esto July tne tsi en Ce eg B81 | 63.4 Smile totaly 28 e1OUsee ge ee et 531 317 59.7 Ancusti7. to Ausust/24, 1913s. 0). oe) 2c. canes te | 601 | AUP La AUTustleconmeptemben 41 Oloaaas 5 serie: 856 715 | 83.6 October 2) tom November ih Ose sso eao eel 467 341 1320 | | 2. The inbred stock This stock originally came from Falmouth, Massachusetts. During the course of the experiments brothers and sisters have been selected and paired for over 26 generations. Its produc- tivity was at the beginning of the experiments relatively high as shown by the large number of offspring produced. On inbreed- ing the stock gradually dropped in productivity as shown in table 2. TABLE 2 Showing the productivity of the inbred stock in successive generations of close inbreeding GENERATION | 1 2 12 | 13 14 3 4 sill 7 slo lil 1 No. offspring | | | | produced.... | 368 | 209 190 184 65| 119 (este aa) _ 159 346 ROSCOE R. HYDE This evidence taken by itself does not prove conclusively that the loss in productivity was due to a decrease in the fertilizing power of the gametes. That this is a very probable interpre- tation however is shown by the fact that in the Fi, generation when the productivity dropped to 159 per pair their fertility was only 32 per cent. And yet the flies of the Fi, generation were producing twice as many fertile sperm and twice as many fertile eggs as shown by outcrossing into the truncate stock (Part I, diagram E). 3. I, and Is stocks In the fifth generation of the inbred stock, when they were producing 184 offspring to the fertile pair, I set aside two dif- ferent bottles besides a number of pairs that were kept to con- tinue the inbred stock. All of these flies had descended from the same grandparents, a single pair of the third generation. In one bottle there were three females and four males. This I shall designate as I,. In the other bottle there were ten males and twenty-two females. This I shall refer to as J];. These flies were set apart April 2, 1912, and received no attention except that the flies were transferred to clean bottles about every three weeks and fresh food added from time to time. It is important to bear in mind that we have three stocks descended from the same germ- plasm. 4. The white-eyed stock This stock arose as a mutation from the inbred stock early in the history of the strain. This stock received the same treatment as I, and Is. 5. The pink-eyed stock This is an eye mutation that arose in one of Morgan’s cultures.’ It had been bred for some three years. I received my stock from Mr. Liff, a graduate student in the department, who has made a study of their productivity. This stock was bred in mass culture for about four months when the present experiments were carried out. 2 Science, vol. 33. FERTILITY AND STERILITY IN DROSOPHILA 347 We have then six different strains, two from different localities and some that had the same germ plasm. I now propose to exam- ine the fertility of the different stocks and their behavior on cross- ing. The combinations were made up as shown in diagram A. METHODS In order to get an exact measure of fertility the eggs were iso- lated as described in Part II. From the stock bottle in each case 15 virgin females and 30 young males were selected. The flies that hatched from June 19 to June 26 were separated every twelve hours from the stock bottles. There can be no question as to the virginity of the females. The males and females were kept in separate bottles and the different combinations made up on June 26. An epidemic of mating took place in all the bottles a short time after the flies came out from under the influence of the ether.* I commenced to isolate the eggs two days later. This process was carried out in the same way as in previous experiments with one exception. ‘The weather was very warm and the larvae emerged from the egg in less than twenty-four hours. In order to exclude this source of error I added another bit of food after isolating the eggs. This served as food for the parents. After six or eight hours this food was removed and a new bit added from which the eggs were later isolated. Tables 3a, 3b, and 3c give the number of eggs isolated each day and the corresponding number that hatched. DISCUSSION “A study of diagram A brings out the essential relations that concern the questions propounded at the beginning of this paper. 1. What is the effect on fertility when germ plasm originally from the same source separately inbred for several generations is recom- bined by crossing? It will be recalled that I, I, and I; represent in this case the stocks under study. It is to be noted that the three stocks although originally brothers and sisters and descended from 3 Mr. A. H. Sturtevant has shown that the mating habit in this species is largely associated with the sense of smell and this fact probably accounts for the phe- nomenon here observed. 348 ROSCOE R. HYDE Fink & 41.6 X Fnk 9 Diagram A Based on the foregoing data, showing different combinations and the fertility in each case. the same grandparents, show different degrees of fertility. The inbred stock in which brothers and sisters had been continually selected had lost most in fertility. In the crosses it will be observed that there is no sudden rise in fertility as one might expect but, on the other hand, a peculiar relation exists in that the stock in each case having the highest fertility is able-on crossing to bring the fertility of the lower stock up to its level and this is true whether through the male or through the female. 349 FERTILITY AND STERILITY IN DROSOPHILA C19 LF | GG | 19 €°6¢ 09 02 ol ele).ers. 6 le) ls) 0 (eels COOO h OO att hice 4ugo 19g 6&8 | 22o | “99F | S86] zeP]| gszs OOF) 699") 917 |) “102 | EY) nO 72 Fete O00 ame ses vey ee ress = eet, Ll OL iit 0% 8 Cl 0 0 | 6 | 0G II | 0z 0 | 0 Mo Melamelrutles (else) stein alisicspanterersucmelate ts CT v LT 96 OF 0€ GE 0 OP S|" Se && 96 | O0& Oar ie Oe vi cepa et nae eae ia! ) éI Iv G9 vE g¢ 9T 06 GG OF 9T GG ) in ene Sa ee ge ie eae ee rea xa Le 8é GL Ae a4 61 IG 09 GL 66 0S 8 Oe. ee eS es a 9F 09 LE 09 Sc | 09 ce | Os 0 0 to OOF SL gi OO a eo ae Cahiers IT GI GG OF GL 96 | OS Ve | GE 96 OS OG ae G2 ele Pol ttn Serge Np eae OT Dee Oe le Bhs \ss¥¥ 96 GP GG 0€ al ¥G Io | 8 g COs Renn pee Bone Boe eae 6 OI cr | OF ve | 02 9& HE LT vG 8G &G | && 6 OCs lio fovea Ree aa 8 JN alice) Ste oo See Sey ng) L9 66 Olea Sia 7 IT OBA esi Baal hea aa cL PE 0g Ig GL L 96 FI O¢ oD 9€ 9& | ¥9 Or CV Aa Circa es aay aye eek a 9 && Oey 0Gie | eS = (eae Gg Iv €2 | 68 | 2¢ LG og 9 Cola! |e Min copes tote oe eS i] &1 96 SGe 008 5h OG OOP eels HCG (596 Se) te Gal O& L of eb) Sema ge ec he aR a ey P 6 GG | GG 6h | O 0 v1 && 0G && ST OF a ose CR Lo Se ee a & 9 0G ST OS ied GE Z GG ZT OF GI GG v Lea tea epee et Leap ee G I€ 8& 96 09 c% | 8P 66 OF 96 OF 61 0& é1 USSSA Rn lee Paks Sa ae ee Aguve Dea meet) lever Tees OOm cee Fee: OZ) 62) cerre-og: | ep gg. (os ce eee a 08 Oa ame ees lair rere, ace. eye. ge | atl te Wo “lop. ape 66 1Ge 1G Pees OV | &Z 09 6& | O8 [EN AS eG See lO OP A ae ae eee eee OUT: synpy Sooq “sympy Soa |suapy SDS SHIM | Ssoq | sinpy| sBsqy | synpy sao synpy | ssaq S161 moe el | Gores) | ll as I eneueey | I * AL a x at (aaeaneo} ee 9 5 F | & j 1 49078 afia-aziym = sy ‘yI0]8 ahia-yuid = gq £y90}8 2107] Sspoom = HM ‘saypwaf 66 “Sajpu OT wou Yyo0js pasqui pry) = *I ‘Sappwaf 7 ‘sap g woLf y90]s pasqui puosas = 7 “pasqui = J faypounsg = J, *wayn) SDM 9]DUL BY} YIN WOLJ 49078 9Y} 13}}2] Puooas ay? ‘uayn) som appuaf ay) YoLyM wos 490) ay} Sjuasatdas UoYDULQULOD Yyova Ut 4397}9) jsuy ay, *suoynurquioa 94} $0 Yoda Worf payaiwYy DY) Laqunu Burpuodsas.soo ay) pun fivp fig Rup pajnjos2 sbBa fo saqunu ay) Bumoys ®€ ATaViL HYDE ROSCOE R. 390 €°29 Z OF gue | 2°89 | _909 [ii 69 £89 Gos 206 | G&8 | S&IT; G28) $08; 972} LS9 | TZy | OFL| ShF| 622) 289) 968); Z9F| 999 18¢ | LT6 ZT 0€ && OF OF) 08 L 0% 6 | 02 0€ OF 66 OF ci | 0& FG | OF bP 0s 66 OF alleles IT ST 1G vE 96 GE GE OF && i CE | GP OF ac OL GL ¥G 6& LT &P 0g 09 IT 0G 96 OF 66 OF €$ | &9 6P OL v9 89 GE Og GG 09 69 | GL 9 08 ys | OOT| 19 OL Ee SIE 0G 09 0& 09 0% 09 ia! Og fo. |. OF GG OF bP 09 bP 09 0€ | 09 G& GL SY GL GG ¢9 LT O¢ 02 |. SE 96 Og vE ¥9 0 0 8& | 92 ete) OF Lv Lg 8 &6 9 GG GG T€ 91 0€ | 8€ T¢ GE se ie he)! 8& SY OG ain ee IZ | 88 8 SY Ge) “cl 1G 0g | 1% GE 06 & 69 | O0r L9 06 08 OTs Sess OSes Saar OF cp | 99 LS 08 | 99 GL && Iv 09 | 08 1 OF Og 6G Tee enee. || 00 OF O¢ L9 8I UES | 8G. 4) OF 66 Og Lv | 08 ST Ss 69 OO. ; Lz Poe ae 09 Lee | 06 2G || 89 eg 69 c% | & 86 eee ne &6 $9 0& 0g ST cp | 2 T€ 96 Og LT GE 61 GG Vales 86 | “ag 0G Ig Og OO. |; &I | 66 ob Th VAL 61 It rea 86 66 i rai Ne |e) Or ST I 61 LG 0g tw | OF Sil eats GG cE 91 0€ &I GG 4 | 06 L 8T 9§ OF 8& 8F eg | 09 8 96 OMe sles GG GP &G SP Go | «GE 61 | 86 te a3) zg 69 cl | G@ LE TZ 86 | 8& TOON Si 96 og 9T 1G PG | OF eg Co jek zp ee | Cer sees) TE | Toe 06. .| Pe. 18h. | 00 Oren ane Ll sins IZ fe | oF | 6s | 02 | Ree loy sets! 11 | Selettsl ee. | Mc) Geo soe |r |) Ge. 9 | ae synpy ssoq | S}[npy| ssoy SuMpy | ssaqV | sy[npy | ssaq spy | ssaq | SyNpy | ssaq ‘simpy | sa3q | S}[Qnpy | ssaq sy(npy ss0q 1X HAN ‘1X HM IX (ndto0 XAT 1X1 HM XI Hat X11 Gowan) 1¢ ATAVL dol FERTILITY AND STERILITY IN DROSOPHILA -9 Dave GIG 9°GP | 9° IP vLg € bl L 68 &'69 €8¢ O16 | GEE) SIO) SEL | 679 | Lie} 60S | 908) Gez| ece| ST9| szP| zee] Fez] 06g 829 626 cT Mealy OL L 0& OLS) (ORs | esr OF él 0% at 0G L GE ial 06 LT 9% 61 LG L GE Sin ch care og 0& OF 9€ OF 1G && 86 0& 6& Og LT SY al ¢9 OL Gra set 0G 66 09 ¥G 0& GG &P It og OF OF Iv 09 61 ug &G SY a4 08 pS g8 09 OL c& 0g 6¢ G9 GE | OL 0& OL €1 G9 1G O¢ ZI 09 Ig OL 9F 09 0¢ 09 OT Oe 0¢ BS.) TS 5 | 09) } Se | 6S | BS | ee ORs Or |) 0c. Se a ey |) 09- | 2 IZ os | &9 &F O¢ IT LG &1 0g L éI ZT ¥G €1 ZI 61 96 & ST Iv vg 1G TS... 98 VE ZL OF ¢ Ol c Z 0 0 ins LE 0 8 99 GL LY 09 | IT | & 6 0g ¥G OF 61 OF &@ OF 09 08 6 1G 92 06 61 | I? 6 1G 0 Or 8 A Ail 0g 9 ST 9T 61 9 GE 1€ SP 8& $8 0g ee) 6 OF G ia! 8G GL €¢ 08 I¢ Lv ZL 0z 8é 8¢ ZL O& G LT I GG rat &P v 0G 86 ag € Or St gg GP c8 ¥G LV &I | 8 GF G L ST 9¢ Or 0c a L IT T€ 96 467 cT GF LI 66 4 LZ P xe L bE 9T 9¢ Or iat 8 GG 1G GE 8€ 19 &1 86 Ol GG or 96 ZI cé ZI 1G ial 0G iat 9€ && 0g £9 08 &@ Ig 8 teh a mil &@ GE i OL IL | dé 6 81 SP 09 96 sé Or ia! 4 06 v 91 6 0€ 0 0 0 0 SI OF PP &¢ 9F 09 fi OL Ee eG 8é OS | 9€ 9 Or 0 0 GG Iv 0& SY synpy ssaq | Synpy| sssq |sinpy} ssaq isynpy | sssq [si(npy | sssq (sinpy) sssq |sznpy| ssaq | S}]Npy | ssoq synpy Sore § MXa MI (een aX Mi Roane) 1Xd aX1 1X At "1X HA £6 1% 2¢ HTAVL 302 ROSCOE R. HYDE The second question under consideration is; What is the effect on fertility when stocks from different sources separately inbred are crossed? The relations expressed by the right hand side of diagram A throw some light upon this question. In two of the combinations high fertility brought the low fertility up to its level while in four of the combinations there is an appreciable rise in fertility beyond that shown by the parents, although by no means as high as one might expect from the history of the crosses made with the truncatesin Part II. But since the rise in fertility occurs here in the combinations expressed by the right side of the diagram and not those on the left it looks as if the rise in fertility in this case is significant. It must be admitted, however, that the rise in. fertility in these crosses is not great enough to base a final conclu- sion upon, in regard to a point as far-reaching as this. It is to be remembered that Woods Hole, Massachusetts, is only four miles distant from Falmouth and it is altogether probable that the two strains had not been separated by many generations when taken into captivity. Consequently, this material would present after all a picture very much like the first case. If the environment can influence the different strains in respect to the ‘factors’ that bring about fertility the influence in this case has been slight, and, after all, from stocks so closely related this is probably what we should expect to find. In any case it is certain that the high-producing stocks can bring the fertility of the low-producing stocks to their level whether descended from the same or from different germ plasms. It might seem from this (although I am far from con- tending at present that such is necessarily the case) that there is a set of factors of some sort for fertility and that when a loss in fertility occurs in a stock (other than that which occurs in the case of mutations) it is the same set of factors that is lost or is changed and this accounts for the fact that there is no rise in fertility on crossing. The stock that has the highest fertility (the largest number of factors) acts as a dominant character and brings the lower set of factors up to its level. In the case of a sudden rise in fertility on crossing such as occurs in the case of some of the muta- tions it is probable that the stocks have lost different factors for FERTILITY AND STERILITY IN DROSOPHILA 353 fertility, and that crossing gives the proper constellation of factors for a marked rise in fertility. It must be frankly admitted that this explanation at best is only tentative and that the door must be left open for further investi- gation. The possibility of transmissible lethal factors is not to be overlooked. The third question under consideration relates to the effect on fertility of crossing certain mutations, namely, pink-eye and white- eye and the effect on fertility when these are crossed into the inbred stock. ‘The triangular portion of diagram A shows that there is a rise in fertility in all cases. In four of the six cases the rise in fertility is very high. This is analogous to what happens in the case.of the truncates when crossed into other stocks. I wish here to add the data from an experiment that has some _ bearing on the foregoing considerations and also some bearing on the question of inbreeding. The object of this experiment is to serve as a check upon the controls used in the foregoing experi- ments, namely, Woods Hole, white-eye, I., I; and the inbred stocks. This experiment was carried out in the same manner as the foregoing. The egg counts began the same day that the former experiment closed, July 15, 1913. The number of eggs isolated, together with the corresponding number that hatched, is given in table 4. I have placed the percentage of fertility of these stocks in dia- gram A and enclosed the numbers in circles. It will be noted that the Woods Hole and white-eyed stocks remain practically the same as in the previous experiment. There is a marked rise in fertility, however, in case of the three stocks originally from the same germ plasm. Later (Aug. 7 to 24, table 1) the Woods Hole stock was tested. Its fertility had risen from 59.7 to 71.7. The Woods Hole stock was again tested August 17 to September 4. This stock now gave a fertility of 83.6. The fact is that all my stocks at this time showed a marked rise in fertility, as is evi- denced by the experiments to be dealt with in the next paper. Even the truncate stock which had been tested many times through the year and had varied from 20 to 26 per cent now gave a fertility 354 ROSCOE R. HYDE of 31 per cent. Table 1 shows that in the case of the Woods Hole stock the fertility gradually fell throughout the year. I fully expected to find that the fertility would be continually reduced on inbreeding, and yet the fertility of the stock rose in August and September until it was actually higher than it was during the September of the previous year by 6.3 per cent. The meaning of this is obscure. It will be recalled that Castle found seasonal fluctuations in productivity in his flies. The period, however, in which low productiveness prevailed in his stocks corresponds in a general way with the period of high fertility in my stocks. In the light of this evidence it would seem that inbreeding as such cannot be the vera causa of the low fertility that usually accompanies the process. It would seem from this evidence and the evidence presented by Moenkhaus that the fertility of a stock could be maintained, and the closest,inbreeding practiced, provided TABLE 4 Showing the fertility of the Woods Hole, white-eyed, inbred, Iz andI3 stocks. 1A = ' Woods Hole stock; 2A = white-eyed stock; 3A = 13 stock; 4A = I, stock; 5A = inbred stock, Fos generation la 2a BA | 4a 5A DATE iz { 1913 Eggs |Adults| Eggs | Adults} Eggs | Adults) Eggs | Adults} Eggs | Adults July 15 95.) 9650. |- 995, A Sis) B7ON | 2582 1 Soke yl eae 16 BOW aes. 4) 5 0se A AT SG Wee, Tote SG ee ae 17 450 30°.) S50 7h Se AGS Sac 1 ee ee con 14 18 AA) Styl" 20 So Bie) 420 60 eat ea ae ales 19 29 14° | .51. | 20" |) 45) B41 50) 26. | S50 a6 20 30 10 | 50 11 | 66 | 40 | 45 | 22 | 30 11 21 31 15 | 31 2 AG HSB it 1.35 16 22 41 24 | 25 3 Sant) -a0nah 3h 14. | 12 6 23 Ob i On! 231 4.) 2519640" )| 22.) 920 6 24 20 | "21 26 Sel alia 136) 40: | 426) a6 9 25 25 12 | 25 Stl ae Go) 22 | VS |) 22a 26 50 | 29°| 25 Bo | ea ier | oe 1) Rg a al 27 25 13) aa Pes ees a) is ames ae ea ae 10 28 15 Sas 5 | 30 18 | 25 12, (25. |. 76 Totals eee 531 | 317 | 498 | 106 | 544 | 412 | 527 | 302 | 467 12 Percents... 59.7 20.9 | 75.7 | 57.3 45.4 FERTILITY AND STERILITY IN DROSOPHILA 515) the proper combinations were made. It seems probable in the case of fertility, as in many other characters, that inbreeding gives a chance for defects to be brought to the surface; and that low fertility is likely to accompany close inbreeding provided it is not guarded by rigorous selection. When a stock has reached a low degree of fertility it seems strange that that same stock should be able to rise again in fertility. Yet thisis exactly what may happen. Take for example the truncate fly which has been selected for 75 generations and has its fertility reduced to about 20 per cent; and yet that fly can throw a form, the long wings, the fertility of which is more than twice as great as its truncate brothers and sisters. The fact that different individuals, brothers and sisters of the same stock, should differ in such a marked degree (so that one is actually able to separate the more fertile ones from the less fertile flies by inspection) is submitted as evidence to show how selection may operate in controlling the fertility in these strains. BIBLIOGRAPHY Darwin, C. Cross and self-fertilization in the vegetable kingdom. Morean, T.H. The origin of five mutations in eye color in Drosophila and their modes of inheritance. Science, vol. 33. FERTILITY AND STERILITY IN DROSOPHILA AMPELOPHILA IV. EFFECTS ON FERTILITY OF CROSSING WITHIN AND WITHOUT AN INCONSTANT STOCK OF DROSOPHILA ROSCOE R. HYDE Department of Zodlogy, Columbia University ELEVEN DIAGRAMS INTRODUCTION In Part IJ of these studies it was shown that although the mu- tant stock truncate, produced a large number of fertile sperm and fertile eggs, yet when the truncate female was mated to the trun- cate male only 20 per cent of the eggs hatched. It was also shown that the truncates were not a homogeneous stock, for the flies with truncate wings give rise to offspring some of which have long wings like those of the wild flies.. This has held true through many generations of continuous selection. These long winged flies in turn are also not homogeneous because they throw some truncates, and this has held despite some twenty (estimated) generations of selection.!. A peculiar phenomenon shown by the long wings is that their fertility when tested together is about twice as great as that shown by their truncate brothers and sisters when tested together. I wish here to present in detail the evidence that bears on the result of crossing within this inconstant stock; and the effects on fertility when both forms are crossed into a wild stock—the Woods 1 Since both forms are under study in this paper it is convenient to refer to this stock as an ‘‘inconstant stock.”’ F 306 FERTILITY AND STERILITY IN DROSOPHILA 30% Hole stock. I shall also consider the behavior of the fertility of the extracted truncates (extracted after crossing out to a wild stock) when tested together and when back-crossed into their low- producing and high-producing grandparents. FERTILITY OF THE EXTRACTED TRUNCATES I wish to deal first with the evidence that bears on the question raised in Part II, as to whether or not the fertility of the truncate stock can be raised by outcrossing. In other words is low fertility _ In this case a concomitant of the truncate wing condition, orcan high fertility be transferred to the truncate stock by crossing it out to a wild stock and extracting? To answer this question a truncate female was crossed to a Woods Hole male. This cross I shall refer to as A. The recipro- cal cross in which the truncate male was mated to the Woods Hole female I shall refer toas B. There were five bottles of the P; made up in each case. A large number of the F; were mated in pairs and the truncates selected from their offspring. The fertility of the extracted truncates was then tested by means of the following combinations: . 1. A truncate female (F, extracted from the truncate grand- mother) was mated with four of her truncate brothers. 2. A truncate female (F, extracted from the truncate grand- father) was mated with four of her truncate brothers. 3. Control: The fertility of the original truncates was tested by placing a number of males with a single female in each case. The results of this test are recorded in tables 1, 2 and 3. The fertility of the extracted truncates when tested together is almost 50 per cent, while the fertility of the truncates used for control is only 22.6 per cent. In other words, from a source of fer- tility of 63.4 per cent (Woods Hole stock; Part III, table 3b, no. 8) there has been transferred to the truncate winged forms (orig- inally with a fertility of about 20 per cent) about 25 per cent of additional fertility. The increase in fertility is the same in this case whether the truncate male or female is used as the grandparent. 3 ROSCOE R. HYDE TABLE 1 Showing the result of testing the fertility of the AQ X Aco (F2 truncates extracted from truncate grandmother) PER CENT AGE OF 2 AVERAGE OF EGGS NO TOTAL NO. NO. EGGS NO. FLIES WHEN LIFE OF ie) NO. WHICH 4 EGGS LAID! ISOLATED EMERGED COUNT IN DAYS EGGS LAID COMPLETE BEGAN PER DAY DEVELOP- MENT 11 861 353 128 5 AT 25.0 33.4 12 211 184 80 5 30 8.5 43.4 13 yeah! Mehr 2 5 18 138 15 513 328 180 5 31 19.8 54.8 16 575 370 13>. 5 35 19.2 49.2 17 403 250 118 5 33 14.4 47.2 24 356 292 134 4 7 27.4 45.8 25 610 392 210 4 32 21.8 53.5 26 445 310 176 4 25 i a 56.7 27 552 370 | 198 4 29 Trl 53.6 28 Seite Ar S468 4 24 20.7} 5088 29 791 441 210 4 Al 21.4 |. 47.6 ARO Galen ee 3638 1786 | TABLE 2 Showing the result of testing the fertility of the B2 X BS (Fe truncates extracted from truncate grandfather) if | 3 | PER CENT TOTAL NO. NO. EGGS NO. FLIES See LIFE OF | ST eee eee EGGS LAID ISOLATED EMERGED COUNT ad DAY COMPLETE BEGAN DEVELOP- MENT 581 370 183 53 ees 19.4 49.2 1 1 5 25 I) Mog7n \ veasi 78 5 27 13.5 3377 | 352 279 158 5 25 17.6 56.6 67 42 13 5 22 4.0 31.0 657 463 242 4 29 26.3 52.3 266 266 116 agen «3 \ Seat 43.6 1652 | 790 | | Per cent of eggs which complete development 47.8. *Laid eggs which went to pieces. FERTILITY AND STERILITY IN DROSOPHILA. 359 TABLE 3 Control: Showing the result of testing the grandparental stock, truncate by truncate | PER CENT = | TOTAL NO. NO. EGGS NO. FLIES pete = LIFE OF | LINEN pean Ae | EGGSLAID | ISOLATED | EMERGED COUNT fe) | ple ae COMPLETE | BEGAN | DEV ELOP- MENT 1 154 26 4 728 iver tse 3 62 62 16 7 163.18 WO 25.8 4 | 287 265 fe. He 26 al) isat 27.5 5 [97 227 43 7 35 8 ie ie along ic F 27°36 36 10 6 14 5 AG T8 8 iy 258i) hk 7 BBR 54 Grea 27 ll eeloee 227, 10 226 126 22 Gratin eal 6.4 17.5 Mijgpal td a. | 980 292 | | Per cent of eggs which complete development 22.6. It is true that the extracted truncates vary somewhat in the length of wing. Moreover, they appear in about the ratio of 1 truncate wing to 14 of the long wings. In all cases I selected the most typically truncate forms, and if low fertility is an accom- paniment of the truncate wing as such, I should have discovered the fact in this experiment. This does not exclude the possibility, however, that some of the low fertility here may be concomi- tant to the truncate wing, for a comparison of the fertility of their long winged brothers and sisters shows that the fertility of the extracted longs is higher (see table 7). In Part II it has been demonstrated by numerous experiments that the truncate stock is deficient in egg production. A compari- son of the egg production-of the extracted truncates together with that of the truncates used in control and also with that of the truncates given in Part I] makes it absolutely certain that these extracted truncates have been benefited also in regard to egg productivity as a result of crossing and extracting from a wild stock. The evidence is conclusive that fertility can be trans- ferred to the truncates. PP) a THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 17, No.3 360 ROSCOE R. HYDE PRODUCTIVITY OF THE EXTRACTED TRUNCATES If it be true that the fertility of the extracted truncates is more than twice as great as the original truncates and if it be true that the egg production of the extracted truncates is increased, it follows that if the extracted truncates be bred together in pairs their productivity should be greatly increased. To judge from the previous history of the truncates in regard to productivity (Part II; table 1) and the attendant low fertility we should expect that by doubling the fertility of the gametes of the truncates that they would produce between 100 and 125 offspring on an average to the fertile pair instead of 50 as formerly. The number of off- spring produced above 125 would, in a rough way, give a measure of the increased egg production. This does not take into con- sideration, any difference that may exist in the length of life between the truncates and extracted truncates. It would seem from the tables that the extracted truncates had been benefited to some slight degree in respect to the length of life asa result of extracting from a wild stock. In order to throw some light on the question just raised, I bred together a number of these extracted truncates in pairs. These flies were from the same source as those used in the previous experi- ments. The pairs were made up August 9 and 10 and discon- tinued on September 1. The final count was made September 11; consequently the total number of children produced is not repre- sented. The results are given in tables 4a and 4b. Ten pairs of truncates were used to control this experiment but it was evident from inspection that their productivity was practically the same as in former experiments and consequently the offspring were not counted. ¢ st No. 178. .0-4 eee chs (10,20 938 | 442 | 47.0 | 102 lea SAMO SOP et NG, P78 se. Woke. ake 8s) 20/ 811| 365| 44.9 | 149] 88 95) Bor KE ON. W784. Dante dee | 10) 28) 613) 201 | 32.9 | 4i }72 Bal Aca GN io! 781d (a aay eae 11] 21] 247) 86) 34.8 | 21) 15 B7| Controle) x Bol nu. eben 7/15 687 265) 41.6 | G4 105 Bl Control Aon Aci. 2.0). en 7/181 754] 396] 52.6 | 148| 147 39] Control T 2 No.178 X To No.178 15| 34, 668} 194| 28.9 | 46) Control B96 Ba..w ese 12) 21/ 981] 510| 54.7 | 80] 120 FERTILITY AND STERILITY IN DROSOPHILA 363 In table 5 I have added as a matter of record the number of long-wings and short-wings that appeared in a number of bottles from which complete counts were made. No definite ratios can be given for comparison with the study of wing ratios made in Part II, where it was shown that the truncates threw 1 long to 7+ truncates, the longs in turn threw 1 truncate to 7 + longs. The ratio of the longs to the truncates in the crosses made here is near equality. The foregoing experiments bear on the effects on fertility when the extracted truncates are back-crossed with their low-producing grandparents. I wish next to consider the effects on fertility when the extracted truncates are back-crossed with their high- producing grandparents—the Woods Hole stock. The experiment was carried out in the usual way. The results are given in table 6. The relations brought out in this experiment are expressed in diagrams C, D, and E. Diagrams A and B show the result on fertility of back-crossing the extracted truncates with the low-producing grandparental truncates. It is to be noted first of all that the controls show the fertility of the extracted truncates to be practically twice that of the original truncates. Moreover, the fertility of the extracted truncates is the same whether the truncate male or female is used as the grandparent. On crossing, there is no great rise in fertility beyond that shown by the parents, as has been demonstrated in the case of the truncates crossed to wild races. The extracted trun- cate female is able to bring the fertility of the truncate male up to its level but not beyond. The extracted truncate male, on the other hand, does not bring the fertility of the truncate female up to its level by about 15 percent. Itis to be noted that the fertility shown in diagram A is practically the same as in diagram B, indi- cating that here the effect is the same on the truncate grandchil- dren, whether descended from the truncate male or truncate- female. In contrast with the conditions found in these crosses, we find, that when the extracted truncates are back-crossed to the high- producing parents there is a sudden and distinct rise in fertility; beyond that of either parent stock (diagrams C and D). The 364 ROSCOE R. HYDE BF x Bd wre ~x wse8 47.8 7.7 81.7 Cc ¥0.! Af x AG wHs HF 49.1 ees 7.7 75.7 D B73 AZ x Ad Bd 8 AGA A poe v0.6 E 64.9 Diagram C Showing the effect on fertility of crossing the extracted truncates B (F2 truncates descended from the truncateo’) with the high-producing grand- parental stock. Diagram D Showing the effect on fertility of crossing the extracted truncates A (F. truneates descended from the truncate @ ) with the high-producing grand- parental stock. ¥ Diagram E Showing the effects on fertility of crossing the extracted truncates A (F, truncates descended from the truncate 2 ) with the extracted truncates B (F: truncates descended from truncate ™). behavior of the fertility of the extracted truncates toward the Woods Hole stock is practically the same whether the truncate flies descended from the truncate grandmother or truncate grand- father. It is to be emphasized that the fertility of the truncates has been practically doubled (as determined by testing brother and sister) as a result of crossing out and extracting; and yet a marked degree of incompatibility exists between the gametes of the extracted truncates. For example, diagram C shows that the extracted male B produced 85.1 good sperm and the extracted female B produced 81.7 good eggs and yet the fertility as shown by testing 365 FERTILITY AND STERILITY IN DROSOPHILA 96 9 €¢ 6° P9 LIL Ess €°18 G°GL | 11s Aa tear jo quad Jog egz | 262 | 822 | 8I9| Grr} 269] Teh | 109} 419) Sc) 8T9 SO) je Gee | OM epogs| G21) ee [29L Fe eee eeltoem or | ren -2¢> | ON aca inecgh |, “Ces coN Ces sce ye ee 7 | Pee oem ltc| cel Fs | OF | PET OSC Po UnrGS | 2G, | OE eae ca) =| = €6 | Tar Ne eeaeee =| OR 2) Sitwil Cetlneeee OG aC lml te: sinS0y-(c ORSenGTen| COeeiie san i ra | Wieroaeerece Meise cr). Gil 80a ARs OIeg eens es nO, 0G) Cies Malley ae aad ier 12 | ¥ 9 Or GT OT GG GG 0€ §I CT ia! 61 tll Pies | er a ee Bie 0G | Bee rtoe| ie eee | Sic nes Cen era ieee OS thee Tia ecaem Coen ier bau le ca se heeetee 61 | Ge oT) Clemle Call O1e cre eres OCe ea Oe ae ee OES Cleo G Ls iloek Se te ce 8I | epolleen (eee tree ai) Os Ie Ope eee eOee ole OGn OG sere AOy Glerega lf Cas |b) aarti iii | Ail ieee ees lceces eee oe ezeelnela se Greh ate el MG sllaver ale Oia ele Oars ej ie gr OGu SS) |= Tce sce IG LE 6€ cP cy | OS Ge | =oG¥. GE Orca tees ae Teas GT Tegel re ie Ol are aioeee Oe ie Os CoO) eats [aes lee COO Ci aie a chee ia | Gilera letecuece wl on lee Gra | ee Si OS Gia lech Cy eres a ee €I Gene epi eee rogue tecerlOp etera (eOS et sOlr 1h Os \ierik |e Olea ety. ict | seer cle ral eof locteiner occes | de a0. 2p! OG, 08" | 200-1 69) 00 | 99 i OF eee | 08 IT ipa 2 \eagajeoye|' 02 | O28. | Ty te | OF Ge 2i |-0e | ep 4 OU ese | al OL ea uemeecelcen | sen 0s om | se) | Ge) Sh ser) Ge cr OP vines Og. | ee 6 meen mcr ce ci ue eon | Ge 108. Orel 02 | 07 Te RI TEx Se riat09| Sai 8 1g Tg 8I If €€ OL Gi lOe && SP za 06 G TW 9€ OF \ ae rece: ane L ‘sny s}[Npy | s3aq synpy | S53 SYNPY | ss3q |synpy| sssq [SHMPY | S99q, | SIIMPV spoq |sinpy | ssdq isynpy | soa E161 aLva pees) | PaXév | 8aXOV |PHAXSHM) PAX OHM | evyXéHm | évXfHM | 6aXLHM -poaquoa sof pasn “ssp yy ‘910 Hf SP00 AI 0 paanjdvo yoojs pyun nau = fy ‘(ajpway IIDIUNA] WOLLPI}ID.YLI %yq) BDIUNAL) PIJID.AjLA = Y + (APU ajpound} WoL, PIIVAZLI FY) BDIUNA) papovi4jxa = _{ °219H spoom = HM "y90}8 2]0H SpooM ayj—sjuaindpunsb Buronposd-ybry 1294) YUM SIDIUNL} PI}IDL}LI OY} Bursso.o-yong Jo hyrpysaf uo yoaffa ay; Bunoys 9 ATAVL 366 ROSCOE R. HYDE Extracted Truncates Original Truncates gi? Bo/ 80 TT et x ad= 478 TS » e— 22.6 471 F 32.9 Extracted Truncates Original Truncates Pom] &73 80 SS Ag x Ad= #71 TH x TP= 226 a) G 34.8 Diagram F Showing incompatibility that exists between truncates and extracted truncates (extracted from truncate <). Diagram G Showing incompatibility that exists between truncates and ex- tracted truncates (extracted from truncate <). together was only 47.8." Diagram D shows practically the same condition. As just stated the extracted truncates produce 81.7 good eggs and 85.1 good sperm (diagram C). Weknow from many former experi- ments that the truncates produce about 80 per cent of good sperm and 55 per cent of good eggs. Yet when these gametes are brought together, as shown in diagrams A and B there is amarked degree of incompatibility. Diagrams F and G based on this and former experiments, will show at a glance the facts that relate to incom- patibility in this case. | It remains to be pointed out that there is a rise in fertility beyond that shown by either parent when the truncates extracted from the truncate grandfather B are crossed with the truncates extracted from the truncate grandmother A (diagram E). It seems strange that there should result a rise in fertility in this cross and also when the extracted truncates are back-crossed into their high-producing grandparental stock (diagrams C and D); and yet there results no rise in fertility when the extracted truncate female is back- FERTILITY AND STERILITY IN DROSOPHILA 367 AF x Ad ast x 8F H 746 Tid Diagram H Showing the fertility of the extracted longs when mated together and when crossed with each other. crossed to the truncate male; and that in the reciprocal cross there is shown not only no rise in fertility but an actual in- compatibility (diagrams A and B). It has been shown that the fertility of the truncates is raised from about 25 per cent to 50 per cent by crossing out to a wild stock and extracting. It must be remembered, that the fertility of the wild stock from which the truncates were extracted was far below 100 per cent. It seems altogether probable that had a more fertile stock been used the fertility of the extracted truncates would have been still higher. FERTILITY OF THE LONG-WINGED BROTHERS AND SISTERS OF THE EXTRACTED TRUNCATES The question still remains as to how the long-winged brothers and sisters of the extracted truncates behave in respect to fertility when tested together. The data recorded shown in table 7 gives an answer to the question. The results of this experiment as expressed in diagram H show that the long-winged brothers and sisters of the extracted trun- cates are more fertile than the extracted truncates (in this case by over 25 per cent). It would seem from this that while it is possible to put a certain amount of fertility into the gametes of the truncates, yet a certain degree of incompatibility remains ~ with the truncate wing as evidenced by the fact that their long- winged brothers and sisters give a higher degree of fertility. It is to be noticed in this case that there is no rise in fertility on crossing. The fertility in the crosses is practically the same as in the controls. 368 ROSCOE R. HYDE EFFECT ON FERTILITY OF CROSSING THE TRUNCATES WITH THEIR LONG-WINGED BROTHERS AND SISTERS AND THE EFFECT ON FERTILITY WHEN BOTH OF THESE FORMS ARE CROSSED INTO THE WOODS HOLE STOCK I wish to return to a consideration of the low fertility of the truncate stock and consider more especially the difference in fertility between the truncates and their long winged brothers and sisters. I shall also consider the behavior of fertility when the truncates are crossed to their long winged brothers and sisters, and also the behavior of fertility when both forms of the inconstant stock are out crossed to the Woods Hole stock. Table 8 shows the combinations made up, the number of eggs TABLE 7 Showing the fertility of the long-winged brothers and sisters of the extracted truncates when tested together and when crossed with each other. A = Fz long-wing extracted from truncate grandmother. B = F:2 long-wing extracted from truncate grand- father AGXace BQXBO AQXEBO BQXag DATE | ate ia a ——|— ane ei, = 1913 Eggs | Adults) Eggs | Adults | Eggs |Adults| Eggs | Adults August 15 )) 30 | Sf jie40 26" 165 5) Sat) s0meas 16 BU: | 28308 21° 45 BB) SOP ees 17 50 | 35. } 38)°)) 19 || 40: 2a) aos es 18 30) bed? WRG. 530) 12 hl eS ees 19 20. cl 32) Bar) 38 | 26 aise 20s 20 23/18" | DOA > TS = esOwy Aan) ea Omelemeee 21 113 71400 | $33 Oh) ae a2OrgIN IS, teal 22 ) 40°) 2860) SB We Bea O a ae Eh as et 23 405) 87) 70. BSN eS a8 570) see 24 80.) 60. | 7d | 5S G05) 955. "|. 60" Wade 25 50.) 37. + 63) Oe s0 ".89": | “AU ie 26 50\)| 35: | 8B 173. 70" 1) “BO, soa 27 40°) | 38.) 520) Bbe55. | Sa) abana 28 50 | 34 | 60 |} 539 4) 30 -| “37 "1 9h0 = pas 29 AQ> |) B24. 40 7) Oe) 50° 9 250) a ae 30 p 47. 40") es eas 40. |) aes oie Tatas steerer. Cate 643 483° 739 | 566 | 729 | 515 | 697 | 520 Per cent of fertility ....... 15.8 | 76.6 Lok 74.6 FERTILITY AND STERILITY IN DROSOPHILA 369 isolated day by day and the corresponding number of eggs that hatched. The truncates in this experiment gave a fertility of 31.3 which is from 5 to 10 per cent higher than in any of the previous experi- ments. The Woods Hole stock shows a fertility of 83.6 which is much higher than that of previous experiments. The fact is that all the stocks at this time showed a rise in fertility. Whatever the meaning of this may be, the former relations hold as to the effect on fertility when crossed, as is evidenced by the fact that there is a marked rise in fertility when the truncate is crossed into the Woods Hole stock; compare diagram J with diagram C, Part. 1. Diagram I brings out the effect of crossing within the truncate stock. The fertility of the truncates when tested tagether is 31.3 while that of their long-winged brothers and sisters when tested together is 51.9. When the longs are crossed to the truncates as the diagram shows, the high fertility of the longs is able to bring the fertility of the truncates up to its level and it would seem that in the case of the cross between the truncate female and the long male that there was a marked rise in fertility. The fact is that in this experiment this cross is higher by 10 per cent than in the case where the truncate female is tested with a wild male. I have made many crosses between the truncatefemaleand the Woods Holemale with the result that the fertility of the combination stands almost invariably at 55 per cent. It would seem from this experiment and the evidence from former ones that the rise in fertility of this particular combination is significant. A study of the diagrams will show a marked degree of incompat- ibility to exist between the flies with the long wings for while 51.9 of their eggs hatch when mated together yet no less than 90.7: of the sperm and 73.8 of the eggs are capable of entering into a combination that results in development, as is shown by crossing into the Woods Hole stock (diagram K). The greatest degree of incompatibility however exists between the flies with the truncate wings for in this experiment they pro- duced no less than 54 per cent of good eggs and 88.5 per cent of 370 ROSCOE R. HYDE jie Rte TS EG. x uz@ 64.4 l ar! TF x wie x wk F? 33 fd 336 54 J 88.9 ee x Le wht x wHt 51.9 836 73.8 K 90.7 Diagrams I, J and K Showing the effects on fertility of crossing within and without the inconstant truncate stock. T= truncate. L = long-wing thrown by truneates. WH = Woods Hole. Diagram I Crosses between truncates and their long-winged brothers and sisters. Diagram J Crosses between truncates and the Woods Hole stock. Diagram K Crosses between the long-winged flies thrown by the truncates and the Woods Hole stock. good sperm and yet only 31.3 of the eggs from the truncate females hatched when mated to the truncate males (diagram J). It remains to be pointed out that a certain degree of incompati- bility exists in the cross made between the truncate male and his long-winged sister. For their fertility when tested together is 47.1 and yet he is producing 88.5 fertile sperm and she is producing 73.8 fertile eggs as shown by outbreeding into the Woods Hoie stock. In the cross between the truncate female and the long- winged male the incompatibility seems to be removed. For while the long-winged male is producing 90.7 fertile sperm the female is producing only 54 per cent of fertile eggs. Here, as I have STERILITY FERTILITY AND IN DROSOPHILA (Aqrttgtey Jo | L°06 | VS G88 6 1g ilts yao Idd CIL 968 682 STS) 9FE | 69F | 601 | coz | 998| 626] z8z| erg] veo| eze| soe| rer] 9c | gen | TPIOL 8I | 0G Le | OF | | | 96 T& QT | 91 Flt |e 8G 9 0G | 9 bobs cal lel CparOrs|arh | 0g ly | 09 166 Sle Cisse eOGe We Pie s| lca) g Ge ee Wet P al eee | Pe | | ase a ae ZN | ONO) Mal sek aI (YS gD se 0 SB ce |G Ig 6& | Of | 0€ | | lL ereeupese T€ OF 1G | JES Maye cS | rea Or | 1 ‘3dag S& OF | 02 GS | | | Ws OF VG te) OL |) 8c OT 0€ | OL Og | Té GE OF bP OF i Cz | 98 | OF GG 0€ 66 6h | 81 Ces elle | cee lOc ad 0¢ 9€ vP L Sia | | as OF 0% LE LG SPS ee GLE Or LG | 66 OF og | OF OF 9G €€ ae le) elect | GG vG'| OF 0€ 09 61 0g 61 0g | 8 &F 0¢ OF OF al LT G | Ga i206 | 09 | & GP GG | OF 1G VE FI OF | 2c GF GS GF 0g &7 Og Gao Sg ae Gy see) 0G G§ 0€ | &F 9% | &¢ | LT Og | 9% 1g | 09 | | | de ath re | 29. | | | 8% | 9g | | | 18 | g¢ OF &¢ GP 1¢ al Lo<3) = Sema VY SIERSY OS al 0€ OF | 9 | ZI Iv | 8% | 0g | 9% a0 cs Efe ards Bi 00 ace OF.) 88 OOT | 6F 09 Gm OL GG 06 Go | F9 | HG 09 €L 46 | OOL|) 9F og | 8I | 0€ 88 OOT LHSIGM NINIVD -ASdOLOV DTHOIGM AGO -OV j «a puys yjoys ay} jo joao, 48 pews, yonprag 402 o _ o NOITL -VuuddO 10 THOLVYN ponunu0D NOIL adv | “ON GSV0O | -vuddo LV NOIL NOILVUddO -vudado do adlva iv LOOGIAO “ON NOILVUGdO | HO NOILIGNOOD seovjins [vouo}1ed SUIAIBA 0} poyoRy -ye pues umeUu0o}t -1ad Aq pepunorins Aqiaveo) =Apoq = Ul SOSSUU Y[OA AUB] SUOISOT] -pB jo ssvur B 8100 aah fiagnency ary umeu -O}IIed Ul pasolo -ue SSBVUI [OA B JO qyaed & pus AivAo 043 ydeoxe spiiq iay3o Aq yno use}ea wiaosta “yof Apog Ajydure o10M BUISEA puv pUuL[s 1194s foIN{eST] puooes oy} pessed pey suou AyT}uUe -redde ynq uor1z10d pesieAel oy} Ul 918M O89q} JO OUIOS ‘SOsseUl sso pous sprey WTA peliy pue osie, AIaA AIBAO uo sesseul Y[OA piey 0} ISByl UMOIS pue suisojue [ou -unj: prey Gove pue YIOA B SuUIpUNnosins yove sesseul x9 1 peureyuod ‘uoTy -Ipuod SurAv] UT pojou JON, pivy [[B e10M soy? ynq ArvAo uO Ssy][OA as1B'T AIBAO UOSHOA [BUI -10U OFI8] UBAVG ‘OW T “BIA F ‘sour 27 "SOUL QT (aun -[0o Sut -pe0 -o1d ut ajou 20g) 9ZET O1TS suUlvId 00¢ ueyy e10UL YsIom parq ® yons jow1e0 “BTA “sparq r9y40 Aq yno u0389 u90q pey B109 STA heay TOHSV OFAT S16 G6CT IT “28]N 606T 0¢ “Ady SIZTUO}TIOg uselds poiny -dni Woy BUT -p9e]q [eusezUy AYIABO Apoq uedo sur -yord spirq 19440 Aq Pelle SISOUIO}SBUB qsig Mojeq =4ysnf pesos osessed pus peyeatmys punoyz svM jgonp Asdojne ye ‘gonp ey} OFUT yovq pasoulojseus pue pue-10j-puo peuiny ‘peqoes -a1 sem dard oy} ‘uoTz10d usouNng -[8 pus snuryyst jo uor}z0d & jo AfIe] -od Jo [Bs19A0Y einyzest] @ AT[BNy -ITA @10Jo19y} SBM yor SISOUIOY -seue oy} pessed pey suryyou 4% 43 poswodde yt Asdoyne 4B ‘sis -OUI0JSBUB PUe-0}F -pue puB snuly}st jo U01}09893T purys [us jo jeae, 48 peyesry JONPIAO SUTAB] BULAR] eye -uBjur "SOUL 6 ‘SOUL 6 “sour £9 606T St “FH 606T Ot “Fa 8061 et AON 89 *9 9¢ ai IT or 403 SOBSBUI S59 JO UOT} -oBjtIjnd IO s1}1U04 -1ad JO USIS OU ST -"Q peulees pirq jo UIsT[OGBeyY “ouvIqG -woew ysno1yy OTQIstA AStp W183 Y}IM B50 [BUI -10U Yserj A[JOaJ10d T pues peqiosqvun [[8 9ouvysqns so jo junouwe |[eus B YIM [ ‘seuBIq -weur sda Ajdwo 6 910M AYIAGO Apoq ut ‘of ipog tu -19Y WOIJ UOT}POsUI 0} enp pervedde s17 -1uo}ied =! aineu04 -l1ed Aq Yo pelea ‘eoBO]d OUT UedO BUIDBA : P1TG BUIAB] B UI §}1ed as0y4 Jo 9ZI8 94} 919M 4nd oy} purysq BUIsBA pue pus [jays jo qasd ‘pus [joys ut AIJeyo par UOUL -U109 B JO 8Z18 SSBUI snoeIBo[eo ‘ss00 SUIGIOSGe pa1eAod euBIquieUl | 8 UI -10u jo dn opvur SOSSBUL BIO Y}IM P9TF yonp ‘ony -es1] Jod’in 9Aoqe ‘ud GT peanjound ‘uolyIpuodD =S UT -AB] Ul JONPIAG jyonp 9} Jo oo0Id 4s1y jo pue 48 ‘ay ‘aime -SI] 94} eAOge ysnt y[OA prey jo dummy B SBM a10q} yng yonp ut sessBur S59 OU ‘UOT}Ipuod SurAB] 0} posiey SUIGIS PET 4SODIB] “YM ‘sy[OA [@Wid0u 8518] § :Seuo 9]/qIsTA ysnt 0} 9918, UO} O2Z18 UL SUIAIBA sa]dT] -[0} OT ‘uortpuos SutAvy ut AIvAO [BursoUqe [RB ‘SOUL $1) $68 IZ eung) wep 105 parE] -o}1Ied pus uory puvy]s [[eys 24} YSNo1y} ss0108 qno SBM JONPIAG (Seinyest] OM4 ueemyoq SSO10v lelbed 404 ~~ eal oD onl Ayiavo Apoq UrI/-ua pey yoNprAo | ‘prey autos !AIBAO -qns put eraszey | yno eqn}) “jonp SHIOA ¢ Qof Apog | ay} josyaed yyOg uo syoX uay | | 06ST |Sl “AVIN| [BUIMOopqy ay} jo uolzoeg | E | 2 os od rc) = ee y ape | ak |laeOjapo| ap | | 8g 8 546 8% BAgiBI5| Sa | $24 So” | os ASdOLAV LY ASdOLAV LY Zee | Eo erie oH) oes ae ae Bon | 2b He PENSE A HS BAKEYND) EN END Paes | Ce tees eta 40 asavo -Vuad0 40 auayN| B28 = JO NOILIGNOD 410 NOILLIGNOO (eyes ll fe es) S| $ il ae ; 5°°0 g SC 4 | Beecie agr emia ese) ec (cea ie ea 2 | io] a | £ : | fe Re | Aa a Ae | Bs if ' “ON GSVO “ON NOILVUAdO penuymop—T AIAV.L pep100er l | aa you suBsI0 190430 punoy UOT}IPUOD SUTAB] | ‘sour Z | | 606T Is SBA OYSs ai1jue pa) 92 806T jo uoIqgipuodg jou JONpPIAQ: | -uou AyyReYy UT “IAT | 6S | cOOT | OLOT eg “Q9qJ| EsNvOIq poT[Iy |-Aouel yonplAg -UBJUT |"SsOUr ) 'g 09q | | | suo] | | "urD oT | papsooas | | qnoqe | you jeuuny jo azis | | | ‘potsed | wineuoj}ied | ‘SurlAvy] ueeq Ae} R] | | | SurAvy ur pesojoua sy]OA | pey yYoryM piiq | BUISCA SIG 10} peqiosqe jo seq |e jo uoTyIpuod Ut AIGAO | | | pue puv[s [[eys jo /as1reyus | -UMNU 918] B POUTe} | BUISEA PUB PUTS | uO Sy]OA paquosqe | ‘sour g | | S161 Wed ‘jauunj ydeoxe | 07 Suru | 806T -u0d A}IABO Apog eqs jo yeq | Ajred pey Aue “SIA F | OO8T FI AlN) sisojnoreqn 7, panowas jonpug -ulsog |'sOUI 9 |pZ ‘AON pep.z0o | e | WIN9UO} | -2a1 JOU BUITGA JO 4[npe -Wed Y}IM pareAod | 9zIs 10 d90ueserd | SUOISeYy PG BuISeA | gnq | sesseul HOA YIM |S uo0rsy10d youuny pasy | | 6061 | 0} eNp eurysojzur |) puw jouuNy ydeoxe | SurAvy 8061 P2e[G AVAGO Apog ydeoxe 9u03 [TV AIGAO UO SHOX ‘soul ) | TIG—| 6861 | OSIZ $o PUNE) Jo ossvddoqg parowumas janpag -aou | ‘sid $1 QZ “AON uey SUTAB] BOBOTD 94} 0} por[d G UI VUISEA OY} JO Surtqios -de A]eso]o BUIsBA | AW | ozI8 04} SBA YOTYM |-qe [VIeAES oOB8Te | aq} jo jaed oy} jo -Avo Apod ulsyfoA|¥UTS eA 4ydeoxa |‘sy[oA aB1v] [eI9 | ‘sour G | II6I Wor}d90xe 94} YIM | ayT9 | g06T ou ‘70f fuaa ipog 903 [[B JONPIAG -AOS JIM AIBAGQ “siX Z| 691Z | 69ZE | OO9T $2 “SNY| eIEp 10; parry parowas qonpug -UBJUT |"SOUI ¢ [ET “AON SUlvIs | Gp SH[OA YIM “gn | BoBOTO 0} patd | ‘80U0 [BULIOU |] BUS -de Ajasopo vulsva | Ajtaeo Apoq | qJueseid uojoo eq} | euI0s pue syHIOA | oy} jo y1ed 949 jo | Ul ae1} JOA ATJUap | 0} peyorzze vursea | Surquosqe [[eus | 6061 mortydeoxe oY} FIM | 9TTt | S06T -IA9 SOSSBUI PIBET joyred oy} AJUQ «|g YM AwAQ | sour, | GIg | C66 | OSFT FT oun umouyuy parowas gonprag -UBJUT |SOU CG eT “AON BoBOTO oyur uedo 4yonp OBS ITB [bu “TAO ‘Suey SUTAeT | -IWIOpqe 9} OF [es | 8 Ul 818d asaq} Jo | -1op sdi] pasny 049 | | 8218 [BUIIOU 94} 0} | afdI[[OJ peangdna | | jo uotj10d 94} pus | pesie[ue pey uor | esiey eymb ouo0) | | WO[Od 949 0} paryd | | | Aytaeo Apoq | ~eredo ay} 4 449] | {WO | PAO” s¥IO‘ | | | | |-dv vursea oy} jo | UI HJOA paqiosqeun | Jonp ayy Jo suoT} | 9 SUOT}IpuoD [UT | | | 6061 | qaed oy} ydeoxe pa | 94 8061 ou “jof huaa fipog -Iod |[BuIs oy, -10U Ul AIBAG | "SOUL 6 | 8OZI | 992% | O90T 6 “SNY)| BJep 1OJ poy] = -Aoular yonpiag~g, -UBJUT |'SOUI G 08 Bye) 405 AT[euse8} -x9 10 A][Bu1EzUT worzeIIeA | yon} Jo souvivedde poeiq jo osuei MOT | OU !pot} 9q OF rvod -9q jou ynq poeiq | -dv you pip sseyy GOL 489 dY} OJ OZIS ODRIOAR | !PU]S [TOYS JO [VM | -B1V] JO “4M tAIBAO soingns poydni MOOG SBF ‘GOGT ‘ET | UT PUNoy HIS jo| WO syTOA [Bw -leyur Aq pura ‘IB SOUS Burkvy | seoord ‘uoty1puod | -1oU OBI] F : UOT} [194s jo worIs10A 606T useq pey pilig SUTAR] [BUIOU UT -Ipuod SUTA] UT eyep IO} pays, | -UL;eurpnfisuoT SUTAB] BOUL Q YT “Gay $2 poMes useq prey ynd 910M onsst? UL [ITIS YPIS ‘]peus eat] @ Ayqissod peciq roy ozs | YSNoYZ[V odvys | suIvIS CF] YsosIv] 1/8 oy4 eSvIIAB JHoqe sxs0 | [eUIIOU Jo puUBys | Jo “ya ‘AIvAO UO Jo a0erd jo [BAOUIOI '606L ‘91 “BIN 90uTS |[[9GS !uor}1puod But | SY[OA OGsRT fF SOT} Aq pues [joys jo | 6061 ple] sey pilg -AB] UT JONPIAOC -Ipuod SUTAV UT BJep 10} PoT[IY =| 9218 Surysturanrqg SUIAR] |'sour g OT “qoq (ye SNUINYAST JO VOUBIY -u9 48 Sse ‘yonp ey} JO usUIN, 9u4 sulvIs UI UOMOII{SUOD OU | (GT JS9D1B] JO “FAA suo] ‘suid fF Yony G06T | PPBUr 41 ING JAB] | "UNO T BAOGV SH[OA e Aq pexor13su00 ‘oT ‘avy Sdo peur}ioyno Ur peAios |[VuIou Ff ‘UOT} yonp jo uoTser USUI 6061 -10U ke ie -o1d sem yony, | -Ipuod SsutAvy] uy Byep OJ poy] | -Nqye jo uowMny SUIAB] SOUL TT 6T “497 GS suo] "ud 47 ¥Boo repnosnur ey} ur| | pue[s [[eYys Ul dso GT 489 yon} &@ duryey Aq 6061 | PETOUS 9308 !3pong | -Dxe] Jo “4M :ATBAO peysturmip yonp “Tuer oourssutAv] | OY} JO UsIS OU UOT} | UO SY[OA 9 ‘UOT} jo uolser usewNgq 8061 useq pry parg -1puoo SutAvy UT -[puod SUTAR] UT BYEp 1OF poy[tyy | -[2 949 Jo UowNT SUIAR] |SOUL g \ZZ “390! ZI | IZ areleblerclenc| as eo8| ah | at |2/é B46 | 82 Bauleho| o- gc2/ 8" | Bs || 8 XSdOLAV LV XsadOLoAV Lv a= Bu aa Bax me ond BoM Fie 6 5 Z SuuV Wau LONGIAO AUVAO 5 e aa a e 4 ° -vyunao do auntyn| 44 8 e G ap lee JO NOIWIGNOD JO NOILIGNOO a a a = » e c) 8 e S | g ‘ e i = > 3 re| 2 | 88) 28] . 3 po eras penurao)—T HTaViL uGHitpuod DULAB] UL Ueaq 1odAd PBY Pd 347 3ey4 9OUIPIAD OU ST O19T]9 Pae pre] you pey jeu -I0U GIBOSIA UOT} -Ipuod =SuIAB] UL ueeq JeAe psy pig oy} yey} sdUepTAa ou {prey jou pezy si50 [eulIOU [B19Aes prey pey AYIAGO Apoq O4UI pee “nao pey ynq prey JaAou pey plig ulese Avy yOu PIP ‘Surmoyjoy Avp ay} uo uorze1edo JO 9UIT} OY} 9B puBys | ]JeUs UT se Yorum 350 044 piey Sul -JUBM puBS [[eqs ey} suoje quouw -BSI] [B1JU9A 9T]} JO yasd @ Suor1yIpuod SUIAG] UI JON DSUIZUBA Pals [[AYS Jo peey 38 JueUTeST] [¥4qUeA ‘!UOT}IpuOD BUIAR] UL JON pups [eqs jo opis [BiquoA oy} Suole aposnur jo sstul pivy v SBM 3104} pue pue Joddn 4e [euli1ou qnq snuiyyst oy9 jo Y}su9] 949 Yno -Yonoig} SuUryovy JUSUTBST] [B1qU9 A BoBOTD 0} uedo BVUISBA ‘AjsnorAeid = sABp OL PIE] peg qorym p4rq 8 Jo yonp e jo 9zI8 JoNp ‘Suryovy quoulest] [B1}U9 A uOoT}Ipuod SUIAB[-UOU UT UOT}Ipuod SUIAL]-uOuU UT AICAO uo Ssy]OA BSurqios -QB 9B1G] [BIBABQ ‘SOUL F ‘sour ‘soul fp ‘sour ty ra 686 ELP 988 6861 £602 9LET SEP 009T OZ9T O6FT GI6T TT “WOW OT6T v6 “UPW OI6T fF oun S10 p10 “SIP OATISA5IC Pe S1}1U0}I19d Jo osnBoeq pe] [I SIyTUO}TIEd Jo Sea OS 2 1D3i O61 touruINs ul pjos A]JUuep -1008 SBM PII peAowed Sao] “UID 40 Juowrest] [easy -U9A 9} JO 9091 PO9AOUIEI SUOT ‘UID T Juoeurest, [e14 -U9A 9} JO 9d9Tq pea ouler SUOT “ud F JUSTABST| [B1y -U9A OY} JO 900TG queut -BSI[ ]BIQUOA 911} -ud JO [BAOUIOY pexBsT] Osye [ouUN; : PUBS [[OYS Jo oprs [B4jueA UO 9[OsSN UT jo sseu punol B SI JI o10yM gyno JUS UTBST| [B13U9 A a9 -uvjut aq -UBjUT ST -wejar Sey -ueyul pur[s eqs ur }Je] sso ‘DaTAR] f | | | 6U6T “SOUL F RT “AON 6061 “SOU J |f “AON | | | 606T ‘SOUL $G$ “AON 606T ‘sour 2 AZ *39O | | 6061 ‘SOULG F “ABI 61 IT GIT 601) 62 86 ~ a 407 THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 17, No.3 408 RAYMOND PEARL AND M. R. CURTIS ‘shrinking’ no doubt accounts for the belief of the early workers that extirpation of the oviduct caused degeneration of the ovary. Our experiments on the removal of the oviduct do not bear upon this point since at the time the ducts were removed (cases 15 to 20) the sex organs were in infantile or non-laying condition and they were not examined for at least seven months after the operation. Cases 2 and 5, however, show that the ovary may be in full functional condition 13 days after the funnel is ligated. THE EFFECT OF INTERFERENCE WITH THE OVIDUCT UPON THE SECONDARY SEXUAL CHARACTERS The early workers Yarrell (27), Bland Sutton (’85), and Brandt (’89) state that the removal of the oviducts is followed by the assumption of male secondary sexual characters by the female. The belief of the early workers that removal of the oviduct caused the degeneration of the ovary no doubt led them to misinterpret the normal growth of the comb, wattles, and spurs, as the assumption of male characters. None of the birds from which we removed or otherwise in- terfered with the oviduct showed any tendency to assume male characters. At the time of autopsy they were as female in all secondary sexual characters as normal birds. This result also agrees with those of Sellheim (’07). He summarized the results of his experiments as follows: Damit ist -fiir die Henne der Beweis geliefert, dass die Entfernung des Legdarmes keine schadliche Wirkung auf die Funktion des Eier- stockes ausiibt. Dass die Legdarmresektion in ihrer Wirkung einer Kastration gleich setzen sei und dass die sekundéren Geschlechts- charaktere, Bartlippchen, Kaimme, Sporen, Gefieder, Beckenform, Stimme, benehmen gegen das andere Geschlecht nach der Exstirpation des Legdarmes sich dindern, ist renes Phantasiestiick. THE EFFECT OF THE LIGATION OR REMOVAL OF THE FUNNEL UPON OVULATION Coste (’74) describes the infundibulum embracing the ovarian yolk in its follicle directly before the time for ovulation. The same phenomenon has since been observed many times by other workers, including the authors. So far as is known the in- PHYSIOLOGY OF THE OVIDUCT 409 fundibulum has never been observed’ in the act of swallowing a free yolk. Further, the work of Patterson (710) and Bartel- mez (12) shows that normally the follicular orientation is pre- served in the oviduct. These considerations make it certain that normally the yolk is ovulated into the enclosing funnel. Coste believed that the pressure of the funnel upon the fol- licle was the probable cause of ovulation and Patterson (’10) apparently accepts this view. The walls of the funnel are mus- cular and at the time it embraces the follicle it is in active peri- stalsis. That the pressure is sufficient to cause or at least ma- terially to aid ovulation is a natural inference. 3’ That yolks set free in the body cavity may subsequently enter an oviduct seemed the most reasonable explanation for the conditions observed in two birds (no. 17K and 397K) autopsied at this Laboratory. Both these birds were pure bred Barred Plymouth Rock hens a little over one year old and each had laid a number of eggs. Bird no. 17K had not laid for four months and bird 397K had not laid for one month. At autopsy the sex organs on the left side of the body of each bird were in practically nen-laying condition. In neither bird was there any right ovary. Each bird had what appeared to be a right oviduct filled with large egg concrements. These concrements were exactly similar to such masses often found in abnormal conditions of the left oviduct. They appeared to con- sist of concentric layers of hardened albumen surrounding hardened yolks. In both birds the upper end of the tube was greatly distended. Its walls were stretched thin as is a left oviduct containing large egg masses. Smooth muscle fibers were visible in the walls. Each bird showed considerable peritonitis and there were adhesions between the walls of the tube (oviduct) and intestine. The funnel mouth could not be distinguished. The lips had apparently adhered together; (this is often true in the case of egg concrements in the left oviduct). The tube continued for several centimeters behind the caudal end of the masses. It ended blindly near the cloaca into which it did not open. There was no dif- ferentiated shell gland. In Bird no. 397K the tube had ligaments very much like normal oviduct ligaments. In Bird no. 17K the ligaments were not exactly like normal oviduct ligaments but the tube was held in a fold of peritoneum. There seemed no doubt that these egg masses were egg concrements formed in the rudimentary right oviducts and little doubt that the centre of these masses were yolks. If this was true the yolks must have been ovulated from the left ovary and have passed across the body cavity (behind the gizzard in Bird 17K and by this path or through a hole which was found in the mesentery in Bird 397K) and there have been picked up by the righ* oviduct. It should possibly be added that other large right oviducts have been observed in the routine autopsy work. One case (Bird no. 276) had both right and left oviducts in nearly functional condition. The typical parts were differentiated in the right duct and it was open into the cloaca. In this case also there was no right ovary. 410 RAYMOND PEARL AND M. R. CURTIS Bartelmez (12) has shown that in the pigeon continued yolk formation increases the pressure within the follcle so that “the egg bulges out when the rupture begins” and “the egg is over a millimeter in diameter greater after ovulation than the whole follicle was before.’”’ He and other workers recognize in this increase in internal pressure an important factor in the rupture of the follicle. That the pressure of the infundibulum is not necessary for the rupture of the follicle is proven by the experiments of Sell- heim (’07), which show that after the oviduct is removed ovula- tion takes place into the body cavity. This result is confirmed by the present investigation which shows that ovulation into the body cavity occurs when the funnel mouth is closed by sew- ing or ligating the ostium (cases 1, 4, 7 and 8), or by removal of the duct (cases 15, 16, 18 and 19). In cases 4, 8, 15 and 17 the birds were killed during a normal period of egg produc- tion and (with the exception of case 17) showed evidence of recent ovulation into the body cavity. In these birds none of the yolks remaining in the ovary were apparently larger than yolks found in the ovaries of normal laying birds of the same breed. Quantitative data are not as yet available on this point, but the observations indicate that absence of pressure exerted by the funnel does not perceptibly delay ovulation. This suggests not only that internal pressure is a sufficient cause for rupture of the follicle, but also that it may be the most important factor in causing such rupture in the case of normal ovulation into the oviduct. THE EFFECT OF PREVENTION OF EGG-LAYING ON BODY METABOLISM It has already been shown that if the oviduct is removed or if the mouth of the duct is permanently closed the ovary passes through normal periods of egg production and the yolks are ovulated into the body cavity. Cases 9, 10, 12, 13 and 14 show that when it is possible for a yolk to enter but not to leave the duct, the duct may be filled with egg masses and ovu- PHYSIOLOGY OF THE OVIDUCT 411 lation then takes place into the body cavity or if the duct be stopped at the level of the shell gland normal eggs may be formed and passed back up the duct into the body cavity. The fate of these yolks and eggs and their effect on body metabolism is of some interest. At the height of a period of egg production a bird is con- suming a large amount of food materials which she is elaborat- ing into eggs. In the cases cited she is discharging these prod- ucts into her own body. Is she able to resorb these eggs and if so by what process is it accomplished? And are the resorbed eggs utilized in the body metabolism? There are in all fourteen cases where the bird at autopsy showed that she had been either ovulating (cases 1, 4, 7, 8, 10, 12, 13, 15, 16, 18, 19 and 26) or backing fully formed eggs (cases 9 and 14) into the body cavity. Four of these fourteen (cases 1, 7, 12 and 26) died from peritonitis which may have been caused by inability to absorb the yolks in the body cavity. The other ten birds, or 71 per cent, were able to resorb the yolks or eggs. Moreover, in none of the four cases of peritonitis were we certain that it was due to the presence of the yolks although this seemed the most probable cause. In addition to the fourteen cases cited in the preceding para- graph there are in the archives of the Laboratory autopsy rec- ords of other birds which at the time of death were ovulating into the body cavity and resorbing the yolks. Two methods of resorption are observed. First, absorption directly through the general peritoneal surface, and second, walling off of the yolks or eggs by peritoneum and subsequent absorption. In eases 1, 7, 8, 14, 154 and 16 absorption was by the first method. None of these cases present such clear evidence of rapid absorption of several yolks as some cases met with in the routine autopsy work. Descriptions of two such cases follow. Bird no. 1406 laid but seven eggs during her life. The last of these was laid August 22 of her pullet year. She was kept 4 In this case the yolk had been entirely absorbed but the follicle on the ovary was large and it therefore seemed impossible that sufficient time had elapsed for absorption by the second method. ’ e 412 RAYMOND PEARL AND M. R. CURTIS until February 6 of her second year when she was killed for data. She was twenty-two months old and was a large and very fat bird. In the uterus was a small membrane-shelled egg weighing only 23 grams. The yolk of this egg weighed only 4.87 grams. The oviduct was normal except that the opening of the funnel seemed small. In the ovary were four yolks above one centimeter and six between a centimeter and a millimeter in diameter. The ovarian yolks were a normal series of maturing yolks, the largest of which was apparently too large to enter the oviduct. The ovary also contained four large follicles. In the body cavity was a creamy oily fluid. This was apparently yolk mixed with serum. There was no appearance of peritonitis; in fact the bird was apparently in perfect health. Bird no. 1375 laid reasonably well during her first two laying years producing 141 eggs the first year and 117 the second but after the second adult molt she laid but one egg (February 14, 1914). On March 7, 1914, she was killed for data. She was thirty-five months old and was very fat. On opening the sheet of fat enclosing the viscera they were seen to be cov- ered with thick fresh egg yolk. The peritoneum was perfectly normal. The oviduct was large and normal in appearance. The finger could easily be inserted into the funnel mouth which was apparently large enough to admit a normal mature yolk. The duct was open throughout. The glandular ridges of the albu- men secreting and isthmus regions were expanded and whitish, as if full of secretion. There was a little albumen between the ridges in the albumen secreting portion. The funnel hung rather loosely from its ligamentary attachments. When the body cavity was opened the ostiwm was situated at some dis- tance behind the largest yolk. The intestines were still capable of normal peristalsis and slight peristaltic movements could be induced in the oviduct. Quantitative data are lacking but the response of the duct seemed less vigorous than the normal re- sponse of the oviducts of laying birds when stimulated imme- diately after death. In the ovary was a large empty follicle apparently just discharged and five others ranging gradually in size from this to one about 1 mm. in diameter. There were also PHYSIOLOGY OF THE OVIDUCT 413 very many which were just distinguishable. The ovary also contained a normal series of maturing yolks. Four of these were above 1 ecm. in diameter. The largest was apparently mature. These two birds then had normal ovaries and at the time of death were maturing and ovulating yolks at a rate com- parable to that shown by birds at the height of a normal period of egg production. They also possessed oviducts which were apparently able to secrete the normal enclosing envelopes. The fact that the yolks did not enter the duct and become the yolks of normal eggs was probably due, in the first case, to the fact that the funnel mouth was too small to admit a full sized yolk, and in the second to a lowered state of tonus in the muscles of the oviduct and oviduct ligaments. They represent a type of non-production previously discussed by one of the authors (Pearl ’12) where sterility is due to ‘somatic’ (physiological) rather than ‘gametic’ causes. From the point of view of the present investigation they show that a bird may ovulate into the body cavity and resorb the yolks from the general peritoneal surface at an astonishingly rapid rate without causing any apparent disturbance of normal metabolism. In the second type of absorption in the body cavity the yolks or eggs are enclosed in separate sheets of peritoneum by which they are attached to the adjacent peritoneal surface. This type of absorption has been several times observed in routine autopsy work as well as in cases 4, 10, 12, 13, 18 and 19. Case 9 shows all the steps in the process. This bird’s oviduct had been ligated at the level of the caudal end of the shell gland before it has enlarged for its first period of egg production. The sex organs developed normally. The duct received the yolks and formed normal eggs which were discharged back through the funnel into the body cavity. At the time of autopsy the body cavity contained five normal and two yolkless eggs, and there were also two normal soft shelled eggs in the oviduct just above the ligature. The condition and position of the eggs in the body cavity were as follows: (1) A collapsed shell, slightly calcareous, containing a small 414 RAYMOND PEARL AND M. R. CURTIS amount of mixed albumen and yolk was surrounded by peri- toneum by which it was firmly attached to the abdominal fat and the right side of the colon at the posterior end of the body cavity. (2) A partly collapsed soft shelled egg, surrounded by peritoneum which attached it to the peritoneum. covering the left kidney, lay between the left kidney and the oviduct just behind the ovary. (3) A soft shelled egg, surrounded by a very thin sheet of peritoneum, was attached to the right side of the intestinal mesentery just opposite the ovary. (4) A normal hard shelled egg occupied the posterior end of the body cavity to the left of the intestine. St 443 GieAscending phases: iiss 2ae to hoes Mare nok e cls Meare ie anita ere 450 IV. Cytological details of series of pedigreed cells in the reorganization | ONKOL OLS] Shen Bee chee sarcsia ek EMORREE ern Ree ae eect ol ieee Hote tae eee Bin Bion a ° 457 AG Series Of cells! tromsbame Wali. cork one tte tsioe eee, Pom one hee cae ne 458 bs seriesyor cellsdrom, ines Titan. ts si. tse ooo alos oo eee ee 469 V. The reorganization process in this race after conjugation, and in other racesvan GespEeclessolebaTrans cClumla yar ee meer an iter 473 VI. The reorganization process and its relation to rhythms.............. 476 VII. The reorganization process and its relation to depression periods..... 481 VIII. Endomixis and its relation to conjugation...... tee tein we ead SL. 489 IX. Endomixis and its relation to parthenogenesis.....................- 492 X. Endomixis and its relations to variation, heredity and the significance Ol COnpeAhLONs, f275 eee Heese to Aer ae eae roar e cota Ge ae tea 494 PEKECEAGUNECILEU RAAT eects ok. i Ses CE: ola oe ot tyne Mua itis pene 498 iB aplamationy Oinplahese ceri. lacks en stereos bose he bial Chee ee ome oe wares 503 I. INTRODUCTION The conception of the Protozoa as primitive animals has led to innumerable studies on these forms—morphologists and phy- siologists alike being animated with the idea that here many of the riddles of life propounded by the so-called higher animals would be presented in a simple form and so be more readily solved. ‘While this view of the Protozoa is undoubtedly true 425 THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 17, No. 4 NOVEMBER, 1914 426 LORANDE LOSS WOODRUFF AND RH. ERDMANN in a broad sense, recent studies on these organisms have only served to emphasize that size and simplicity do not necessarily go hand in hand and that a unicellular or non-cellular animal with all its varied life processes performed within the confines of a protoplasmic unit presents difficulties which differ in kind rather than in degree from those encountered in the metazoa. No better instance of the truth of this could be presented than that afforded by studies on the life history of the Infusoria. Dujardin’s ideas of simplicity—brought forward to combat Ehren- berg’s mistaken interpretations of the complexities which he observed—gradually were displaced by the revelations of Bal- biani, Biitschli, Engelmann, Maupas, Hertwig and others on the complicated phenomena of conjugation, while the results of more recent work on conjugation from the standpoint of dynamics and heredity have only served to emphasize the intricacies of the infusorian life processes. The present study fully describes still another complex nuclear phenomenon! which we have dis- — covered in the life of Paramaecium aurelia and which we believe affords the key to certain apparent contradictions in recent results on the life history of the Infusoria. The problems of protoplasmic senescence and the significance of fertilization have afforded the stimulus for a long series of investigations on the life history of Infusoria since, more than three-quarters of a century ago, the leading students of micro- scopic organisms, Ehrenberg and Dujardin, theorized on the potential immortality of these forms. It remained, however, for Biitschli ('76) and Engelmann (’76) to attack the problem experimentally, in the light of zodlogical advance in the inter- vening years, and to reach the conclusion that continued repro- duction by division results in degeneration and death; while the classic studies by Maupas (’89) on the life history of a number of different species of Infusoria afforded such a wealth of evi- dence pointing in the same direction that conjugation as a sine 1 We have already given brief outlines of this process in the Proc. of the Society for Exper. Biol. and Med., vol. 11, no. 3, February 18, 1914, and in Biol. Cent., _ Bd. 34, 1914. In the latter several errors are present owing to the war prevent- ing a revision of proof. PERIODIC REORGANIZATION IN PARAMAECIUM 427 qua non for the life of the Infusoria seemed to be placed upon a firm empirical basis. A series of careful investigations by R. Hertwig (’00—-04) and Calkins (’02—’04) confirmed Maupas’ general conclusion that In- fusoria after a more or less definite number of divisions degen- erate and finally die if conjugation is prevented. Further, the latter author found that artificial stimuli of different kinds may be substituted with success for conjugation since by the oppor- tune use of artificial stimulation he was able to prolong the life of one culture of Paramaecium caudatum to the 742d gener- ation. This significant discovery that the death of infusorian cul- tures, bred on a more or less constant medium of hay infusion, may be temporarily deferred by artificial means was corroborated by Woodruff (’05). Enriques (’03) studied the same general problem and reached the conclusion that the degeneration and death observed by Calkins and others was due to the presence of bacterial poisons in view of the fact that he succeeded in breeding Glaucoma scintillans for 683 generations without signs of degeneration when this factor was eliminated. Although Enriques’ interpre- tation of the cause of degeneration in Calkins’ cultures is open to question, the significant fact remains that he kept his ani- mals nearly twice as long as did Maupas or Calkins without conjugation or artificial stimulation, thus suggesting that degen- eration is not inevitable, and that if suitable conditions are supplied reproduction by division can proceed indefinitely The problem was attacked from another point of view by Woodruff (1907 to date) who investigated the possibility that the degeneration observed in the previous investigations was induced by too great uniformity in the conditions of culture, or by the culture medium being deficient in something essential for the continued well-being of the organisms. A race of Para- maecium aurelia (I) was isolated and bred on infusions of various materials found in the natural environment of the animal, while a sub-culture was subjected to the relatively constant hay in- fusion culture conditions employed by Calkins (’04). The result was that the cells bred in the constant hay infusion medium died 428 LORANDE LOSS WOODRUFF AND RH. ERDMANN out after a typical Calkins cycle, while those bred ‘on the ‘varied environment’ medium did not pass through periods of marked physiologica' depression or show morphological changes which could be interpreted as abnormal. This race is still (June, 1914) in a normal condition, having attained over 4500 generations without conjugation or the use of artificial stimuli. The success with the varied culture medium naturally led to the question whether the longevity of Paramaecium on a varied environment is dependent upon intrinsic stimuli from the fre- quent changes of the medium, or whether a constant medium of hay infusion is unfavorable because it lacks some elements which are essential for the continued existence of the organism. Ac- cordingly, Woodruff and Baitsell (11 a) bred a sub-culture of this race for a period of nine months on a constant culture medium of beef extract. The continued health of the organisms on this constant medium throughout the experiment, which was continued sufficiently long to include a Calkins cycle if such was inherent, indicated that it is the composition of the medium rather than the changes in the medium which is conducive to the unlimited development of this race without the necessity of conjugation or artificial stimulation. From a study of various species of hypotrichous Infusoria, as well as his main culture of Paramaecium aurelia, Woodruff found that minor periodic rises and falls of the division rate occur from which recovery is autonomous. He termed these fluctuations ‘rhythms’ and contrasted them with the so-called cycle, which comprises a varying number of rhythms and, ac- cording to Maupas and Calkins, ends in the death of the race if conjugation or artificial stimulation is not resorted to. The problem of the rhythms was studied intensively by Woodruff and Baitsell (’11 b) who showed that when Paramaecium aurelia is subjected to the most constant environmental conditions it is impossible to eliminate the rhythm and thus resolve the graph of the multiplication rate into an approximately straight line. Accordingly there are inherent rhythmical changes in the phe- nomena of the cell which produce slight changes of the divis on rate. The. same result was reached by Woodruff and Baitsell PERIODIC REORGANIZATION IN PARAMAECIUM 429 (llc) from a study of the temperature coefficient of the rate of reproduction of this culture which showed that the rate of cell division of Paramaecium is influenced by the temperature at a velocity similar to that for a chemical reaction—except when the rhythms interfere. The results from the study of this pedigreed race of Para- maecium aurelia have led Woodruff to conclude that this organ- ism, when subjected to suitable culture conditions, has the power of unlimited reproduction by division without conjugation or artificial stimulation—the only necessary variations in the rate of reproduction being the normal minor periodic rise and fall of the division rate (rhythm), due to some unknown factor in cell phenomena, from which recovery is autonomous. Calkins ang ‘Greeory: 1s. 9p 507))" 2 5 a dommot share his optimism, however, and can only say that while his results are remarkable, his race is not yet dead. Is there any clue to the particular make up of this race of Paramaecium aurelia?’’ Accordingly Calkins sought the explanation of the diametrically opposite results derived from his and from Wood- ‘ruff’s cultures of Paramaecium in variations in the tendency to conjugate which have been observed by Jennings (’10) and him- self to exist in different races of this organism. Thus he empha- sized the fact that he could readily induce conjugation in his culture whereas experiments. to secure conjugation in Woodruff’s culture were without effect. Calkins, therefore, stated that ‘‘the two races cannot be compared in regard to vitality, since normal conjugation was prevented in the conjugating race, whereas in the non-conjugating race there has been no artificial prevention of a normal process.’’ ‘‘ Woodruff’s Paramaecium aurelia is evi- dently a Paramaecium Methuselah belonging to a non-conju- gating line the life history of which is not known in any case.’’ However, conjugation was finally secured in a mass culture seeded from Woodruff’s race (’14) thus demonstrating that this race is a conjugating race when the proper conditions for its consummation are realized. Therefore, there is no evidence extant that a non-conjugating race of Paramaecium exists. 430 LORANDE LOSS WOODRUFF AND RH. ERDMANN With this theory swept away by a fact the results derived from this culture stand where they stood before and demonstrate that the very limited periods in which Maupas, Calkins, and others observed degeneration have no significance for the ques- tion as to whether degeneration and death are inevitable results of reproduction without conjugation. In other words, this one positive result from Woodruff’s race outweighs all the negative evidence derived from work on the Infusoria, and justifies the statement that these organisms can live indefinitely, when sub- jected to favorable environmental conditions, without conjuga- tion or artificial stimulation. Although morphological or physiological variations winch could be interpreted as the result of degeneration were never observed in Woodruff’s race of Paramaecium, it was early noted (’08) ‘‘that various nuclear changes which are not at present recog- nized occur normally in the life history of Paramaecium,”’ and suggested that possibly when conjugation is prevented a reorgan- ization of the nuclear apparatus within the individual cell occurs. Erdmann (’08) independently reached an essentially similar view from a consideration of the published data on this culture and a critical study of infusorian life histories, and further, in an experimental study of Amoeba diploidea (713), suggested that a relation exists between sexual phenomena and rhythms. Accord- ingly the present authors have collaborated in a study of the daily cytological changes of this race of Paramaecium aurelia (1). As a result of our study of a large series of animals preserved daily during the past half year and of specimens preserved at various periods throughout the existence of this culture of Para- maecium, we have discovered that the rhythms in the division rate are the physiological expression of internal phenomena which involve the formation of a complete new nuclear apparatus, by a definite sequence of normal morphological changes which simulate conjugation. This nuclear reorganization, in essence, consists of a gradual disintegration and absorption of the macro- nucleus in the cytoplasm. Simultaneously a multiplication of the micronuclei is in progress. Certain of the resulting micro- nuclei degenerate while the remaining one or two form the new Text fig. 1 Diagrammatic survey of the reorganization process. Descending phase, 1-5; climax, 6-9; ascending phase, 10-14. Dot in circle = micronucleus; isolated dot = degenerating micronucleus; lines and dots = macronucleus; stip- ple = chromatin body; crosses = macronuclear anlage. macronuclear and micronuclear apparatus. This results in the reorganization of the cell without the fusion of two animals (text fig. 1). | 431 : + 432° LORANDE LOSS WOODRUFF AND RH. ERDMANN II. MATERIAL AND METHODS A. METHODS OF CONDUCTION OF THE MAIN CULTURE I Since the results presented in the present paper are based on a cytological study of the main pedigreed race (I) which has been carried on for the past seven years, a brief outline of the methods employed in its conduction must be given. For further details the reader is referred to earlier papers on this race (Wood- rat. 408; (11); A specimen of Paramaecium aurelia was isolated from a labo- ratory aquarium on May 1, 1907, and placed ‘n about five drops of hay infusion on an ordinary glass slide having a central de-— pression. When the animal had divided twice, producing four individuals, each of these was isolated in fresh media on a sepa- rate slide, thus forming the four main lines (Ia, Ib, Ie, Id) of this culture, Paramaecium aurelia I. This pedigreed race has been maintained by the isolation of a specimen from each of the four lines practically every day during the more than seven years of its existence. The isolations have been made with a capillary pipet and a Zeiss binocular microscope, oculars 2, ob- jectives 55a. This daily isolation has prevented the possibility of conjugation occurring, afforded fresh culture medium, and enabled an accurate record of the generations to be kept. There- fore this is a pedigreed culture of Paramaecium. The rate of division of the culture has been determined by averaging the daily rate of these four lines and the accompanying graphs show this again averaged for various periods as indicated in the respec- tive legends (text fig. 2). The culture material supplied to the main lines of the race has consisted of thoroughly boiled infusions of materials taken practically at random from ponds, swamps, etc., in an endeavor to supply the general type of material ordinarily met with by this organism in nature. The depression slides holding the ani- mals have been kept in a glass moist chamber on a laboratory table and therefore the organisms have been subjected to the ordinary fluctuations of temperature, light, etc., of the room. The morphological condition of the animals has been followed (‘smyphys ypu pasnfuod aq JOU JsnU aLofatay}] puv SUOYYPUOD ainzjna paripa ay) fo qjnsas ay} fizfarya aap aasna ay} ut suoyonjony ay]) “paurezye atom AdYy YOTYA ul spotted ay} 9AOQeB poovid o1v pue suor}eIBUS quasoidal ‘049 ‘QOOT ‘00¢ SeInSYy oY], ‘“SIvaA IVpUdT[VO BY} JO SJIVMT] AY} O}PBOIPUL SOUT] USYOIg [BOI}IOA OYJ, °O9BP 0} OJI] SI Jo yYJWow Yow IOJ PIBVIOAV UIVSE IIN}[ND IY} JO SOUT] INO} VY} JO UOISIAIP JO 94¥I A[IVP OSBIOAL OY} JUASOIdaI SayVUIPIO VY J, “WOIB10Ues YIFEEPF 949 4e ‘FIGT ‘T Avy 0} ‘2061 ‘1 AB] GO 4148 WIOIJ VITOING UMIDBBVUIBIEY JO (J) 9DBI padisIpod oY} Jo UOTJONpoOAdo. Jo 9381 OY} SuIMOYsS Yduinn 7 -SYy 4yxoz, Elél Z161 GL OBL 6061 8061 LO6t ra) S LL AD ere | (U1 nb OOO0F 00S OOOE 00S2 0002 OosL Ooo! 00S ' ' 1 U i t ' 1 ' ' ' | i -<<-<-—- 434 LORANDE LOSS WOODRUFF AND RH. ERDMANN in a general way by periodically preserving specimens left over at the daily isolations. The culture was started and conducted for two months at the Biological Laboratory of Williams College, and since has been carried on at the Biological Laboratories of Yale University, except during a part of each summer when it was transferred to the Marine Biological Laboratory at Woods Hole, Mass. It is under these conditions then that this culture (I) has been successfully maintained without loss of vigor for over 4500 gen- erations—all the experiments which have been conducted on animals from this race having been made on sub-cultures started from the animals left over from the lines of this culture at the daily isolations. B. METHOD OF CONDUCTION OF SUB-CULTURE IE For the intensive study of the daily cytological changes of this race of Paramaecium aurelia, on October 27, 1913, six ani- mals left over from Line Ia of the main culture at the 4020th generation were isolated to start a new sub-culture, designated IE, of six lines. This sub-culture was subjected to practically constant environmental conditions. The culture medium was the 0.025 per cent beef extract which Woodruff and Baitsell (11a) found to be a most favorable medium for this race of Paramaecium. The sub-culture was kept in a thermostat set for 26°C., and such variations (about 1°C.) from this as occurred were well within the optimum zone for the animals of this race as determined by Woodruff and Baitsell (11 ¢). An animal was isolated every day from each line of this sub- culture from its initiation to April 27, 1914—a period of six months. The isolated animal from each line was placed in fresh culture medium, while one or more of the remaining animals was preserved for study. Thus for half-a-year permanent prepara- tions were made from day to day of sister cells, the exact ancestry of which was known in every case. It is important to emphasize the fact that by this method only, 1.e., the study of practically each cell generation, could the sequence of nuclear changes which we describe be determined. This became evident very soon as the PERIODIC REORGANIZATION IN PARAMAECIUM 435 work proceeded and consequently the study was concentrated particularly on one (VI) of the six lines—the other lines being merely controlled to show the occurrence of the process, at which time all the cells available from these lines were also preserved. In many cases all the animals of certain lines were killed in the reorganization process so that every animal could be studied. The places of these lines were supplied by new ones from one of the remaining five lines. Thus by killing of various lines and the branching of others to take their places a total of fifteen sub-lines of longer or shorter duration were formed. These to- gether with the six original lines made a total of twenty-one lines whose cytology was investigated.” The individual animals were fixed in Schaudinn’s sublimate- alcohol (stronger solution), stained with Delafield’s hematoxylin and mounted in cedar oil—the specimen being watched under a Zeiss binocular microscope throughout the operations, and trans- ferred with a capillary pipet from one depression slide to another as occasion demanded. Differentiation was effected with acidulated alcohol (70 per cent alcohol + 0.002 per cent HCl) under a com- pound microscope. The following staining reagents were tried but the above methods gave the best results—bearing in mind that each animal had to be carried along under constant observation through each of the fluids until it was in the cedar oil undera coverglass: Heidenhain’s hematoxylin and Bordeaux red, methyl green and eosin; Delafield’s hematoxylin and eosin; safranin and methyl green; Mannsche Farbung and Giemsa feucht. Animals preserved in bulk from mass cultures were submitted to the same staining methods and again Delafield and eosin were found most satisfactory; Heidenhain staining the trichocysts too greatly. Sections (5u) were made of animals in various stages of the proc- ess but they did not afford important details which could not be made out in the total mounts. The sections were stained with safranin and Heidenhain according to the suggestions of En- riques (712). 2 After the definitive experiments were formally concluded at the end of six months, certain lines were continued to June 14, 1914, in order to secure some further details. 436 LORANDE LOSS WOODRUFF AND RH. ERDMANN In addition to this material preserved daily from Sub-culture IE, we had at our disposal, as already mentioned, preparations preserved at various periods throughout the life of the main culture (I). These had been fixed in a saturated solution of corrosive sublimate with 5 per cent glacial acetic acid, stained with Ranvier’s picrocarmine and mounted in damar. III. DESCRIPTION OF THE CYTOLOGICAL CHANGES IN THE REORGANIZATION PROCESS The general outline of this remarkable cytological process which accompanies the rhythms clearly shows that a complete internal reorganization of the Paramaecium cell occurs without cell fusion. The details to be presented were obtained as already described from Sub-culture IH, from October 27, 1913, to April 27, 1914, and certain stages were substantiated with specimens which had been preserved from the main culture (I) at various isolated periods during the previous six and one-half years of its existence. Figure 1 (pl. 1) shows a typical specimen of Paramaecium aurelia isolated from the main culture (I) on April 11, 1908, at the 424th generation, i.e., 3596 generations before Sub-culture»IE was started. Figure 2 (pl. 1) represents an animal in the 4020th generation at the time Sub-culture IE was branched from the main Culture I. Pay The reorganization process resolves .itself naturally into three periods: the descending phase, the climax, and the ascending phase. A. DESCENDING PHASE The macronucleus of a typical Paramaecium aurelia, which is not undergoing the reorganization process, consists of fine chromatic granules enclosed within a relatively thick membrane. The finer structure has been fully described by.Maupas (’89, p. 217) and Hertwig (’89, p. 9). The two micronuclei, in a simi- lar period with respect to the process and in the resting stage between two cell divisions, are more or less compact and homo- geneous, and often lying in a pair near the macronucleus which sometimes obscures them from view. ‘The cytoplasm of the cell PERIODIC REORGANIZATION IN PARAMAECIUM 437 contains many small vacuoles some of which may be filled with bacteria (figs. 1 and 3, pl. 1). As the low point of the division rate approaches, the granules of the macronucleus become more and more coarse, its staining capacity increases, its form becomes more kidney-shaped and its shorter axis elongates. Still more obvious changes of the macro- , nucleus indicate the actual beginning of the regulation process. 1. The macronucleus At this stage projections appear at the end of the macro- nucleus (figs. 4 and 5, pl. 1) which are either merely thin membranes devoid of chromatin or filled with small granules. Hertwig (’89, p. 7) describes them as follows: “gleichzeitig verliert die Oberfliche ihr glattes Aussehen; Einkerbungen erstrecken sich mehr oder minder tief in das Innere hinein und zerlegen den ‘Kern nicht selten in drei ungleich grosse Lappen oder es werden an den Enden fingerférmige Fortsitze deutlich oder leisten-und riffartige Vorspriinge.”’ According to this author they are not evident during the latest stages of vegetative cell division, or in the first changes incident to conjugation. Though Hertwig could not observe these projections in the living animal he did not believe them to be artifacts because he could demonstrate them with every fixation fluid. We have never observed them in vegetative cell divisions but frequently have noted their presence several generations before the definite onset of the process. We have no data in regard to their relation to conjugation. Figure 5 (pl. 1) gives a good idea of these projections. The animal figured is from Line VI, 4094th generation, and is nine generations before the beginning of the reorganization process which started in the 4103d generation. These long finger-like projections are preceded by the appearance of smaller ones (fig. 4, pl. 1) which seem to indicate that plasmatic currents are present in the macronucleus though the membrane at this stage is still intact. — 438 LORANDE LOSS WOODRUFF AND RH. ERDMANN The next stage which is peculiar to the process is the separation of chromatin bodies from the macronucleus. In the early stages of the reorganization process cells are seen with only two or more chromatin bodies (figs. 6 and 7, pl. 1) which are more or less spherical and consist of large and small granules. Figure 8 (pl. 1) shows a macronucleus which is surrounded by several of these bodies which have been ejected from it. This method of nuclear disintegration is not unique in Infusoria because similar morphological phenomena are described by Neresheimer and Buschkiel. Neresheimer (’08) mentions that chromatin bodies are ejected from the macronucleus of Ichthyopthirius before the beginning of the sexual process. This species during vegetative life has its macronucleus and micronucleus in one body, but this difference is unimportant because the chromatin bodies of Ich- thyropthirius contain the combined material of the macronucleus and micronucleus, and at the time of extrusion the vegetative or sexual character of the chromatin bodies is determined. In Paramaecium aurelia merely vegetative nuclear material is ex- truded by the macronucleus. A connection between the macronucleus and the chromatin bodies of Paramaecium as described by Buschkiel in Ichthyop- thirius (’08, p. 81) could not be discovered and seems not to exist because these bodies are formed in the macronucleus and arerejected through openings in the membrane. The formation of the chromatin bodies begins in the macronucleus and is evi- dent by the condensation of granular material at various points within the membrane (fig. 8, pl. 1), other parts of the macro- nucleus becoming devoid of chromatin as segregation proceeds. The chromatin bodies are not yet surrounded by clear areas. The resulting membranous condition of the old macronucleus is shown in figures 6, 10 and 11 (pl. 1) and figure 14 (pl. 2). Two characteristic stages in the elimination of chromatin bodies by the macronucleus are given in figures 8 and 9 (pl. 1). The chromatin bodies are removed from their place of origin by movements of the cytoplasm, thus emphasizing the well- known cytoplasmic currents in the cell. The specimen shown PERIODIC REORGANIZATION IN PARAMAECIUM 439 in figure 10 (pl. 1) has seven of these bodies arranged in a semi- circle at the posterior end of the cell. The anterior end contains some irregularly shaped chromatin bodies which seemed to be undergoing involution, while three others more recently ejected are still near the macronucleus. The disintegration of the macro- nucleus progresses until it is entirely devoid of chromatin (figs. 35, 36, pl. 4) while the wrinkled and ruptured membrane with but slight staining capacity remains in the cell (fig. 14, pl. 2.) In the later period of the reorganization process the membrane is resorbed and the old macronucleus has finished its function in the paramaecium cell. These changes of the macronucleus have their analogy in normal conjugation. There the old macronucleus is destroyed but this is effected by the formation of the so-called ‘‘wurst- formige Schlingen,’’ which form a tangled mass of chromatin ribbons. Their origin is figured by Maupas (fig. 10, pl. 12; figs. 17-20, pl. 18); and by Hertwig (figs. 6-9, pl. 1; figs. 1-9, pl. 2) and in the present paper (fig. 44, pl. 4). However, in conjugation it is not until after the animals have separated that the chro- matin ribbons are totally fragmented and more or less spherical bodies are free in the cell. Later an involution of these occurs, the details of which are not described by either Hertwig or Maupas. Likewise, the persistence of the membrane of the macronucleus is not figured by these authors. This ribbon-like formation of the macronucleus we have found to be characteristic of conjugating animals from this race of Paramaecium aurelia (fig. 44, pl. 4). Figure 32 (pl. 3) shows a pair of conjugants with their macro- nuclei in this form. In this period then the differences between the macronuclear changes during conjugation and during the process are only mor- phological; on the one hand, the macronucleus forms ‘wurst- formige Schlingen;’ on the other the macronucleus eliminates its chromatin by extruding it in the form of spherical bodies. The physiological effect is the same. The result is the destruction of the old macronucleus. 440 LORANDE LOSS WOODRUFF AND RH. ERDMANN 2. The micronucleus While these changes are taking place in the macronucleus the micronuclei do not remain unaltered. They move from their accustomed position, which is more or less close to the macronucleus, and migrate in the cytoplasm (fig. 3, pl. 1). The same appearance was noted in non-conjugating Para- maecium caudatum by Calkins and Cull (07, p. 383), who state that in stages of a pedigreed race, which Calkins calls de- pression periods, the micronucleus migrates in the cytoplasm, and they interpret this as an abnormal condition. In conju- gating animals the migration of the micronuclei at the beginning of the sexual process is described in Paramaecium aurelia by Maupas (’89, p. 212) and by Hertwig (’89, p. 21); in Paramae- cium caudatum by Calkins and Cull (07, p. 383); and in Para- maecium bursaria by Hamburger (’04, p. 200). This migration seems to be a general phenomenon at the onset of the changes incident to conjugation and occurs also, as described above, in isolated paramaecia before the beginning of the reorganization process. . The micronuclei in their new position proceed to divide with the result that finally eight are present in the cell. Figure 14 (pl. 2) shows a specimen in which there are two solitary micro- nuclei and two groups of three micronuclei. All the micronuclei show a perfectly normal structure but some variation in size is evident. In typical conjugating animals the eight so-called re- duction micronuclei are described by Hertwig and Maupas as all being of the same size, but Maupas did not lay great stress on cytological details so that only Hertwig’s statement is of weight. It is probable that the differences of volume shown by micronuclei in the process are due to variations in the amount of nuclear sap preliminary to the cessation of their functional activity. It is impossible to determine whether these micronuclei of the animals in the process are actually reduction micronuclei because, here, as in the conjugation of Paramaecium aurelia, it is not possible to count the number of chromosomes. PERIODIC REORGANIZATION IN PARAMAECIUM 441 The formation of the eight micronuclei does not begin until the disintegration of the macronucleus is about finished. Figures 36 and 37 (pl. 4) illustrate the coincidence of the end of the macronuclear disintegration and the beginning of the multipli- cation of the micronuclei. The animals in figures 35 and 36 show the typical form of the macronuclear destruction which takes place in the reorganization process without the formation of ‘wurstférmige Schlingen’. A slight resemblance to the char- acteristic macronuclear condition during conjugation is given in the isolated paramaecium of the 4087th generation, thus indi- cating that the macronuclear destruction in conjugation and in the reorganization process may sometimes show similar features. The animals shown in figures 36 and 37 possess either three or four ‘reduction’ micronuclei. The identification of these’as ‘reduc- tion’ micronuclei is based upon their position in a homogeneous protoplasmic layer, and the absence of a micronuclear membrane, together with the general morphological structure of the cell and its fate. The diagnostic characters of the ‘reduction’ micronuclei are still more prominent in later stages. Figure 11 (pl. 1) shows a macronucleus partly devoid of chromatin, eight chromatin bodies and three micronuclei. Two of the micronuclei are apparently starting to divide. The anterior micronucleus seems to be in the process of forming another by an unequal distribution of the chromatin inside the membrane. This is the only micronuclear change of this type which we have observed and may well be due to some irregularity. The formation of mitotic spindles with long distinct threads and crescents have not been seen, and apparently exactly the same type of mitosis as described by Hertwig does not occur in the reorganization process. However, figure 31 (pl. 3) illustrates an animal, from a sub-culture of this race in which conjugation was allowed to occur, having division spindles which approximate to a certain degree some of the spindles figured by Hertwig. It is well known that the mor- phology of the mitotic apparatus varies at different phases of cell life, as for example was found to be the case in Amoeba diploidea by Erdmann (’11, p. 336). THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 17, NO. 4 442 LORANDE LOSS WOODRUFF AND RH. ERDMANN Calkins and Cull (’07, p. 383) note that, during periods which they call depressions, the typical structure of the micronucleus of Paramaecium caudatum is lost and it becomes abnormally large with the chromatin in a loosely granular condition. It is a remarkable coincidence that this occurs also in Paramaecium aurelia in the descending phase of the process. Figure 12 (pl. 1) shows the macronucleus, which now has become smaller, together with fourteen chromatin bodies free in the cytoplasm, and two just leaving the macronucleus. The cell possesses three micronuclei: two, lying near each other in a cytoplasmic layer free from granules, are in early division stages, while the third has nearly finished division. Thus the cell soon would have contained six micronuclei. Though four micronuclei are formed in normal vegetative division the begin- ning of the ‘reduction’ division is distinguishable from that of vegetative division because the four vegetative micronuclei are usually either lying in closely associated or widely separated pairs in the cell (fig. 7, pl. 1) while the ‘reduction’ micronuclei remain clustered together in a layer of homogeneous cytoplasm. Later stages showing seven and eight micronuclei are given in figure 13, pl. 1, and figure 14, pl. 2, already described. Thus at the end of the descending period of the reorganization process the paramaecium cell possesses eight micronuclei, a shrunken membrane of the old macronucleus, and numerous chromatin bodies which have passed out from the macronucleus (fig. 14, ply 2): The important multiplication of micronuclei is not without precedent in non-conjugating animals. Maupas mentions a non- dividing animal with three micronuclei. Popoff (09, p. 30) describes a specimen of Paramaecium caudatum with two micro- nuclei which are just preparing to divide and form four without cell division ensuing. Also Kasanzeff and Calkins note the multiplication of the micronuclei in Paramaecium caudatum with- out cell division. Popoff states further that in Stylonychia mytilus, after treatment with a medium with CO, in solution, the normal four micronuclei (p. 15) multiply until eight may be present. It is unnecessary at this point to discuss the theoretical PERIODIC REORGANIZATION IN PARAMAECIUM 443 bearings of these cases (cf. p. 483) but merely to emphasize the fact that the multiplication of the micronuclei in an isolated cell is not an infrequent event, and that the fate of the cell alone gives the significance of the phenomena. ¢ B. CLIMAX The period of the reorganization process, designated the climax, is the most important of the three phases. Morphologically defined it extends from the total disintegration of the old macro- nucleus to the formation of the new macronuclear anlagen. Phys- iologically speaking it represents a stage in which the vegetative functions of the cell are relatively in abeyance and only the potentialities of the micronuclei are in evidence. At the end of the descending phase the paramaecium cell possesses eight micronuclei, twenty to thirty spherical chromatin bodies and the membranous remains of the old macronucleus (fig. 14, pl. 2). Figure 13 (pl. 1) gives an animal just as the degeneration of the micronuclei sets in. It shows six micronuclei, one of which is dividing, scattered in the cell, and numerous chromatin bodies but no trace either of the old macronucleus or of the macronuclear anlagen. Four micronuclei are closer together than the other three, indicating that the two groups arise from each of the two original vegetative micronuclei. The scattered posi- tion of the micronuclei indicates the beginning of their degen- eration (see fig. 44, pl. 4) which shows the same feature in a conjugating animal. The next change consists in the practically complete disappear- ance of the old macronuclear membrane and in the degeneration of the so-called reduction micronuclei to the number two (or one). It was not possible to trace the method of degeneration of the micronuclei, but it is positive that in the generation which follows the above described stage only one or two are present. Maupas and Hertwig figured the degenerating micronuclei in conjugating animals as homogeneous minute dots the structure of which is not discernible (Hertwig ’89, fig. 7, pl. 1; Maupas ’89, fig. 13, pl. 12). Hertwig at first could not determine 444 LORANDE LOSS WOODRUFF AND RH. ERDMANN whether one or two of the reduction micronuclei remained and formed immediately without a further division the stationary and migratory nucleus, but in his final publication (’89, p. 31) he states definitely that seven reduction micronuclei degenerate but his figures do not absolutely prove this point (fig. 6, pl. 1; fig. 24, pl. 4). A micronucleus remains which is the source of the new micronuclear and macronuclear apparatus. Maupas (89, p. 219) stated earlier than Hertwig that ‘‘Sept d’entre eux vont, en effet, passer a l’état de corpuscles de rebut et dis- paraitre en serésorbant. Un seul persistera et continuera l’évolu- tion fécondatrice,’’ but as already mentioned, he does not give cytological details so that a perfect description of the degener- ation of the micronuclei in Paramaecium aurelia does not exist. It is easier to trace the degenerating micronuclei in conjugating animals because there the ribbon-like formation of the macro- nucleus is still intact (fig. 44, pl. 4). Many chromatin bodies are not free in the cell and consequently cannot be mistaken for degenerating micronuclei. Thus each tiny homogeneous gran- ule in the cytoplasm can be identified as a micronuclear remnant, whereas in the reorganization process unconnected chromatin bodies are scattered freely in the cell and minute homogeneous granules intermingled with them may or may not be identified as degenerating micronuclei. However, certain bodies which have been observed we are inclined to interpret as micronuclear remnants. These will be indicated in the description of the plates. The reorganization process proceeds in two different ways, as shown by the animals which were preserved at the climax, depending upon whether or not a cell division takes place at this period. The usual method is for a cell division to occur. In the cell which possesses no macronucleus the chromatin bodies change their shape, becoming more and more elliptical while each granule begins to migrate to the surface so that the bodies appear hollow. The one remaining micronucleus is large. Fig- ure 15 (pl. 2) shows such cells which have just completed the division at the climax. Each has one micronucleus while the anterior cell has eight and the posterior twelve chromatin bodies. PERIODIC REORGANIZATION IN PARAMAECIUM 445 The posterior animal still contains a piece of the old macronuclear membrane. It is possible that the two micronuclei in this divid- ing animal (fig. 15, pl. 2) are the result of a division of a single ‘reduction’ micronucleus which alone remained in the parent cell. It is also possible that both of these micronuclei are the ‘reduction’ micronuclei which did not degenerate and which are now, after a long period of passivity, distributed to each of the daughter cells. In this case the formation of a new spindle could not have occurred and the absence of the micronuclear division would be ' proof that the process is effected in this way. In eight cases of the process the cytological study has shown that a cell division actually does occur. Such cases fall under the heading Ia and Ib in text figure 3. The discussion of the micronuclear events preceding the cell division in the climax will be postponed until all the cytological data from the study of the climax have been presented. The morphology of the cell after the cell division in the climax was definitely determined from several specimens (see fig. 15, pl. 2, and figs. 41 and 40, pl. 4). Clearly the reorganization of the nuclear apparatus is effected by a single micronucleus in all cases in which a cell division occurs in the climax. Figures 41 and 42 (pl. 4) give two cells in the 4436th and 4437th generations respectively. Cell 4436 has one micronucleus and numerous chromatin bodies and is an animal after the cell division in the climax. Cell 4437, which is a product of the first division of the sister cell of cell 4436 after about forty-eight hours, already has one macronuclear anlage. This proves that, after the discussed cell division in the climax, the one micronucleus forms, by two subsequent divisions, four micronuclei, two of which become macronuclear anlagen which are distributed by the succeeding cell division to each of the two daughter cells. In figure 16a (pl. 2) is given a more advanced stage of one of the cells immediately after division. This anterior cell - (a) has just emerged from the climax, as is evident from the fact that the formation of the macronuclear anlagen is completed. The sister cell (b) has not undergone the same changes and one micronucleus and several chromatin bodies, of which only five 446 LORANDE LOSS WOODRUFF AND RH, ERDMANN at the anterior end are figured, can be detected. Figure 39 (pl. 4) shows essentially the same stages, the only difference being that the anterior cell has effected one more micronuclear division before the first reconstruction cell division than lI6a. While the morphological changes which occur after the cell division at the climax are clear, those which take place before this division (i.e., from the formation of eight ‘reduction’ micro- nuclei to the distribution of two to sister cells) must be dis- cussed further. In the instances where two micronuclei remain, it is possible that a micronuclear division has not taken place, but that merely a shifting of one of the micronuclei into each of the two newly formed sister: cells has occurred. This method of distribution, which. gives an equally satisfactory explanation of the observed data, we call Case Ib (text fig. 3). The results of cases Ia and Ib are the same—two cells with one micronucleus and many chromatin bodies at the climax. Only twice during these experiments has a paramaecium been observed which had a macronucleus half devoid of chromatin, two well formed micronuclei, two macronuclear anlagen and several chromatin bodies. In these cases the latter were so small and homogeneous that they readily could be mistaken for ‘reduction’ micronuclei. Here it is evident that no division during the climax took place and that the formation of the anlagen has occurred preco¢iously. There are thus two possi- bilities as to the origin of this condition: (a) One so-called reduc- tion micronucleus may remain and divide. The two micro- nuclei thus arising form the anlagen (text fig. 3, IIa). (b) Two so-called reduction micronuclei may remain and divide, thus form- ing immediately the macronuclear anlagen (text fig. 3, IIb). The result in either case is the same—one cell with two micro- nuclei and two macronuclear anlagen. Data will later be presented (cf. p. 495) which afford strong physiological evidence, that a third micronuclear division produc- ing gametic micronuclei (stationary and migratory) must be absent. It is important to note that if no micronuclear division takes place between the last so-called reduction division andthe divisions which form the new micronuclei and give rise to the oo oO (0) (- VY qt . “ssoooid Uoryeziuvs1001 oY} JO Neu] oY} 4e UOTSTAIp [Jeo pus IVITONUOIOIUI JO Spoyys|, & “BY 4xoq, ©] [© @} /© @) (@ Oe) QO) |© ©] |©} [O} [© og \OO OO; \OG OO (oxo) OO} \CO OO, OO OO OO (ote) (oXo) %q Ig 2p ‘D tp lp % lg 2p tp ag hg ey ty ) © 9@ y ®: :) = i sH ie OQ —_ ssod0rd u01VezIue3.10077 — uo1lyesnfuog 448 LORANDE LOSS WOODRUFF AND RH. ERDMANN macronuclear anlagen (Ib, IIa, IIb), a primary distinction between the morphological phenomena in conjugation and in the process here described is established, i.e., the important micronuclear division which forms the migratory micronucleus and the sta- tionary micronucleus is absent. But we wish to emphasize the fact that if a third micronuclear division occurs, it can equally well be a precocious division in the ascending or re- construction phase of the reorganization process. The diagram presented herewith (text fig. 3) summarizes the different cases which are possible from the observed facts. All the described cases are distinguished from conjugation and autogamy by the absence of the formation of a syncaryon. In the process only in case Ia already described is it possible that a third micronuclear division at the climax takes place. The difference between conjugation and the process in case Ia consists in the occurrence of one more cell division in the latter than in conju- gation. The possible variations in cases Ib, IIa and IIb agree -in the absence of the third micronuclear division under discussion. Cases Ila and IIb have no cell division before the formation of the macronuclear anlagen, one or two so-called reduction micro- nuclei remaining in the same old cell. Case IIb is peculiar in the shifting of one micronucleus into each new cell. But from a study of all the observed sequences of generations at this period there is some evidence that case Ib presents the usual method of nuclear changes at this period. Line VId had four and Line VIb had three animals in the reorganization process, all of which were at the first stage of the ascending phase—for- mation of the macronuclear anlagen. This one fact is crucial evidence that a cell division has occurred before the formation of the macronuclear anlagen, but no direct observation can be pre- sented which determines whether the discussed micronuclear division has or has not taken place, because one spindle in a cell filled with chromatin bodies without a macronucleus or with a macronuclear membrane devoid of chromatin can represent either: (a) the discussed third micronuclear division or (b) the division which forms in cases Ia, Ib, and Ila the two new micro- nuclei which in the next division form the macronuclear anlagen. PERIODIC REORGANIZATION IN PARAMAECIUM 449 Figure 40 (pl. 4) shows an actual cell to illustrate these state- ments. The animal has many chromatin bodies and two micro- nuclei which have just divided. Only because the fate of the cell (VIh, 4355th generation) is known, is it certain that these two micronuclei are the two micronuclei which, by another micro- nuclear division, form four micronuclei, two of which become macronuclear anlagen. If the cytology of the cells related to this cell were not known, its two micronuclei could be interpreted as the products of the discussed third division, which, if it occurs, should take place before the cell division in the climax. We have in our pedigreed lines no indication of its occurrence. Only on the assumption that this important division in the climax has different morphological structures from the other divisions of the reorganization process would it be possible to recognize it. However since we have only seen small elongated spindles during the ascending phase there is no reliable morpho- logical criterion to determine the presence of the third micro- nuclear division. One might expect that the condition of the chromatin bodies would give some indication of the extent to which the process has advanced and therefore of the number of micronuclear divisions which has occurred. But this is not a positive criterion because the disintegration of the chromatin bodies does not progress with equal rapidity in all cells. Weare certain, from the combined evidence from our cytolog- ical study and from physiological data which will be presented later, that the micronuclear division which would be comparable to that which forms the stationary and migratory micronuclei in conjugating animals is absent in the reorganization process. In conjugation the reorganization of the nuclear apparatus is consummated by the synearyon. In the reorganization process the so-called reduction micronuclei or their descendants give rise to the new nuclear apparatus. This significant feature is obviously of great importance from the standpoint of the theo- retical interpretation of the reorganization process which we describe. A brief survey of all the stages from the climax which we have figured in the plates substantiate, we believe, the cytological 450 LORANDE LOSS WOODRUFF AND RH. ERDMANN data which we present and our interpretations of them. The third micronuclear division is absent; two ‘reduction’ micronuclei remain which are shifted into two cells. Figures 15, 16 and 39 show the cell division in the climax; figure 41 gives a single cell which has completed that division; figure 40 shows the next micronuclear division preceding the formation of the macro- nuclear anlagen; figure 16, anterior cell, figures 17 through 24, and figure 39, anterior cell, give the formation of the two macro- nuclear anlagen and the micronuclear changes in each stage. The distribution of the macronuclear anlagen is shown in figures 25, 42, and 43. C. ASCENDING PHASE The ascending phase of the reorganization process, which is the longest of the three, extends from the formation of the macro- nuclear anlagen to the restoration of the typical paramaecium cell with one macronucleus and two micronuclei. It is identical with the periods F and G which Maupas describes as the first divisions after typical conjugation in Paramaecium aurelia (’89, p. 221). Two micronuclear divisions occur in a very short time, half the products of the second forming the two macronuclear anlagen. The two untransformed micronuclei divide again and the first cell division ensues. The following divisions of the cell are exactly similar to typical vegetative divisions and can only be distin- guished by the fact that the involution of the chromatin bodies is In progress. The details of this period are remarkable in various ways. Figures 17, 18, and 19 (pl. 2) give details of the formation of the macronuclear anlagen. The chromatin bodies are omitted from the drawings. These three preparations are counter-stained with eosin, which, according to Calkins (’07, p. 383), stains the non-chromatic parts of the micronuclei of Paramaecium cau- datum. The posterior micronucleus (fig. 17) has just divided and of the products of this division one remains a typical micro- nucleus while the other shows the beginning of the development of a micronucleus into the fundament of the macronucleus. PERIODIC REORGANIZATION IN PARAMAECIUM 451 Chromatin has just separated from the achromatic threads but still retains its micronuclear structure. The next step in form- ing the anlagen is the distribution of the threads and the arrang- ing of the coarser chromatic granules underneath the membrane. This step has been effected in the anterior anlage shown in the same figure. Figure 18 gives a clear idea of the completed an- lagen and the two micronuclei from which they arose. Here the granules have become smaller and the plasmatic character of the structure is more evident because its staining capacity for eosin has increased. The micronuclei show the chromatin surrounded by a plasmatic layer highly stained with eosin. In figure 19, the next division, which is the first in the new rhyth- mical period of the life history, is completed in one pair of micro- nuclei and is in progress in the third micronucleus which is lying just above the one chromatin body shown in the figure, though many were present in the cell. All the chromatin bodies are represented in an animal from Line VI (fig. 21, pl. 2). Nearly all are more or less spherical and the dissolution of the chromatin is evident. Two macronuclear anlagen, lying close together, and stained a reddish-blue, show marked paucity of chromatin. The two functioning micronuclei are lying at the opposite ends of the cell, while a spindle, in which chromatin has passed to the two poles, is also visible. This spindle represents the same forma- tion as that figured by Maupas in a conjugating animal (’89, pl. 13, fig. 27). Above one of the chromatin bodies is seen a small chromatin granule which may be interpreted as a degen- erating micronucleus. The disintegration of the chromatin bodies, which are now often surrounded by a clear area, is also progressing in the ani- mal represented in figure 20 (pl. 2). The macronuclear anlagen have lost their round contour and appear more or less irregular while the two micronuclei are ready for division. The micro- nuclear divisions are far more advanced in the specimen shown in figure 22 (pl. 2) and the two spindles present indicate an ensuing cell division. One micronucleus is apparently degen- erating. Large vacuoles are usually present in the cytoplasm during this period. 452 LORANDE LOSS WOODRUFF AND RH. ERDMANN The last period of the ascending phase is characterized by the disappearance of all chromatin bodies. Figure 25 (pl. 3), Line Vib, represents an animal, after the first cell division following the origin of the macronuclear anlagen, showing two micronuclei at the left, twenty chromatin bodies and several large vacuoles. A single new macronucleus is visible. The other macronuclear anlage, of the two in the parental cell, has been distributed to the sister cell which was kept to continue Line VIb. This method of distribution of the two macronuclear anlagen to each of the new animals is the same as in typical conjugation of Paramaecium aurelia. Maupas (’89, p. 222, pl. 13, figs. 23-27) gives a full account of this distribution. Hertwig’89, p. 38) shows that one of the products of each of the two dividing micronuclei becomes a macronuclear anlage and the other a micronucleus, but this is not evident from his figure (pl. 3, figs. 9and 10). The destruc- tion of the old macronucleus occurs in most Infusoria during and after conjugation by the formation of ribbon-like structures which are finally resolved into chromatin bodies and disappear. This is the method which obtains in Paramaecium caudatum and Paramaecium aurelia (Maupas, Hertwig) and in Paramaecium putrinum (Doflein). But the accounts of this stage given by Maupas, Hertwig, and Calkins do not afford details of the ulti- mate fate of the chromatin bodies, since they merely mention that, by the formation of the daughter cells, these bodies decrease in number and their remnants become pale and disappear in the cytoplasm. Collin (’12, p. 223) however, gives a thorough ac- count of the old macronucleus in Acineta papillifera and figures (text-fig. 63) the gradual resorption of the chromatin remnants in the cytoplasm. A different fate of the macronucleus was observed by Ubisch (13, p. 72) in a study of Lagenophrys, who states that the chromatin bodies, which appear very similar to micronuclei, are actually ejected from the macrogamete after the formation of the synearyon. The extruded chromatin bodies were finally observed between the animal and its test. The total disappearance of the chromatin bodies in the reor- ganization process of Paramaecium aurelia is thoroughly described on page 455. It is distinguished from the similar stage in con- PERIODIC REORGANIZATION IN PARAMAECIUM 453 jugation by the fact that chromatin ribbons are not formed. There is no evidence that the chromatin bodies are ejected from the cell or that the remnants fuse with the new macronucleus. After the reorganization process the totally dissolved material of the old macronucleus probably remains, at least temporarily, in a changed chemical form in the cell. From our studies of conjugation in animals derived from the main culture of the pedigreed race of Paramaecium aurelia we can show clearly that the first division after conjugation dis- tributes to each cell two micronuclei and one macronuclear anlage (fig. 33, pl. 3). There is no degeneration of nuclear mate- rial as Maupas (’89, p. 204) and Klitzke (’14, p. 8) have described in Paramaecium caudatum. Maupas as well as Klitzke state that eight micronuclei are formed from the syncaryon by two mitoses quickly following each other, four of the resulting micro- nuclei forming the macronuclear anlagen, three degenerating and one remaining as the definitive micronucleus of the reorganized cell. By two cell divisions, accompanied by divisions of the micronuclei, the typical vegetative stage in Paramaecium cau- datum is restored. However, Calkins and Cull (’07, p. 387) do not mention a degeneration of nuclear material in this stage but state that four of the eight micronuclei are transformed into macronuclear anlagen and, by two cell divisions without division of the micronuclei, four typical vegetative animals result. Doflein in his account of Paramaecium putrinum gives the same inter- pretation as Maupas and Klitzke, i.e., three micronuclei degen- erate and the four potential macronuclear anlagen are distributed by two cell divisions to four animals. Hamburger (’04, p. 223) states that in Paramaecium bursaria two micronuclei and two macronuclei are formed which are distributed in one of the two following ways to the sister cells which arise: either one micro- nucleus and one macronuclear anlage pass to each new cell; or one macronuclear anlage and one micronucleus (the productof an additional micronuclear division) pass to one sister cell, while one macronuclear anlage and the other micronucleus (resulting from the extra micronuclear division) passes to the other sister cell.. The normal vegetative state of Paramaecium bursaria, 454 LORANDE LOSS WOODRUFF AND RH. ERDMANN completely free from disintegration products of the old macro- nucleus, is often not completed before the third division. All descriptions of the behavior of the paramaecium cell of the different species after the formation of the syncaryon agree in the origin of four or eight new micronuclei, but these micro- nuclei resulting from the synearyon are either all used in the further development of the cell (Calkins and Cull, Paramaecium caudatum; Maupas, Hertwig, Woodruff and Erdmann, Para- maecium aurelia; Hamburger, Paramaecium bursaria), or some degenerate (Maupas, Klitzke, Paramaecium caudatum; Doflein, Paramaecium putrinum). Therefore the weight of evidence at hand indicates that the latter alternative is due to a wrong inter- pretation. Similarly ‘in the process under discussion in Para- maecium aurelia there is no evidence of a wasting of chromatin material in this period and its morphological features are iden- tical with all statements of the nuclear changes in typical con- jugation in this species. The observations of Hamburger on Paramaecium bursaria prove that the first micronuclear division after the development of the anlagen may occur either in the old cell or later in the two daughter cells which arise from it. Before the first cell division in the reorganization process two spindles are formed (fig. 22, pl. 2) which indicate that the first cell division after the forma- tion of the macronuclear anlagen restores the normal nuclear _apparatus of the cell. During the next two generations the rem- nants of the old macronucleus are completely eliminated. This is effected by complete dissemination in the cell of the chromatin bodies (fig. 27, pl. 3) in the form of cloud-like masses of faintly staining chromatin granules which soon undergo totalinvolution in the cytoplasm. This disintegration process does not occur simultaneously in all the chromatin bodies as is shown in the figure under discussion in which four bodies are still intact. The animals which have undergone their first vegetative divi- sion are filled with vacuoles as shown in an animal from Line IIIb, 4313th generation (fig. 27, pl. 3). The macronucleus has assumed the peculiar spherical form which it frequently shows immediately after. cell division. . The climax was in the 4312th PERIODIC REORGANIZATION IN PARAMAECIUM 455 generation and the involution of the chromatin bodies was com- pleted in the 4316th generation. Of the eight cells which were formed by the three following divisions (43138 to 4316), four in the 4316th generation were without chromatin bodies. The pre- ceding figure (fig. 26, pl. 3) represents an animal from Line VI, 4185th generation. Two divisions have occurred since the climax at the 4183d generation. Here the chromatin bodies which still remain have not lost their original shape, but if one considers the four division products which result from the divisions 4183 to 4185 then it is evident that a second method of decreasing the number of chromatin bodies is by their distribution to the daughter cells. As an example of this distribution the following ease may be cited (cf. text fig. 8 for sister line): Line VI 4183d generation has 19 chromatin bodies Line VI 4185th generation, animal a_ has 6 chromatin bodies &Line VI 4185th generation, animal b_ has 5 chromatin bodies: Line VI 4185th generation, animal ec was not observed Line VI 4185th generation, animal d_ was kept alive to continue the line Two further divisions took place in twenty-four hours: Line VI 4187th generation, animal a_ has still 3 chromatin bodies (fig. 29, pl. 3) Line VI 4192d generation has 1 chromatin body Therefore the ascending phase of the process can be followed in nine generations; that is, from the 4183d generation, which is at the end of the climax, to the 4192d generation which is a typical vegetative paramaecium cell except that one chromatin body still persists. Since the beginning of the process was ob- served at the 4182d generation, it is evident that the nuclear changes in this case extended over ten generations. The typical reduction of the division rate at the climax of the process to one division in approximately thirty-six hours occurred in this case at the 4182d to 4183d generation, and the increase in the rate to three divisions in twenty-four hours which characterizes the ascending phase of the process occurred at the 4189th to 4192d generation. When the second or third division after the formation of the macronuclear anlagen has occurred the new macronucleus has 456 LORANDE LOSS WOODRUFF AND. RH. ERDMANN assumed its typical macronuclear characteristics. The elongated form is the prevailing one in the interval between two cell divi- ° sions, the broad axis is relatively short, while the few granules within the membrane are small. The appearance of an old and new macronucleus is markedly different and the two stages can- not be confused (compare figs. 3, 4 and 5, pl. 1, with figs. 28, 29 and 30, pl. 3). Figure 29 and figure 30 (pl. 3) show the new macronuclei. The animal portrayed in the former is from the above-mentioned Line VI in the 4187th generation. One large and two small chromatin bodies are visible. The latter shows a cell which was taken from the main culture in the 1201st gen- eration. The staining capacity of the macronucleus is very marked; the micronuclei present no unusual features; and two chromatin bodies still persist. A dividing animal from the main culture at the 1432d to 1433d generations (fig. 28, pl. 3) further illustrates the distribution of the chromatin bodies. This anintal has undergone four divisions during the previous twenty-four hours and is now just completing the fifth cell division, while the micro- nuclei have divided precociously for a sixth cell division. This affords another example of the acceleration of the division rate which is characteristic of the ascending phase of the process. The ensuing divisions in the rhythmical period efface the last trace of the chromatin bodies and, therefore, of the reorganiza- tion process. In conjugation, as has been shown by Maupas and Hertwig, the chromatin bodies undergo the same fate at the same time. The reorganization of the cell is completed. In the description of the process nearly all the examples pre- sented have been animals from Sub-culture IE (4020th to 4359th generation) bred for this particular purpose under constant en- vironmental conditions. As already described in the section on technic, animals had been preserved at irregular intervals through- out the first six years of the life of the main Culture I and on the basis of these it was early stated (’08) that profound nuclear changes occur which cannot be interpreted as abnormal and which demand further study. The character and significance of these changes obviously could only be determined by such an intensive daily study of all the available animals as the pres- PERIODIC REORGAWIZATION IN PARAMAECIUM 457 ent paper involves. It is important, however, at this point to emphasize that the process herein described is not a phenomenon peculiar to the organisms of this culture after the 4020th gen- eration, but that it was in progress throughout the life of the race. For this purpose various specimens from earlier gener- ations are presented in the plates. For example, figure 23 (pl. 2) represents an animal from the main culture on May 1, 1910, at the 1755th generation with its nuclear apparatus in a char- acteristic stage of the process, 1.e., the formation of the macro- nuclear anlagen. This should be compared with figure 20 (pl. 2) which shows an animal which is a descendant of the former after a lapse of 2329 generations, i.e., it is in the 4084th generation. The similarity of nuclear conditions also is evident in figures 21 and 22. Cases such as this could be multiplied but for brevity it will suffice to note that among the preserved material there are specimens in the 426, 910, 1201, 1452, 1498, etce., generations which exhibit reorganization stages in all respects the same as those the animals have shown during the past six months of study. Consequently, when convenient, these have been figured to illustrate the process. IV. CYTOLOGICAL DETAILS OF SERIES OF PEDIGREED CELLS IN THE REORGANIZATION PROCESS FROM SUBCULTURE IE In the preceding description of the cytological changes in the process the data from all lines were presented together in order to afford a composite picture of the nuclear phenomena as they appeared from the entire culture. It is the purpose of the pres- ent section to resolve this picture into its component parts in order to emphasize the exact sequence of events as actually observed in individual pedigreed series of animals from single lines and thus to demonstrate that our description of the process is not derived merely by combining isolated stages found at different times in the different lines, but by combining the data from a large number of series of animals whose genetic relation with each other was exactly known—each series showing a con- siderable part of the process. In combining these series there THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 17, No. 4 458 LORANDE LOSS WOODRUFF AND RH. ERDMANN was little chance for error for their cytological phenomena fre- | quently overlapped. As has already been indicated, it is obviously impossible to study every animal every day and so have a complete picture of the animals of a single line at any one time, since one cell must be maintained if the line is to be kept alive. In certain cases it was deemed best to keep more than one animal alive in order to further safeguard the life of the line because death some- times occurs during the process. However, it has been possible actually to determine the entire sequence of events by preserving all of the animals of certain lines at selected periods in the pro- cess. The fact that the process occurred nearly synchronously in the various lines and at comparatively regular intervals ren- dered this procedure particularly valuable. A. SERIES OF CELLS FROM LINE VI AND ITS DERIVATIVES Line VI at the 4057th generation afforded the first proof of the reconstruction of the cell, two characteristic macronuclear anlagen and twenty-four chromatin bodies being present. The animal is in the beginning of the ascending phase of the process. In the 4060th generation the process was completed as the ani- mal does not show any macronuclear remnants. The large num- ber of typical animals which followed in this line for nearly fifty divisions was finally interrupted by the appearance of a cell at the 4094th generation which has two chromatin bodies. Its sister cell (4094) has no signs of the beginning of the process and the following generations up to the 4102d generation showed no further traces of reorganization. This indicates that the cells when the critical period is approaching may start the process but not carry it to completion. The total reconstruction of the nuclear apparatus took place between the 4100th and 4107th generation. Text figure 4 gives two stages of the process; the disintegration of the old macronucleus in A, and A, and A; and the formation of the new macronuclear anlagen in Ay. In the 4100th and the 4102d generation the macronucleus is merely eliminating chromatin bodies while in the 4103d generation the PERIODIC REORGANIZATION IN PARAMAECIUM 459 . individuality of the macronucleus is destroyed. The animal in the 4105th generation was lost. The cell in the 4107th gener- ation had already formed two macronuclear anlagen and numer- ‘ous chromatin bodies. One of the latter was still present at the 4110th generation so that in this case the process persisted over about nine generations. The next critical period began at the 4140thgeneration. As already mentioned, Line VI made an abortive attempt at the process in the 4094th generation by the extrusion of a single Ag VI 4100 VI 4102 VI 4103 VI 4107 Text fig. 4 Conventional signs, same as in text fig. 1 chromatin body. Line VlIa, which was branched from Line VI at the 4068th generation,.showed stages of the process from the 4084th to the 4089th generation so that the time of appearance of the process in Line Vla coincides closely with the precocious trial in Line VI at the 4092d generation. In Line Vla the for- mation of the macronuclear anlagen took place in the 4084th generation (see text fig. 5, B,). The animal 4084, was not pre- served but 4084; in the descending stage is shown in text figure 5, B:. Cell 4084, continued the line. This cell must have been in the same stage of the reorganization process as 4084;, because 460 LORANDE LOSS WOODRUFF AND RH. ERDMANN 40882 Via 4088, Via 4087 Text figure 5 Via 4084; Via 4084, PERIODIC REORGANIZATION IN PARAMAECIUM 461, in the following generations the process is still going on from the 4084th to 4088th generation (Line VIa, text fig. 5, Bs, Bs, Bu, B;). Therefore, the process does not progress exactly synchron- ously in sister or cousin cells. This is clear from the diagram, text figure 6. The stages of the process in Line VIa represented in text figure 5 show the climax and the ascending phase (Bs, B,, Bs). Cell 4087 is about to form the macronuclear anlagen. Their formation took place in cells 4088; and 4088, (B,, B;). Line Ya ys: onto Text fig. 6 Diagram showing a case of nonsynchronism of the reorganization process in closely related cells. The fact that the sister cells show slightly different stages in the process of course has no significance from the standpoint of the synchronal appearance of the process in sister lines; that the latter occurs is apparent from tables 1, 2 and 3 (pp. 462-463) which give the occurrence of the process in the lines branches from Line VI. Thus in VIa the process was going on in the 4084th generation, in VIb in the 4092d generation. An exami- nation of table 1 shows that the process was observed in Line 462 LORANDE LOSS WOODRUFF AND RH. ERDMANN TABLE 1 Showing the occurrence of the reorganization process in Line VI and its derivatives, Culture IE 2 TOUR beac Meek: cE era Re oe SON. | VI | 4057 | 4104 4183 4231 | | Died in 4020-4231. CBee aie aie | __ process Via | 4084 q Sn. ne 4068-4095 | | | Vib | | 4092 | 4174 | | Killed in 4083-4174 | | a process Vie | | ~4101.—«| 4177 | | Killed in 4099-4178 | | | process Vid | |* 4180 : 10) \aRatieaiin 4178-4180 | | | | process Yer | | | 4181 | Killed in 4178-4181 | | process vif | 4189 4229 [ae eal Killed in 4178-4228 | | process Vig i | 4184 ~ Killed in 4180- 4186 | | process Vih | 4236 4315 | 4355 | Killed in 4189-4358 | i OL process Vii | 4297 | Killed in 4246-4297 | iis Snes eo | | process Vij" | | 4312 | Killed in 4306-4315 Tite process VIk* | | | Killed 4325-4336 | during low di- vision rate; animal not in Aye leit i) har ih ate! Sw process vil* | | | 4355 Killed in 4348-4355 | process * Sub-lines j, k, | were branched from i. VI in the 4104th generation, in VIc in the 4101st generation. Further evidence of the synchronal occurrence of the process is afforded by the next period in table 1 which shows the process in Line VI at the 4183d generation and in six branches from this line (VIb, VIc, VId, Vle, VIf, VIg) between the 4174th and 4184th generation. These facts are substantiated by further data as presented in the graphs of the division rate. The third repetition of the reorganization process in Line VI took place between the 4140th and 4146th generation. This was relatively early since the second occurrence of the process PERIODIC REORGANIZATION IN PARAMAECIUM 463 TABLE 2 Showing when the reorganization process occurred in Line IIIT and derivatives, Culture IE | | | 20-4050 4050- L 150-4200 | 4200-4250 | 4250- - 50- shame ea Saas 4100-4150 4 50-4 200 4 Bee) 4 50-4300 4300-4350 4350-4400 ieee | | WI 4065 | 4189 4237 | | 4315 | Killed at 4020-4358 | | | end of | experi- | | | ments Ila | | | | | 4271 iDied 4946-4275 | | | | | after Ve le | | | | process IIIb | | 4313 | Killed; 4304-4321 | | | animals es | |. | | small lites > >| | | | Killed; 4327-4346 | division ie tate low, | | not in, | | process was at the 4104th generation. The ascending phase of the process was completely secured (text fig. 7). Cell C, (4140th generation) has two macronuclear anlagen; C, (4141st genera- tion), resulting from one division, has one macronuclear anlage; C; (4142d generation) shows the distribution of the chromatin bodies in the daughter cells. The next occurrence of the process in Line VI, represented in text figure 8, gives nearly a complete picture of each stage in the process in a single line. The animal (4180-4181) just divid- ing was typical, while the following cell (4181-4182) shows simply the extrusion of one chromatin body and the remnant of a TABLE 3 Showing when the reorganization process occurred in Lines I, II, IV, and V, Culture TE anions = ACID totem pee, sen RATERS ae = : s | abe fs eo eed : de é Line I from 4020 to 4057 gen..| 4038 Died in process; last animals small due to abnormal divi- sion; lived 4 days without johsas he ns et _division. _ i Line II from 4020 to 4172....... 4021 4069 4172 _| Line killed in process? Line IV from 4020 to 4075...... =, 4066. =| SSS Animals small; died Line V from 4020 to 4170...... 4023 | | 4115 | Killed; not in process 464 LORANDE LOSS WOODRUFF AND RH. ERDMANN division spindle. This proves that in generation 4182 the com- plete extrusion of chromatin bodies will be completed and the ‘reduction’ divisions of the micronuclei and their degeneration must also occur here during the thirty-six hours that this cell persists as an individual. The low division rate at this point indicates that the profound nuclear changes temporarily inhibit the reproductive acitivity. The animal undergoing the next division (4182-4183) was not preserved but evidently would have shown each of the arising cells in the climax of the process Cy ©) C; C, VI 4141 VI 4142 VI 4148 Text figure 7 with one or two micronuclei and numerous chromatin bodies. However, one of the completed products of this division, 4183d generation (D;), shows the two macronuclear anlagen completely formed. The next stages (4185,, 4185,, 4187) show the distri- bution of the chromatin bodies. Cell 4184 was not preserved but evidently it had one macronuclear anlage, so from gener- ation 4182 to 4192 (D;) with one chromatin body, we have the process characterized physiologically by the lowering of the divi- sion rate in the 4182d to 4183d generations when reproduction was deferred for about thirty-six hours, and then the sudden acceler- cOIF IA Z8IF IA "SIF IA 8 9INSY 4X9], 'SSTP IA S8IF IA G8IF-I8TF IA I8Th-O81F IA 465 466 LORANDE LOSS WOODRUFF AND RH. ERDMANN ation of the rate at the 4187th to 4192d generation when three divisions occurred in the twenty-four hours. The morphological characteristics for the single periods are well established as text figure 8 (D, to D,) proves. The descending phase of the process is of the shortest duration and this makes it particularly difficult to secure all the stages—several coming and going in a single generation. On the other hand the ascending phase is the long- est and, therefore, the stage most often seen in all lines is that with two macronuclear anlagen. The synchronism of the proc- ess in Lines VI, VIb, VIc, VId, VIe, VIf and VIg made it possible Ei E2 E3 Ex Vid 4180; VI d 41802 VI d 4180; VI d 4180, Text figure 9 to kill whole lines undergoing the process and in this way some of the most interesting stages were secured. Text figure 9 (E, to Es) gives four cells (line VId) all at the same stage with the formation of the two macronuclear anlagen completed. The next stage will be the distribution of the two macronuclear anlagen into two sister cells. This stage is simul- taneously completed in four cells of Line VIb at the 4174th generation. Text figure 10 (F, to F;) shows three of these cells PERIODIC REORGANIZATION IN PARAMAECIUM 467 which result from two divisions in forty-eight hours. The fourth animal (F,) was lost. Since each of the three animals preserved has one macronuclear anlage it is certain that each of thetwo parent cells in the 4173d generation had too macronuclear an- lagen, one of which passed to each daughter cell at division. The fourth animal (F,) which was lost during preparation obvi- ously must have been in the same stage as those figured. The next text figure, 11 (G, to Gs), shows again that at least seven or eight generations are necessary to complete the process. The F, F, . F; F, VI b 4174, VI b 4174, VI b 4174; VI b 4174: Text figure 10 animals are from Line VIf in the 4184th, 4188th, and 419Ist generation and show the extrusion of the chromatin bodies from the old macronucleus, the new macronuclear anlagen and the nearly reorganized paramaecium cell. The last occurrence of the process in Line VI extended from the 4229th generation to the death of the line at the 4231st gen- eration from the accidental infection of the culture medium with a deleterious strain of bacteria, as mentioned on pages 478 and 489. Five times in 211 generations the cell reorganized its nuclear 468 LORANDE LOSS WOODRUFF AND RH. ERDMANN apparatus by the process; six times the division rate (see p. 478) was relatively low. Five times this was synchronous with the process; once with bacterial infection. The process occurred in Line VI and its sublines at intervals of about forty to fifty generations. For example, Line VI had the process in progress at the 4231st generation, Line VIf at the 4229th generation, and Line VIh at the 4215th generation. The stages of the process which occurred in Line VIi in the 4279th generation and Line VIh in the 4315th and 4355th gen- Gi VI f 4184 VI f 4188 VI f 4191 Text figure 11 eration, and in Line VIj in the 4312th generation did not show any new features. Simultaneously with Line VIh, Line VII underwent the process in the 4355th generation. Line VIk was killed after eleven generations, when the division rate fell slightly, with the expectation of securing early stages of the process. However the process was not found but judging by closely related lines it would have appeared about twenty generationslater. Line VI together with its twelve branches and subbranches exhibited the process a total of twenty-one times from the 4020th to the 4355th generation at clearly recognizable periods, showing that PERIODIC REORGANIZATION IN PARAMAECIUM 469 there is an inherent tendency for sister lines to undergo the process synchronously. The significance of this will be consid- ered later. The death of Line VI during the process indicates a fact which we have noted several times in these experiments and one to be expected when the complexity of the reorganization process is appreciated. The animals at this time are in a relatively sus- ceptible condition and consequently more readily succumb to slight environmental changes. It is undoubtedly at such periods that many cultures not bred under the most favorable conditions have become exterminated, and the attending cytological con- ditions have naturally been interpreted as the results of de- generation. The cells during the process seem to be more opaque than usual, their breadth is relatively greater, their total volume is somewhat increased and their movements are considerably more sluggish. Cells with this appearance defer division for.‘about thirty-six hours and are at the climax of the process. The ani- mals at this time are more difficult to handle with a pipet owing to the fact that they have a tendency to adhere to the glass or to minute particles of débris, thus indicating the so-called misci- ble state noted by Calkins, Erdmann, Popoff and others in con- jugating forms. Also, animals in this condition will sometimes burst on transference to fresh culture fluid medium or to a fixing fluid. Thus, by the continued daily study of the animals and their rate of division one finally becomes able to tell, with a consider- able degree of accuracy, the occurrence of the process merely by the use of a low power of a Zeiss binocular microscope. B. SERIES OF CELLS FROM LINES I, I, Ill, IV, V Line III, which was not under such close observation as line VI, had the process at the 4065th, 4189th, 4237th, 4315th gen- erations. Text figure 17 gives the graph of the division rate averaged for five-day periods. One lowering of the division rate occurred at generations 4140 to 4144. During this time the 470 LORANDE LOSS WOODRUFF AND RH. ERDMANN cells from Line III were not preserved but judging by the division rate of the line and the occurrence of the process at this period in Line VI, it is almost positive that the process took place in Line III at this time. Cell 4148 was preserved and found to be a typical cell without traces of the process. Text figure 12 (H, to Hs) give examples of the ascending phase of the process | from Line III. The cells 4063, and 4063, had formed the macro- nuclear anlagen. Two divisions later, cell 4065 (H;) had only seven chromatin bodies, showing a reduction of these bodies by Hy Hees H; III 4063, III 4063. IIT 4065 Text figure 12 distribution and absorption from twenty-seven (H;) to seven dur- ing these two cell generations. Text figure 13 (I, to I;) give some of the later stages of the ascending phase. The three animals which could be observed had a few chromatin bodies. Cell 4240, had two, 4240. one, 4240; one chromatin body. Cell 4240,, which continued the line, gave rise to typical animals at the . following division, 4241st generation (Is). Text figure 14 (J, to J3) shows the process after the formation of the macronuclear anlagen in cell 4189. After two divisions the new macronucleus is seen in the sister cells 4191, and 4191,. s PERIODIC REORGANIZATION IN PARAMAECIUM 471 Thus the distribution of the two macronuclear anlagen took place in the 4190th generation, which affords further proof that the two macronuclear anlagen are distributed by one cell division to each of the resulting sister cells. In Line III and its branches, IIIa, IIIb, IIIc, the process was actually observed six times, while undoubtedly it also occurred several’ times unobserved, judging by the division rate and the periodic occurrence of the process in sister lines (see table 2 and text fig. 17). ieee ip 1p iG I; III 4240, III 42402 IIT 4240; III 4240, III 4241 Text figure 13 The Lines I, H, IV, V, which were not branched, show the process a total of seven times, each time at due periods coin- ciding with Line VI. Some of the stages observed in these lines are presented in the plates as they supplied some of the details of the process as, for example, Line II, 4069th generation (fig. 9, pl. 1) which give a particularly clear idea of the extrusion of the chromatin bodies. One point which is shown by a study of tables 1, 2, and 3, which give the life history of the twenty-one lines, is the occur- rence of death in certain instances during the process. Of these 472 LORANDE LOSS WOODRUFF AND RH. ERDMANN lines seventeen were killed intentionally, while of the remaining four lines, three died during or shortly after the process. Some of these animals showed irregular divisions resulting in very small cells which apparently lacked the power of growth. It is important here to emphasize that fatalities may occur as a result of faulty reorganization during the process just as has so often been observed to occur in conjugating animals. Both cdnjuga- tion and the reorganization process, however, are normal phe- nomena which have no pathological features. Ji Je J3 III 4189 IIT 4191; III 4191. Text figure 14 The detailed discussion of Lines I to VI and their branches and subbranches gives clear evidence that the process, resulting in a reorganization of the nuclear apparatus which simulates typical conjugation, is going on with strictly determined morpho- logical and physiological features at relatively definite periodic intervals of time in the pedigreed race of Paramaecium aurelia which is the basis of this investigation. PERIODIC REORGANIZATION IN PARAMAECIUM 473 V. THE REORGANIZATION PROCESS IN THIS RACE AFTER CONJUGATION AND IN OTHER RACES AND SPECIES OF PARAMAECIUM So far it is clear that the nuclear reorganization occurred throughout the life of this race bred under conditions which absolutely precluded the occurrence of conjugation, and the ques- tion arises here: Would the process again occur periodically after conjugation had taken place? Animals were accordingly isolated from this race at the 4102d generation and with them a large mass culture was started in a stender dish. Within afew daysa number of conjugating pairs were observed and about twenty were isolated. Some of the animals were preserved for details of con- jugation in this race, while the descendants of other ex-conjugants were bred on depression slides by the same isolation culture methods as those used in the main culture from which they were derived at the 4102d generation. A study of the animals pre- served from these ex-conjugant lines demonstrates that the same reorganization process was resumed wn all the lines within a rela- tively short tume after conjugation. More positive evidence could hardly be presented to prove that the process is a fundamental normal periodic phenomenon in the life of this race of Paramaecium aurelia. This being established, the question arises: Is this a peculi- arity of this race or does the process occur generally in Paramae- cium aurelia? This we can also answer conclusively. Erdmann on August 11, 1912, isolated a specimen of Paramaecium aurelia from a canal of the river Spree in Berlin and bred its descendants by the daily isolation method on a culture medium of beef ex- tract. Specimens from Erdmann’s race, bred by this method which absolutely prevents conjugation, were preserved from time to time and figure 24 (pl. 2) shows one of these in a characteristic stage of the process under discussion, i.e., the formation of the macronuclear anlagen. Thus it is evident that the same nuclear reorganization which has occurred throughout the life of Wood- ruff’s race started with a specimen of Paramaecium aurelia iso- lated at random in America in 1907 also occurred in Erdmann’s race similarly isolated in Germany in 1912. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 17, No. 4 A474 LORANDE LOSS WOODRUFF AND RH. ERDMANN Therefore the data justify the conclusion that this reorganiza- tion process is a normal phenomenon and probably occurs in all races of the species Paramaecium aurelia. Most suggestive observations on nuclear changes in Para- maecium aurelia were made incidentally by Hertwig during his study of conjugation in this species. This author noted (’89, pp. 74-75) in a mass culture, in which conjugation had not been observed for a long period, certain animals whose nuclear struc- ture apparently indicates isolated steps in the phenomenon which are described in detail in the present paper, and the significance of which is here elucidated. Hertwig grouped the stages which he observed under two gen- eral classes, one of which he interpreted as merely a vegetative phenomenon, in which a new nuclear apparatus is not formed: Die zuerst eintretenden Veranderungen besitzen kein Analogon in den Vorgingen einer normalen Entwicklung. Wahrscheinlich zerfallt der Hauptkern erst in gréssere dann in kleinere Stiicke, ohne das regel- missige Auswachsen in Fortsitze, welches im Lauf der geschlechtlichen Entwickelung der Paramaecium eintritt. Ich fand bald 2, bald 4 Nebenkerne entweder in Form der ruhenden Kerne oder hiufiger in Form von Spindeln, wie ich sie ebenfalls sonst nicht beobachtet habe. Our data show this to be a mistake. These stages are clearly some steps in the descending phase of the process just before the beginning of the climax. The second group of changes of the nuclear apparatus described by Hertwig comprised: ‘‘Tiere mit vergrésserten Nebenkernen, mit Sichelkernen mit 2, 4 und 8 Spindeln, Tiere, bei denen die Teilung in die Haupt-und Nebenkernanlagen vollzogen war.” From this brief note it is clear that the animals with two and four spindles may possibly represent ordinary vegetative divi- sion because the striking disintegration of the macronucleus is not mentioned. We have never discovered an animal with eight spindles, but ‘‘Tiere, bei denen die Teilung in die Haupt-und Nebenkernanlagen vollzogen war” are apparently identical with those animals which we figure (figs. 21, 22, pl. 2) and represent the important reconstruction of the new nuclear apparatus. If it be true that in Hertwig’s culture conjugation had not oc- PERIODIC REORGANIZATION IN PARAMAECIUM 475 Text figure 15 curred (it is impossible to prove this since he employed mass cultures in the work) then it is also true that he discovered at least one of the steps in the process which is outlined in the pre- sent paper but obviously he failed to recognize its general funda- mentalimportance. This is evident from the trend of his work and that of his students, on the life history of Infusoria, during the subsequent quarter of a century. Another question naturally suggests itself at this point: Is the process confined to Paramaecium aurelia? Figure 34 (pl. 3) is from a pedigreed race of Paramaecium caudatum which we have studied by the same experimental methods, and shows a characteristic stage in the process. This indicates that it occurs, at least with essentially similar features, in Paramaecium cau- datum also. It will be shown in a later section of this paper that certain of the morphological changes interpreted by Calkins, Popoff and others as degeneration phenomena in Paramaecium caudatum are in all probability the same process. It may be mentioned here, however, that Doflein (’07) figures an abnormal conjugating pair of Paramaecium caudatum, one animal of which appears to be a typical vegetative animal while the other shows what he interprets as conjugation nuclear phenomena in a single 476 LORANDE LOSS WOODRUFF AND RH. ERDMANN animal. From our results we would interpret this as a stage in the ascending phase of the process, because he figures the four macronuclear anlagen typical for Paramaecium caudatum and terms them Mikronukleusderivate (see text figure 15, which is a copy of Doflein’s figure). Finally, does the process occur in other genera of Infusoria, and in other classes of Protozoa? Later in this paper we shall discuss the literature relative to this point but it can be stated here that while no positive evidence is extant to answer this question, because pedigreed cultures involving a daily detailed cytological study have not been made, nevertheless, such more or less random facts as may be gleaned from various investiga- tions strongly suggest, at least, that the authors are dealing with comparable nuclear reorganization phenomena. VI. THE REORGANIZATION PROCESS AND ITS RELATION TO RHYTHMS Earlier work on this culture has shown that there are rhythms in its rate of reproduction which are not the results of environ-. mental variations, but which are due to some periodic internal phenomena of unknown character. Minor fluctuations are evident in the division rate of other species of Infusoria bred by the daily isolation method first employed by Calkins. The fluctuations in the culture graphs of Calkins on Paramaecium caudatum, Woodruff on Oxytricha fallax and Gastrostyla steinii, Gregory on Tillina magna, and Moody on Spathidium spathula appear to be of the same character as those demonstrated in this culture. However, - since the authors in the study of the life history of these species have not eliminated the possibility that the fluctuations in the division rate are the results of environmental changes, it is im- possible to state positively that the fluctuations observed by them are actually ‘rhythms,’ though it is highly probable that such is the case. It is important to emphasize the fact, although it is well known, that relatively slight changes in culture medium or tem- PERIODIC REORGANIZATION IN PARAMAECIUM 477 perature will change the division rate. Indeed it may be said that the rate of reproduction is a function of the environment of the cell—except as the rhythms interfere—and consequently the fluctuations which appear in the graph of the division rate of a culture are not, ipso facto, true rhythms, i.e., due to “inherent rhythmical changes in the phenomena of the cell’? (Woodruff and Bantsell: “11. b, ps. 3o7). The present cytological study demonstrates the nature of the inherent changes in the cell which have their obvious physio- logical expression in the rhythms of the rate of reproduction. The relations of the process to these fluctuations can be most readily appreciated by a consideration of graphs of the rate of divisions of Lines VI and III of Sub-culture IE, which was sub- jected to the constant culture medium of beef extract and to practically constant temperature conditions (about 26°C.) with- in the optimum zone for this race. The five-day period was adopted in the presentation of our results because this was the method of constructing the graph emphasized in the original study of rhythms in this culture. It is realized, of course, that a five-day period is largely an arbitrary one and that the ideal graph would present the momen- tary changes in the metabolism of the cell. Data for such a curve being absolutely impossible to secure, it might seem at first glance that the daily record of division would approach most nearly to this ideal condition. As a matter of fact, the twenty-four-hour period is as arbitrary as the five-day period when it is considered that this is a long period when compared with the metabolic changes in the cell and that the daily record, made at approximately 11 a.m., would merely give the divisions actually completed during the previous twenty-four hours. for example, let us assume that, at the time of isolation, two animals are present, representing one division during the previous twenty- four hours. The record for that day is one division. One ani- mal is then isolated and it divides within an hour and each of the resulting cells again divide twice before the next isolation. The record for this second day is three divisions, thus the record for the two days shows a different division rate for each day, 1.e., 478 LORANDE LOSS WOODRUFF AND RH. ERDMANN one division against three divisions, whereas a more true, but not a perfect, picture of the state of affairs is given by the state- ment that four divisions occurred in forty-eight hours. One might follow this argument to its logical conclusion and assume that the best method of presentation would be to average for considerable periods, e.g., 10 or 30 days, but this obviously would tend to obliterate any fluctuations in the rate which are not of relatively long duration. The adoption of the five-day period was made in recognition of both of these contingencies, and it was of a duration particularly well suited to show the effect of the process on the reproductive rate, because the process extends over about nine cell divisions or a pericd of about six days. Consequently the effect of the process makes itself evident in the five-day plot. Certain apparent irregularities in the coin- cidence of the phenomena are, from an actual study of all the data at hand, clearly due to the fact that the five-day period is not ideal. f te In a consideration of the relation of rhythms to the reorgani- zation process it will be convenient to consider first the data from Line VI, Subculture IE, since animals from this line were preserved every day of its existence and, during the process, every animal was preserved except the one needed to keep the line alive. Inspection of the graph of this line shows at a glance five fluctuations in the reproduction rate which would naturally be interpreted as rhythms (cf. text fig. 16) while a study of the cytological condition of the specimens shows that the low point of each of the first four fluctuations is coincident with the pro- found nuclear changes of the reorganization process. The fifth fluctuation, during which the division rate fell to nearly three- quarters of a division per day, was brought about by the con- tamination of the medium at the time of transference with a deleterious strain of bacteria, from the effects of which the ani- mals of this line never completely recuperated and succumbed upon the recurrence of the process for the fifth time. Conse- quently it is evident that four out of the five fluctuations are PERIODIC REORGANIZATION IN PARAMAECIUM 479 Text fig. 16 Graph of the rate of division of line VI, subculture IE, averaged for five-day periods. The periods during which the reorganization process oc- curred are indicated by an X. actually rhythms, 1.e:, not obviously due to changes in the envi- ronment, and at the low point of each of these rhythms the process was in progress. ‘Table 1 shows that the process in Line VI was in progress at the 4057th, 4104th, 4140th, 4183d, and 4231st generation when the line died. These generations represent ap- proximately the climax of the process. Thus it is clear that in the line under consideration the process occurred at intervals of about 40 to 50 generations or about every twenty-five to thirty days, and that the low point between each of the first four fluctuations was coincident with the nuclear reorganization. This evidence from Line VI shows clearly a relationship between the rhythms and the nuclear reorganization. Cumulative evidence which establishes a casual relationship between rhythms and the process is afforded by a study of Line III which was carried continuously for eight months. Aniinals from Line III were not preserved every day at the time of iso- lation (as was the case with Line VI) owing to the great amount of labor involved, but at intervals of several days duration. That is, no attempt was made to have a complete pedigreed series of stages showing its daily cytological condition but merely a broad 480 LORANDE LOSS WOODRUFF AND RH. ERDMANN view of its nuclear state. By this method the process was dis- covered at the 4065th, 4189th, 4237th, 4315th generation (cf. table 2). Text figure 17 gives the graph of the division rate of this line, plotted the same as that already presented for Line VI, and shows the periods in which the generations involving nuclear reorganization were discovered. Here again in four out of five cases, the process coincides with the low point in division rate, i.e., between two rhythms. The occurrence of the process of the 4315th generation, however, coincides with the early ascending phase of the rhythm and is an exception for which the data afford no evident explanation. Gt erates S ql | Text fig. 17 Graph of the rate of division of Line IIT, subculture IE, averaged for five-day periods. The periods during which the reorganization process oc- curred are indicated by an X. The other main lines were carried chiefly for the purpose of affording a sufficient supply of animals in the process and, after the first couple of months, were preserved only when the process was suspected from the appearance of the cells, the rate of divi- sion and the length of time and the number of generations since its last occurrence. Consequently it is unnecessary to consider these lines in detail from the standpoint of the rhythms but simply to emphasize that the evidence derived from them is entirely corroborative of that presented from Lines VI and III (ef. table 3). In our description of the cytological changes in the process it has been shown that the phenomenon extends over about nine cell generations, and that at the climax divi- sion is deferred for a period of nearly 36 hours. Therefore, it is PERIODIC REORGANIZATION IN PARAMAECIUM 481 evident not only that the reorganization process is coincident with the low point between two rhythms, but also that there is a causal relation between the reorganization process and the rhythms. VII. THE REORGANIZATION PROCESS AND ITS RELATION TO DEPRESSION PERIODS The results of the great majority of studies on various spe- cies of Infusoria, from the pioneer experiments of Biitschli and Engelmann, have indicated that these forms will not reproduce indefinitely if conjugation between two animals is prevented, but that the race shows signs of lowered vitality accompanied by marked morphological changes which sooner or later end in death. Improved methods of conducting the cultures, however, have enabled succeeding investigators to continue the races longer and longer, until studies on the culture which forms the basis of this paper have shown that under proper environmental con- ditions Paramaecium aurelia can be bred indefinitely without conjugation or artificial stimulation. In other words, there is no evidence of a life cycle, as understood and emphasized by Maupas, Calkins and other authors, which comprises a more or less definite number of cell generations beginning after con- jugation with the high potential of vitality of youth and maturity and leading to either old age and death or conjugation and rejuvenescence again. A critical survey of the literature, however, shows that another type of variation occurs in the life history of Infusoria under culture conditions which must be distinguished from the cycle. In certain cases this has been recognized by the author, but as a rule it has been passed over without comment owing to the fact that only by the daily isolation method of culture is it possible to demonstrate its reality. Apparently Hertwig (’00- 04) from his study by mass culture methods of the life history of Actinosphaerium eichhornii, Dileptus gigas and Paramaecium caudatum was the first to contrast minor periods and deep periods of physiological depression, the latter resulting in death. The lesser periods were marked chiefly by a slight lowering of 482 LORANDE LOSS WOODRUFF AND RH. ERDMANN the general physiological activities of the cells, and interpreted as the result of relatively slight disturbances of the nuclear con- dition which were overcome by internal readjustment. In the periods of deep depression, on the other hand, marked signs of degeneration were evident, for which the sole panacea was conjugation. This led Hertwig to the natural suggestion that the nuclear phenomena observed in physiological depression and those which occur at conjugation have fundamentally the same raison d’étre. Calkins (’04, p. 424) in his studies on Paramaecium caudatum noted that the well-marked cycles which resulted in death, unless Fhyfhm Fhythm Fracess Text fig. 18 Diagram illustrating the relation of the reorganization process to rhythms. drastic methods of artificial stimulation were resorted to, were of about six months duration, while intermediate cycles of less importance occurred at intervals of approximately ninety days, recovery from which took place without purposeful stimulation. In these smaller cycles morphological signs of degeneration were not observed, but in the well-marked cycles cytoplasmic and nuclear changes occurred. These varied somewhat in character at the low points of the various cycles, but in several instances Calkins was able to restore the normal condition by the oppor- PERIODIC REORGANIZATION IN PARAMAECIUM 483 tune use of artificial stimuli. At the last cycle when all the methods of rejuvenation which were tried proved of no avail Calkins observed that the signs of degeneration were apparent in the micronuclei and, therefore, concluded that at last ‘ger- minal death’ occurred. Some very suggestive nuclear conditions are figured by Calkins (04) from animals in his depression periods. Calkins’ figure 15, plate 2, shows an animal of his B series just before its death in the 502d generation, which had been treated with beef extract. He states ‘‘that the micronucleus has divided three or more times and the daughter nuclei have accumulated at one end.” His photograph does not convince us that these are really micro- nuclei. If they are they clearly represent an early ‘‘reduction’’ phase of the process as described by us for Paramaecium aurelia. We are inclined, however, to interpret these ‘micronuclei’ as chromatin bodies since we have found that in Paramaecium caudatum, as in Paramaecium aurelia, this is the method of dissolution of the macronucleus during the process. This figure of Calkins should be compared with our figure 34 (pl. 3) which shows the chromatin bodies in Paramaecium caudatum. Text figure 19 (Calkings’ fig. 16, pl. 2) shows an animal from his A series in the 602nd generation treated for twenty-five minutes with phosphoric acid. It was transferred to hay infusion and killed twenty-four hours afterwards. The macronucleus is broken into frag- . ments; the micronucleus has divided and one part (left center) seems to be forming a new macronucleus. (This individual offers the only evidence obtained of nuclear fragmentation and reconstruction through artificial means. ) This suggestive comment by Calkins, given merely in his description of plates, apparently hits the mark. We must inter- pret this depression animal of Calkins, which he attempted to ‘rejuvenate’ by phosphoric acid, as at the end of the climax of our process with many chromatin bodies and two macronuclear anlagen. The micronuclear condition is not clear from the photo- graph (compare text figs. 19 and 20, p. 484). Maupas’ observation that in certain hypotrichous forms the micronuclei in some periods may be increased to a number 484 LORANDE LOSS WOODRUFF AND RH. ERDMANN Text figures 19 and 20 beyond that typical for the species was substantiated by Wood- ruff (05) who observed marked macronuclear fragmentation as well as a tendency to micronuclear reduplication at periods of low division rate in a culture of Oxytricha fallax. The fact that Woodruff was able to ‘rejuvenate’ the race when apparently on 7 the verge of extinction led him to state that ‘‘we are hardly justified in assuming that Protozoa, when dividing at a low rate, with nuclei fragmented, etc., are exactly ‘abnormal,’”’ and “‘sug- gests that we are justified in regarding these changes as phases in the life history of Infusoria which occur under certain con- ditions after a considerable period of vegetative reproduction.”’ It now seems probable that this tentative suggestion points in the right direction. The discovery of atypical conditions, chiefly during the period of deepest depression and throughout the month that it persisted before ‘rejuvenation’ occurred, would seem to indicate that owing to unfavorable environmental con- ditions the race of Oxytricha was in an unhealthy state which resulted in a series of abortive attempts to reorganize itself, the final one of which was successful. PERIODIC REORGANIZATION IN PARAMAECIUM 485 Popoff (’07) found in the culture of Stylonychia mytilus, which he bred for three and one-half months, that after three periods of depression had occurred the race finally became extinct during a period of ‘‘sehr tiefe Depression.’ Popoff does not make a distinction between these various depressions, except in regard to their intensity. He records the fact that at these times char- acteristic nuclear changes as well as a tendency to conjugation were in evidence though both were most pronounced during the very deep depression. Again, in a culture of Paramaecium cau- datum, the data from which he does not present in as much detail, he found essentially similar depressions and nuclear meta- morphoses. He identifies the deepest periods of depression with those described by Maupas as ‘“‘dégénérescence sénile”’ by Hert- wig as ‘‘physiologischer Tod,” and by Calkins as ‘‘depression periods.”’ In a later study Popoff (’09) stated that he was able to produce, by treatment with various chemical reagents, identical variations in the nuclear apparatus of cultures of Stylonychia mytilus and Paramaecium caudatum, which in turn he considered similar to those observed by Kasanzeff in starved Infusoria. In other words, Popoff concluded that the nuclear phenomena in all his depression periods are exactly the same as those induced by starvation, chemical stimuli, ete., and further that they cannot be distinguished from those which occur at the onset of normal conjugation. In text figure 21 (Popoff’s fig. 25, pl. 2) is shown an animal with two micronuclei in mitosis from a culture of Paramaecium caudatum which Popoff had treated with ammo- nium. This specimen we would interpret as an early ‘reduction’ division in the reorganization process under discussion. We have found that it is possible to retard: or hasten the occurrence of the process by the character of the culture medium; for example, it may occur a few days earlier in animals not supplied daily with fresh culture fluid than in the regular lines. Consequently we can readily believe that treatment with the reagents employed by Popoff would influence its onset. Popoff figures specimens of Stylonychia mytilus in periods of depression and also after chemical treatment which he interprets as showing 486 LORANDE LOSS WOODRUFF AND RH. ERDMANN Text figure 21 the multiplication of micronuclei until there are a dozen or more in a cell. Although it seems clear that these represent stages in the process, it is impossible from his figures to be sure that many of the bodies which he interprets as micronuclei are not chromatin bodies which have arisen from the macronucleus as occurs in Paramaecium aurelia. However, Popoff’s paper shows that he recognizes a general similarity of the so-called depression phenomena and the cytological changes incident to the early stages of conjugation. Our data, secured by a daily cytological study of pedigreed animals, throw light upon Popoff’s isolated stages and indicate with considerable certainty that they are stages in the sequence of normal nuclear changes in the process which closely parallel conjugation both in morphological and physiological reatures. Gregory (’09) in a study of the life history of Tillina magna points out that ‘‘ . . . .~ the curve which represents the general vitality of the protoplasm shows the normal rhythmic fluctuations observed by Woodruff,’’ and in an analysis of the PERIODIC REORGANIZATION IN PARAMAECIUM 487 data secured by Popoff in his study of the life history of Sty- lonychia mytilus shows that ‘“‘if the curve of Stylonychia is plotted from average records of five and ten-day periods, it will be found to correspond to the curves for Paramaecium, Oxy- tricha and Tillina, each showing the rhythmic periods of high and low vitality.”” We reproduce here (text fig. 22) Gregory’s curve plotted from Popoff’s data, with the depression periods indicated. ‘This curve is obviously strikingly similar to the ones which we give in the present paper to show the relation between the process and the rhythms. It is to be noted that at the low points between Popoff’s ‘rhythms,’ just as at the low points between our rhythms, nuclear changes occurred. This agree- ment, taken in conjunction with the fact that Popoff’s animals show evidence of the process we describe, probably makes it now safe to state that Popoff’s small depression periods are rhythms. Such being the case, we are in a position to identify with more certainty the minor depression periods of Hertwig and Calkins as rhythms also. When one considers the diverse culture methods used by the various authors and that merely cells selected at relatively long intervals and practically at random were studied by them, it is remarkable that so many indications of the process may be gleaned from their data. This affords additional evidence of the general occurrence of this phenomenon in Infusoria. We do not desire, however, to deny that abnormal cytoplasmic and nuclear conditions can and do occur in infusorian cultures, or to seem to attempt to interpret as stages of the process all the so-called degeneration stages figured by various authors. Now and then an animal in the culture of Paramaecium under con- sideration has failed to divide for several days and has finally died without signs of the process. . IODIC REORGANIZATION IN PARAMAECIUM _—, , WOODRUFF AND RH. ERDMANN Woodruff, Erdmann and Bradley, del. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 17, No. 4 NOVEMBER, 1914 ie PLATE 2 EXPLANATION OF FIGURES Climax 14 Main culture J, Line ¢, 426th generation, April 9, 1908. Old macronu- cleus merely in the form of a membrane from which the numerous chromatin bodies have been ejected and are free in the cytoplasm. Eight so-called reduc- tion micronuclei two of which are lying isolated; the others in groups of three. 15 Main culture, Line b, 1498 to 1499th generation, December 12, 1909. A division stage at the climax of the process. Numerous chromatin bodies in each sister cell. One shows a membranous remnant of the old macronucleus. Each has only one micronucleus. 16 Mass culture started from main Culture I, October 12, 1913 (see page 435). One cell is in same stage as both cells in figure 15. The other cell shows the for- mation of the macronuclear anlagen, several chromatin bodies and two micro- nuclet. ‘ 17, 18 and 19 Culture B, Berlin race, October 6, 1913 (see page 473). Three animals illustrating the details of the formation of the macronuclear anlagen. Figure 17 shows two micronuclei, one macronuclear anlage ccmpleted and the other macronuclear anlage just beginning to be evolved from a micronucleus. Figure 18 shows two micronuclei and two completely formed macronuclear an- lagen. Figure 19 shows the same except that the micronuclei have divided for the next cell division. In these three figures the chromatin bodies have been omitted. The cells have been counterstained with eosin. 20 Subculture IE, Line Vla, 4084th generation, December 4, 1913. Animal shows two macronuclear anlagen which have lost their initia] form. Chromatin bodies beinning to disintegrate. Two micronuclei. 21 Subculture IE, Line VI, 4107th generation, December 16, 1913. The same stage as shown in figure 20, except that one of the micronuclei has completed divi- sion and the other is in process of division: 22 Subculture IE, Line III, 4063d zeneration, November 23, 1913. Essen- tially the same stage as former but showing both micronuclei in division. 23 Main culture I, Line c, 1755th generation, May 1, 1910. Animal figured to show that the complete reorganization of the cell occurred in the same manner at a relatively early pericd in the history of the main race. 24 Culture B, Berlin race, March 15, 1913. Animal figured to show that the complete reorganization occurred in an animal from a different race. Only one micronucleus is evident, the other being obscured by the chromatin bodies. 908 Se hen ‘PERIODIC REORGANIZATION IN PARAMAECIUM Lb. BiSD Daas AND RH. ERDMANN Woodruff, Erdmann and Bradley, del. THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 17, No. 4 NOVEMBER, 1914 PLATE 3 EXPLANATION OF FIGURES Ascending Phase (Except figures 31 to 34) 25 Subculture IE, Line VIb, 4174th generation, January 18, 1914. Animal after the first cell division. One macronuclear anlage. Some chromatin bodies in process of disintegration and enclosed by vacuoles. One micronucleus has divided and the other is in process of division for the ensuing cell division. 26 Subculture IE, Line VI, 4185th generation, January 25, 1914. A charac- teristic stage of the ascending phase of the process. New macronucleus with fine chromatin granules. Two micronuclei present in cell but one near end omitted from drawing. Chromatin bodies disintegrating. Cf. sister cells, text fig. 8. 27 Subculture IE, Line IIIb, 4313th generation, April 3, 1914. New macro- nucleus, two micronuclei, four chromatin bodies and the cloud-like chromatin remnants of the others. 28 Mainculture I, Lined, 1432 to 1433d generation, November 7, 1909. Young animals just after division during the ascending phase of the process. Micro- nuclei in each show various stages of division for the following cell division. The, cell division just being completed is the fifth in about twenty-four hours (see page 456). A few chromatin bodies still present. 29 Subculture IE, Line VI, 4187th generation, February 27, 1914. New macro- nucleus, two micronuclei and three chromatin bodies. 30 Main culture I, Line b, 1201st generation, June 7, 1909. New macronu- cleus, two micronuclei and two remaining chromatin bodies one of which is under the macronucleus. 31 Mass culture seeded from main Culture I. An epidemic of conjugation in progress. Animal with micronuclei in process of division. 32 Pair of conjugants, showing first micronuclear division after formation of the synearyon (cf. page 429 and page 453). 33. An exconjugant from isolated conjugating pair. First vegetative division completed. Macronuclear anlagen assuming typical macronuclear form. Note finely granular condition of chromatin typical of new macronucleus. Two micro- nuclei are present. Five food vacuoles with bacteria and promiscuous culture material in which conjugation was secured. 34 Paramaecium caudatum, Culture Y, Line 2, 139th generation, April 29, 1914. Descending phase of the process showing the formation of the chromatin bodies in this species. The one characteristic micronucleus of caudatum has divided. Only optical section of micronuclei is drawn. I RIODIC REORGANIZATION IN PARAMAECIUM L. L. WOODRUFF AND RH. BRDMANN Woodruff, Erdmann and Bradley, del. 513 THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 17, No. 4 NOVEMBER, 1914 514 : PLATE 4 EXPLANATION OF FIGURES Additional Stages 35 Subculture LE, Line VIb, 4092d generation, December 12, 1913. Typical form of macronuclear disintegration. Macronucleus small. Numerous chroma-. tin bodies. Three micronuclei present. 36 Subculture IE, Line VI, 4102d generation, December 14, 1913. Typical form of macronuclear disintegration. Macronucleus of typical size, practically devoid of chromatin. Numerous chromatin bodies. Four reduction micronuclei; (cf. figs. 1l and 12). Ao, text figure 4, is sister cell. 37 Subculture IE, Line VIa, 4087th generation, December 6, 1913. An atyp- ical form of macronuclear disintegration, slightly resembling the ribbon-like formation characteristic of conjugation. Beginning of micronuclear reduction. 38 Mass culture started from main Culture I, October 12, 1913. Transverse section through an animal at the climax. Seven chromatin bodies; two micro- nuclei in a cytoplasmic layer are visible. Stained with safranin. 39 Mass culture started from main Culture I, October 12, 1913. The pos- terior cell is in the same stage as both cells in figure 15. The anterior cell, after the formation of the macronuclear anlagen, has effected one more micro- nuclear division than the animal in figure 16a. The food vacuoles present contain remnants of material from hay infusion medium. 40 Subculture IE, Line Vth, 4355th generation, April 24, 1914. Stage at the end of the climax after the cell division characteristic of this period. Half the products of the next micronuclear division will become the two macronuclear anlagen. ; 41 and 42 Subculture IE, Line VIn, 4436th and 4437th generations, June 12, and 14,1914. Cell 4436 has completed the cell division in the climax. Numerous chromatin bodies and one micronucleus. The sister cell in the 4436th generation, which kept the line alive, formed the macronuclear anlagen and by a division in abcut forty-eight hours distributed these to each of the two cells of the 4437th generation. One of these cells is shown in figure 42. Numerous chromatin bodies, one macronuclear anlage and four micronuclei. The micronuclear divi- sion indicates the quick succession of cell divisions characteristic of the ascend- ing phase. 43 Subculture IE, Line VIq, 4437th generation, June 13, 1914. Animal after the first division in the ascending phase. New macronucleus with condensed chromatin. Three micronuclei. 44 Mass culture from main Culture I. Shows ribbon-like formation of macro- nucleus in aconjugating animal (cf. figs. 32and37). Eight reduction micronuclei, three of which are drawn. 516 a ae ~*~ . NIZATION IN PARAMAECIUM * Woodruff, Erdmann and Bradley, del. 517 THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 17, No. 4 NOVEMBER, 1914 THE REACTION OF EMBRYONIC CELLS TO SOLID STRUCTURES ROSS G. HARRISON Osborn Zodlogical Laboratory, Yale University FOURTEEN FIGURES It is generally recognized that the movement of cells and cell masses 1s an essential factor in morphogenesis.!. The most obvious movements concerned in the development of the Metazoa are those of cell aggregates, but these are often complicated by their association with growth or increase in mass. On the other hand, the movements of single cells, involving growth only to a very slight extent, are less complex and, at the same time, with our present methods of tissue culture, are readily amenable to observation and experiment. The mechanism of this move- ment is the streaming of the cell protoplasm, and the ontogenetic results depend upon the physical properties of the protoplasm itself and the stimuli which act upon it. The importance of such factors in development was fully recognized twenty years ago by Roux? in his experimental study of the behavior of isolated cells of the frog’s blastula, and by Herbst* in his ‘Reizphysiologie.’ Even before this Loeb* had considered the tropisms of cells, and had shown that the arrange- ment assumed by certain chromatophores in the fish embryo is dependent upon a stimulus emanating from the circulating blood. Somewhat later Driesch® also took up the question and found a 1 The various types of movement occurring in ontogeny have been carefully classified by Davenport in his ‘‘Preliminary catalogue of the processes concerned in ontogeny.’’ Bull. Mus. Comp. Zoédl. Harvard Coll., vol. 27, 1896. 2 Cytotropismus. Arch. f. Entw. Mech., Bd. 1, 1894; Bd. 3, 1896. 3 Biol. Zentralbl., Bd. 14 and 15, 1894, 1895. Also Formative Reize, Leipzig, 1901. 4 Jour. Morph., vol. 8, 1893. > Arch. f. Entw. Mech., Bd. 3, 1896. o21 THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 17, No. 4 Hoe ROSS G. HARRISON striking case of a cellular tropism (taktische Reizbarkeit) in the behavior of the mesenchyme cells of the sea urchin embryo. It must likewise not be forgotten that as early as 1892 Ramén y Cajal had stated his theory of chemotaxis to account for the outgrowth of the nerve fiber. Inexplicable as it may seem, very little of a definite nature has been added to our knowledge of this field since the period just cited. In my work on the development of peripheral nerves’ it was definitely shown in confirmation of Cajal’s theory, that the active factor in the spinning of the nerve fiber is a small mass of amoe- boid protoplasm at the end of a cell process. The formation of long fibers by means of this mechanism was found to be a char- acteristic of embryonic nervous tissue, but other somewhat different forms of activity of the same general nature were seen in all cells of the frog embryo when cultivated in clotted lymph.® No rigorous proof of the reaction of moving cells to definite stimuli was given in the course of these experiments. One of the most striking circumstances noted, however, was that when the lymph clotted firmly the movement was nearly always active, while in the cases in which the medium remained fluid only rounded cells were seen and these failed to undergo characteristic changes of form and locomotion. These observations led to the hypothesis that the cells were positively stereotropic and therefore unable to leave the solid masses to move out into purely fluid media. The work of Burrows with tissues of the chick embryo lent support to this hypothesis and it has since received further confirmation from other sources.’ Prior to this L. Loeb,!® ® La rétine des vertébrés. La Cellule, T. 9, 1892. 7 Anat. Rec., vol. 1, 1907; Jour. Exp. Zo6l., vol. 9, 1910. 8 Cells with processes of hyaline protoplasm were seen by Roux in the course of his work on cytotropism. The pseudopodia were described by him as occurring on cells that had become attached to the bottom. In one passage Roux states (op. cit., p. 163) that it is doubtful if any eytotropic movements take place without the solid base. * Burrows, Jour. Exp. Zo6él., vol. 10, 1911; Carrel and Burrows, Jour. Exp. Med., vol. 14, 1911; Holmes, Univ. California Pub., Zodlogy, vol. 11, 1913. 10 Arch. f. Entw. Mech., Bd. 6, 1897; Bd. 13, 1902; Zeitschr. f. Krebsforschung, Bd. 5, 1907. REACTION OF EMBRYONIC CELLS TO SOLIDS wae from his studies on the growth of epithelium and carcinoma, was likewise led to the view that these tissues are stereotropic. Nevertheless, none of these cases established the occurrence of this phenomenon beyond doubt. It seemed to be only a probable explanation of certain scattered facts. The experiments described in the present paper were designed to put the hypothesis to a more rigorous test. Examples of stereotropism (thigmotaxis) have long been known in plants and there is a wide range of reactions in animals that may be gathered under this head, some of which have been considered by J. Loeb! and by Verworn.!2 These reactions, are however carried out mostly by complex mechanisms and have but little direct resemblance to the cellular phenomena here considered. The best known case of positive stereotropism in single cells is that of the spermatozoa of Periplaneta, originally described by Dewitz,!* who called attention to its significance in the act of fertilization. Jennings’s™ description of the move- ments of amoeba indicate that here also reaction to solids plays an important part in locomotion. METHOD OF EXPERIMENTATION The plan of experimentation in the present work consisted in varying the three principal factors involved in the cultivation of tissues, viz., the tissue itself, the fluid medium, and the solid support. The results have shown that all three of these factors have some determining relation to the movement observed. Each tissue has certain characteristics, as seen especially in the differences manifested by epithelium, connective tissue and nervous tissue; the constitution of the fluid medium has a marked effect upon the vigor of the movement; the solid support influences the form and arrangement assumed by the moving cells, and its 11 Heliotropism der Tiere. Wirzburg, 1889; Studies in general physiology, Chicago, 1905; Dynamics of living matter, New York, 1906; Winterstein, Ver- gleichende Physiologie, Bd. 4, article on Tropisms. 12 Bewegung der lebendigen Substanz, Jena 1892; Allgemeine Physiologie. 18 Pfliiger’s Archiv., Bd. 38, 1886. 14 Behavior of the lower organisms. New York, 1906. 524 ROSS G. HARRISON indispensability is shown by the fact that no movement takes place in its absence. The first experiments were made with tissues from embryos of Rana palustris in the stage just after closure of the medullary folds. These were followed by a series of experiments with the tissues of chick embryos of various ages.!® The fluid media used were physiological sodium chloride, Locke’s and Ringer’s solutions in varying degrees of concentration, blood plasma and serum of the frog and the chicken respectively. The means used to support the planted tissues were the fibrin net work of the clotted plasma, spider web, the surface of the cover-glass, and, in some accidental cases, the surface film of the fluid drop. Control experiments were made in large drops of the fluid media, in which the tissue was prevented by the size of the drop from touching the cover-slip. At the time the experiments with frog tissues were made it was not fully realized that the cover-slip might serve to support moving cells, so that in respect to the solid support only cultures upon spider web and in free hanging drops were made. In some of the latter the pieces of tissue came into contact with the cover-slip accidentally and cells began to wander out on the glass. The technique used was in the main that employed in the study of the development of nerve fibers,'® together with the modifi- cations introduced by Burrows.!? Sterilized apparatus and fluids were used throughout. The cultures were all made by the hang- ing drop method, the cover-glass being placed over the hollow of a deep depression slide, or upon a thin glass ring of about 20 mm. diameter and 2 mm. height and sealed on with vaseline. The frog tissues were kept at room temperature, while those 15 A preliminary account of the experiments with frog embryo cells was pub- lished in Science, vol. 34, 1911. The results upon chick tissues have been referred to briefly in several general papers (see Anat. Rec., vol. 6, 1912; Trans. Am. Cong. Phys. and Surg., 1913). The chick experiments were made with the assist- ance of Mr. (now Dr.) Paul G. Shipley, to whom it gives me much pleasure to express my thanks. 16 Jour. Exp. Zo6l., vol. 9, 1910. 17 Jour. Am. Med. Assoc., vol. 65, 1910; Jour. Exp. Zoél., vol. 10, 1911. REACTION OF EMBRYONIC CELLS TO SOLIDS bYAS from the chick were incubated at about 39°C. The unusual hot weather which lasted during almost the whole period when the experiments with chick tissues were under way rendered unnecessary any precautions to keep the tissues warm during their preparation and examination. The preparation of the spider web for the experiments was the only innovation in technique that needs special description. It was necessary to have the web tensely spread over a suitable frame so that it would support the drop of fluid from below. For this purpose glass rings were employed, since they were easy to sterilize and well adapted to form the wall of a moist chamber. A number of rings were placed in clean glass aquarium Jars and a single spider introduced, which in the course of a day or two spun a web covering the whole bottom of the jar. The rings were then lifted out with forceps and the web cut or torn off around them. The slight roughness of the rim of the glass was sufficient to hold the web tense. The rings with the web upon them were sterilized by dry heat, the web standing a temperature of 150°C. without injury. In making the spider web preparations, the rings were first fastened to the slide with vaseline. A small drop of the culture fluid was then placed upon the web and the tissue afterwards introduced by means of a capillary pipette. It was unfortunately necessary to use very small drops of fluid since the weight of large drops caused the web to sag and touch the bottom of the chamber. The small size of the drops is sufficient to account for the relatively unfavorable results ob- tained by this method, the tissues usually not growing with very great luxuriance. After mounting the tissue upon the web, a cover-slip was placed over it and sealed on by vaseline. The weight of the cover-slip flattened out the small drop and the tissue was thus in contact both with the cover and the, web. In most cases the covers themselves were coated with web so that the tissue was kept between two layers of this material. During the early part of the season there was some difficulty in finding a sufficient number of suitable spiders for the purpose.'® 187 am greatly indebted to Prof. A. Petrunkevitch for advice and assistance in collecting this material. 526 ROSS G. HARRISON Tiginaria, which may be caught in old tree stumps, proved to be the best adapted of those which were obtainable early in the season, and two large specimens wove enough web for all of the experiments with frog tissues. Later the common grass spider, Agalena, was obtained in abundance. This form spins an ex- tremely fine but dense web which is admirably adapted for the purpose. The individual threads of the web are amazingly thin and even under the oil immersion lens they appear as fine lines, the thickness of which can scarcely be measured. EXPERIMENTS WITH. TISSUES OF THE FROG EMBRYO From this material seventy-one cultures were made in all. They are grouped in several series, of which the first two were of a preliminary nature, having been designed to ascertain what fluid media could be used. Series I. Only inorganic media were used, as follows: tap water, 0.325 per cent NaCl, Ringer’s solution without sugar, Locke’s solution half diluted. Solid support was afforded in all cases by spider web. The tissues used were: medullary cord alone, medullary cord with axial mesoderm attached, and ectoderm. Eight preparations showed cell movements. Of the six that gave negative results, three were those which were mounted in tap water; they disintegrated on the second day. Two were in the dilute sodium chloride. One specimen in dilute Locke’s solution lived seven days and the others from one to four days. Series II. In these experiments both inorganic solutions (Ringer and 0.4 per cent NaCl) and defibrinated serum were used. Thir- teen pieces were supported on web and eight were put up in hang- ing drops. None of the latter showed any movement of cells except one in which the drop touched the bottom of the chamber. In this some cell movement took place on the glass. Only one of the cultures in the saline solutions on the web gave positive indi- cations of movement, the others showing no promise from the beginning. Of the nine in serum five were on web and four unsupported. Four of the former showed active movement, one was evidently injured since histolysis of the tissue began REACTION OF EMBRYONIC CELLS TO SOLIDS 527 almost immediately. None of the unsupported ones gave positive results. Series III. This series of twenty-four experiments was de- signed to test rigorously the influence of the spider web upon the movements of the tissue, and as the preliminary experiments indicated that serum afforded a better medium than the inorganic solutions, the former alone was used. The serum was obtained from three different specimens of Rana clamitans, a species different from that of the embryonic tissue isolated. The cul- tures in this series were made in pairs. The central nervous system with some mesoderm attached was dissected entire out of the embryo in saline solution. It was then divided into two parts, one part being mounted upon web and the other in a plain hanging drop. In some cases the cephalic half of the medullary cord was placed on the web and the caudal half in the plain drop, while in others the order was reversed. The results of this series were entirely convincing. Out of the twelve preparations mounted upon the web eleven showed characteristic wandering of the cells with definite relations to the web fibers (fig. 1). The remaining one was disarranged accidentally the first day and gave no results. Cell movement began in some cases as early as eight hours after explantation and on the day following it was in full swing in all of the cultures. Of the twelve mounted in the large drops none showed any movement of cells except one of the cases in which the drop spread out, allowing the tissue to come into contact with the cover (fig. 3).!° In this one case six days after the culture was made a number of cells appeared on the cover and later some pigment cells were found. In the large drop preparations the isolated cells, which soon became very numerous,?° always remained rounded (fig. 2), but in spite of considerable disintegration into single cells the main mass of tissue was in every case left intact. That it was alive was 19 The figure, which is much like the preparation referred to, was drawn from another case. 20 Tn this respect embryonic frog tissues differ markedly from those of the chick, in which the elements remain closely bound together and the whole mass rounds off its outer surface (see p. 539 and figs. 11 and 14). Unless otherwise stated figures were drawn from living specimens. Fig. 1 Experiment 8. W. 33. Medullary tube of embryo of Rana palustris cultivated 27 hours on spider web in serum from R. Clamitans. Some cells have moved out in sheets on the cover-slip; others are adapted to the web fibers. 50. Fig. 2. Experiment 8S. W. 34. Medullary tube from same embryo, cultivated 28 hours in large hanging drop of same serum as in figure 1. Considerable dis- integration of tissue, but isolated cells are rounded and exhibit no active move- ments. X 50. Fig. 3 Experiment S. W. 66. Medullary tube (and notochord) of palustris embryo cultivated two days in Locke’s solution. The drop spread, leaving tissue in contact with cover. Detached cells are all on the glass. X 50. 528 iaose -! e ee . < uo”. eS “ue ad 4 a® 6 e* x) % @°o » 530 ROSS G. HARRISON shown by the fact that the cells of the myotomes differentiated into muscle fibers, which after about five days began to contract sporadically, indicating a functionally intact neuro-muscular mechanism. The results of this set of experiments are sum- marized in table 1. Series [V. The twelve preparations of this series were put up in the same way as the last except that Locke’s solution was used instead of serum. The results were not so conclusive as in the last set, though they presented no contradictory evidence. Only two of the six spider web preparations showed cell move- ment in any considerable degree, two others were doubtful, while the remaining two gave no promise of activity at any time. TABLE 1 Experiment III. Tissue in all cases medullary cord and myotomes CELL | STRIATED MUSCLE DURATION NO. CHARACTER OF PREPARATION eae PIBERS ere pe tan ae PLACE COUSIN SDD) OBSERVED PLACE } IN DAYS 31 Web se 8 32 Round drop + + 22 33 Web AF 4 34 Round drop 22 35 Web =F 3 36 Spread drop 12 ant Web + 2 38 Round drop + 22 39 Web + 29 40 | Round drop 8 a | Web + 6 42 | Round drop + 22 Ze Web a - 6 44 | Spread drop 12 45 Web + + 22 46 | Spread drop 12 AT | Web ar 12 48 | Round drop | + 22 49 | Web | =| + 29 50. | Round drop | ae + 22 Si Web! | 1 52 Spread drop? = | + 22 53 Web — | ; + 22 54 Spread drop | + 22 1 Injured. 2 A few outwandering cells found on cover. REACTION OF EMBRYONIC CELLS TO SOLIDS 531 In the cultures mounted in plain drops the tissue was found in two cases to be in contact with the cover-slip, upon which a few cells wandered out from the main mass (fig. 3). A small number of nerve fibers were seen growing out upon the cover in one case. None of the others showed any cell movement. It is clear from the experiments that the inorganic media are not so favorable for the cultivation of tissues as plasma or serum.?! Table 2 gives a summary of the experiments with frog tissues. The general character of the different types of culture is shown in the figures. Figures 1 and 2 represent two preparations from TABLE 2 Summary of results of experiments with frog tissues CELL WANDER- | PER CENT cuanscrn ov cuuuns | XEUMBRE OF |G occORRED | _ wo cm, | nORERTYE Ox SUPPORT | SUPPORT Defibrinated serum DIG erHweDecm acs ewes 17 15 2 | 88 Wareecdropes caer nie eee ete? 11 11 | Spread Gop. asco tee 5 1 4 | 20 Saline solutions | | | Spideriweure water re tee 251 | 11 14 44 argevdnope re ees on 5 42 | 5) 2 3 40 SJOMONG! ClO aa conenbceonoacct 1 Exclusive of three cases in tap water in which disintegration began very early. 2 In one case tissue touched bottom of chamber and some outwandering oc- curred. Series IIT mounted respectively upon spider web and in a plain drop. In the latter (fig. 2) there are a great many loose cells which have separated from the main mass of tissue and assumed spherical shape, while the main mass itself looks to be upon the point of disintegration, though, as subsequent observation showed, this did not take place. The drawing was made about twenty- eight hours after the tissue was prepared. In figure 1 the com- panion preparation is represented. The mass of tissue is held against the cover-slip by the spider web and is much more flat- 21 Cf. M. R. and W. H. Lewis, Anat. Rec., vol. 5, 1911. Boe ROSS G. HARRISON tened than in the other case. A comparatively thin fringe of cells has formed around it and many cells have left the mass entirely, being scattered at various distances from the latter. These cells are at two levels. The upper layer is in contact with the lower surface of the cover-slip, where the outwandering cells are apt to assume a polygonal form, though they are sometimes influenced by the web fibers attached to the cover. In the lower layer the cells are arranged with reference to the web fibers, and they are usually drawn out into long processes which are closely applied to the latter (see also figs. 4 to 7). The preparation shown in figure 3 is from Series ['V, being one of those mounted in Locke’s solution. It shows the effect of contact with the smooth cover-slip only. The drawing was made about forty-five hours after the culture was prepared. The tissue consisted of a piece of the medullary cord with some mesoderm. and a small piece of the notochord attached. It has been flattened in pancake form against the cover-slip by the spreading of the drop of fluid. A small number of isolated polygonal cells have moved off upon the cover, and a thin fringe of cells surrounds the tissue as in the preceding case. The adaptation of the cells to the web fibers is shown more clearly in figures 4 to 7, which are all taken from the same series of experiments. Figure 4 shows a cell with two hyaline processes attached to the web fibers running approximately at right angles. Figure 5 shows two bipolar and a tripolar cell, drawn when the preparation was two days old, figure 6 two cells from an eight day culture, and figure 7 pigment cells from another preparation of six days. The pigment is of two kinds, a granular black melanin and a diffuse yellow pigment (probably a lipochrome) present in two of the cells in large quantities. The yellow cells contain a little of the dark granular pigment, though much less than the other cells, and the latter contain none of the yellow. Nerve fibers developed in only a few of the cases and in none of these was anything notable shown. All of the nerves observed were growing upon the surface of the cover slip. The web fibers do not seem to be a favorable support for them. Fig. 4 Experiment S. W. 47. Cell from medullary cord with two processes attached to crossed web fibers. 300. Fig.5 Preparation 8. W. 53. Bipolar and tripolar cells from medullary cord attached to spider web fibers two days after explantation. 300. Fig. 6 Experiment 8. W. 39. Similar preparation eight days old. X 300. Fig. 7 Experiment 8. W. 49, showing the two types of pigment cells; six days old. x 300. 533 534 ROSS G. HARRISON EXPERIMENTS Wi1TH CHICK TISSUES One hundred and two experiments divided into ten series were made with chick tissue, comprising altogether 142 different cultures since in some cases more than one culture was made under the same cover. In three of the series the medium used was Locke’s solution, and in the other seven it was defibrinated serum. Clotted plasma was used for comparison in all series. When plasma and serum were used the blood was always taken from the same hen so as to have an identical fluid medium in all cases. Fixed support for the tissues was afforded by spider web prepared as described above, by clotted plasma and by the lower surface of the cover-slip. The latter was rendered available by mounting the tissue in small drops. In other cases contact was brought about after a time by the spreading of the drop which was originally large. This gives a firm attachment but rarely a luxuriant growth because of the extreme thinness of the fluid film.?2. Tissues from embryos varying in age from two to nine days were used. The results are given briefly in table 3, which shows the behavior of cells in the different media without, however, attempting to analyze the behavior of different tissues. Plasma preparations are 100 per cent positive. The cover- slip preparations (i.e., those mounted in small drops) show also a very high percentage of positive results while the spider web cultures are less favorable, due probably as pointed out above, to the extremely small amount of fluid used. The large drop preparations are all negative except two which show some cell wandering on the surface film. Several of the small drop cul- tures also show slight cell movement on the surface. The film cannot, however, be a very favorable surface for movement; otherwise we should expect to find cells upon it much more fre- 22 Burrows (Trans. Am. Cong. Phys. and Surg., 1913) has shown that the amount of cell migration depends upon the thickness of the layer of medium. There is an optimum thickness above and below which less movement takes place; in the thick layer because of insufficient oxygen and lesser concentration of repelling waste products (acid), and in the very thin layer (below the optimum) because of the small amount of nutrient medium. REACTION OF EMBRYONIC CELLS TO SOLIDS jae TABLE 3 Summary of results of experiments with chick lsswes Guido or coptans | oe neue /DEEINE ON carte, O50 eemo onl | enemas ARATIONS | SOLID FILM WANDERING ON SOLID SUPPORT SUPPORT PB lsismaac Lotie: Meer tie hace ea 34 34 100 Defibrinated serum Siderewebnctas ke sae eit oe 31 23 8 74 An COSCO Ds wat ees Ary eens 14 il 13 SHV Cher waweaa tages aeeee as 5 5 | _ 100 SJ ONACENGL(G U0) O). cel tatarm ms eas oun eee ce 6 bear 5 | 17 Locke’s solution | Spiderwebee. were eke as: 30) ii ao Pests ileal 63 Lareexdro peek 2 eee noe ee aes % | Abd Mi | Smalltdropr...a2.. ote ee a0 tere 12 10 2 1 83 SpreadadTopse ss eee aes 3 3 100 quently. It is of interest to note that in several of the large drop preparations where the tissue was at first not in contact with the cover, it afterwards did touch the glass and then cell movement began. These have been classed with the small drop (contact) group. The main results of these experiments can best be presented by the description of four different cases taken from the same series and shown in figures 8 to 11. The cultures are all from pieces of the duodenum of a nine-day chick embryo, and the drawings were made from specimens preserved two days after preparation. The first (fig. 8) has been cultivated in clotted plasma. In this there is a very characteristic ring formation with bands of tissue extending across the clear space. Isolated mesenchyme cells are present and the epithelium (endothelium?) shows its usual tendency to form membranes. The second (fig. 9) was cultivated in a small drop of defi- brinated serum and the tissue was in contact with the cover- slip from the beginning. The striking feature of this case is the formation of a wide membrane extending out from nearly the whole circumference of the original tissue. In one region conditions have prevented this movement, but at a distance from 536 Figs. 8 to 112% Duodenum of nine-day chick embryo cultivated with different kinds of solid support. Drawn from specimens preserved two days after explan- tation. X 39. Fig. 8 In clotted plasma. Fig. 9 In serum in contact with cover glass. Fig. 10 In serum on spider web. Fig. 11 Free in large hanging drop of serum. Fig. 1223 Portion of figure 10 under high power. X 300. Explanation of details in text. 23'These preparations, which were used to illustrate a general paper read before the American Congress of Physicians and Surgeons, have been redrawn (except fig. 31) for the present account in order to show greater exactness of detail. 537 THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 17, No. 4 538 ROSS G. HARRISON the main mass the membrane from the two adjacent sides has united, leaving a large hole, and there are also several smaller spaces left uncovered. A considerable number of isolated branched mesenchyme cells are to be found outside the area covered by the membrane. All of these structures are fixed to the cover-slip. The cells forming the membrane are one layer thick.2t. Those near the periphery of the membrane have the greatest superficial area, and there is a general decrease in their dimensions as the tissue mass is approached, though there are irregularities in this respect. There are no mitoses present, which indicates that membrane formation is due to movement of cells and not to growth.” There isno way of ascertaining whether this particular membrane is endothelial or from the epithelium lining the intestinal tract. The third culture (fig. 10) was supported upon spider web in the manner described above. It does not show such extensive cell wandering as some other preparations but it is very char- acteristic. A membrane (m) has been formed on the cover glass at one side of the preparation, and at another place, quite detached from the main mass, there is a small group of cells (a) which have moved out along the web fibers. On the opposite side of the tissue a considerable mass of cells (b) projects sharply outward and is firmly attached to the web. What is most inter- esting in this case is the banding together of spindle cells upon bundles of web fibers or upon single threads to form structures which closely simulate embryonic nerves with their sheath cells (c). In several places web fibers covered with such cells cross one another and here there is some accumulation of cells. Spindle shaped cells are also found singly (e). Some necrotic tissue (d) is present at places around the periphery of themain mass. The arrangement of these structures is shown in figure 12, which represents part of the preparation under higher magnification. 24 Cf. M. R. and W. H. Lewis, Anat. Rec., vol. 6, 1912. Also Burrows, 17th International Congress of Medicine, London, 1913; Section General Path. and Path. Anat., p. 225. 25 This is in harmony with the observations of many authors on the covering of wound surfaces by epithelium. REACTION OF EMBRYONIC CELLS TO SOLIDS 539 13 Fig. 13 Experiment C19. Piece of telencephalon of five-day chick embryo cultivated two days in clotted plasma, showing fringe of nerve fibers growing out into clot. X 39. Fig. 14 Experiment C15 Piece of telencephalon of five-day chick embryo cultivated two days in defibrinated serum. X 39. In the fourth case (fig. 11), which was cultivated in a large drop where the tissue never was in contact with the cover glass, there are no evidences of cell wandering whatever. The tissue is rounded off and has an almost smooth contour. One series of experiments was made to test the effect of the different kinds of support upon the growth of nerve fibers. Pieces of the telencephalon of a five-day embryo were used and twelve cultures were made, four of each kind, in a large drop, on spider web and in clotted plasma respectively. Those mounted in the plasma clot all showed the development of nerve fibers one day after explantation. In the more favorable cases they formed a veritable fringe around the whole periphery of the specimen. The fibers were much matted together and some were branched (fig. 13). Neither the preparations on web nor those 540 ROSS G. HARRISON in the large drop showed the same condition. Two of the former showed cell movement, i.e., some wandering upon the cover and the web fibers but no nerves were present. In all of the large drop preparations (fig. 14) the contour of the tissue was relatively smooth and no cells wandered out. These experi- ments tend to corroborate the results obtained with frog tissues, that the web fibers are not favorable to the outgrowth of nerves, though they are too few in number to be at all conclusive in this respect. DISCUSSION The foregoing experiments lead to the conclusion that solid objects are an important and even necessary factor in the move- ment of embryonic cells, such as mesenchyme and epithelium. Leaving out the cases of movement upon the surface film, which, moreover, has certain properties of solids, there are no exceptions to the rule that movement takes place only when contact with solid material is attained. Furthermore, each of the three kinds of solid support used in the experiments influences the cell move- ment in its own way, as is shown by the different arrangement assumed by the cells after a certain period of cultivation (figs. 8 to 10). The reactions to solids take place whether the fluid medium is a saline solution or serum, though the latter is conduc- ive to more active movement (and growth) than the former, owing no doubt to its nutritive qualities. The question now arises whether these reactions are to be re- garded as a manifestation of stereotropism (thigmotaxis), which is a response to mechanical stimulation (pressure), or whether the solid acts only indirectly by inducing conditions that give rise to chemical or some other form of stimulation. Burrows?? has shown that the centrifugal movement of cells observed in almost all cultures, i.e., the movement from the implanted cell mass out into the culture medium, may be explained by the acidity produced in the main mass of cells through the accumu- lation of waste products. That a condition of acidity does 26 Cf. M. R. and W. H. Lewis, Anat. Rec., vol. 6, 1912. 27 Proc. Am. Cong. Phys. and Surg., 1913. REACTION OF EMBRYONIC CELLS TO SOLIDS 541 obtain in the center of tissue cultures after a few hours of incu- bation has been proved by means of indicators.28 A given cell would thus be exposed on one side to a more acid medium diffus- ing from the tissue, and on the other to the more alkaline culture medium. This condition would, in accordance with the Quincke- Biitschli surface tension theory of amoeboid activity, bring about the formation of pseudopodia on the side of the cell turned toward the fresh medium and thus produce a centrifugal movement. Similarly the flattening of an isolated cell or the spreading of a sheet of cells on a smooth surface would be accounted for. Bur- rows thinks that no movement takes place as the result of mere contact with solids without the secondary stimulus resulting from the chemical change. These considerations at least show that even under very simple conditions, where no chemical stimuli have been intentionally applied, the local activities within a culture may nevertheless give rise to such conditions as might stimulate the cells chemi- cally on one side and thus direct cell movement. It however remains a fact that the chemical stimuli in question are pow- erless to call forth these movements in the absence of solid sup- port, else the large drop cultures would not behave as they do. The acidity theory also offers no adequate explanation of the adaptation of single cells to such minute structures as the web fibers, nor of the fact that outwandering cells rapidly bridge a gap between two separate pieces of tissue in the same culture. While it must therefore be admitted that chemical stimuli may play an important part in influencing the movements of cells in simple cultures, as Burrows has pointed out, the facts show that the cells are also stimulated by solids as such and respond to them by an orienting movement. Response to tactile stimuli is of such general occurrence in animals that there is nothing anomalous in the manifestation of the same kind of sensitiveness in cells. Though in the metazoa the responses are brought about by complex neuro-muscular mechanisms, the reactions to mechanical stimuli given by tissue cells and the protozoa are closely comparable with one another. 28 Cf, Burrows, ibid, and Rous, Jour. Exp. Med., vol. 18, 1913. 542 ROSS G. HARRISON The latter have been studied by Verworn?? in Rhizopods and by Putter? in Flagellates and Ciliates. These observers have shown that different parts of the same organism respond differ- ently to contact stimuli. In Protozoa, where the movement is sufficiently rapid to be directly observed, the location within the cell of both the sensitive region and the responding mechanism can often be made out, whereas it has as yet been found im- possible to do so in the case of the slowly moving tissue cells. This is a difficulty that applies alike to the study of the reactions to all kinds of stimuli and renders uncertain for the present any attempt at exact description of the process. At the same time the case of Chilomonas, for instance, in which according to Pitter stimulation of one of the flagella by solids calls forth a positive response, while similar stimulation of the other a negative one, shows that we have in this kind of sensitiveness an important directive factor. By analogy in the case of tissue cells, if dif- ferent parts of the same cell are unequally sensitive, it is likely that contact with an even surface would induce movement in a definite direction. Barring some such regional difference in the irritability of the cell, the kind of reaction called forth by con- tact with a solid object could only be a clinging to that object or a recoil from it. The former or positive reaction is that shown by the embryonic cells here studied. Without any further stimulus it is conceivable that the reaction to such contact might result in the change of shape of the cell, either its flattening upon the surface of a glass cover-slip or its elongation upon the fibers of the spider web (fig. 12). Its locomotion, however, must depend either upon some other stimulus or upon some local internal differences in the cell. As to the propriety of calling the reaction to solids a tropism, the position here taken differs from that of Loeb who holds that it is no real tropism, in as much as lines of force do not exist.*! It is true that in the case of tactile stimuli the source of the 29 Allegmeine Physiologie, 5th Edition, p. 519; Bewegung der lebendigen Substanz, Jena, 1892. 8° Arch. f. Anat. u. Physiol. physiol. Abt. Supplementband, 1900. 31 Dynamics of living matter, New York, p. 156. REACTION OF EMBRYONIC CELLS TO SOLIDS 543 stimulus is not placed at a distance from the organism, for it is the relative motion of the organism stimulated and the stimu- lating object acting only when they come into contact, that gives rise to the stimulus, and in the case of a moving organism coming into contact with a fixed object the energy is supplied by the organism itself. However this is not sufficient ground for draw- ing general distinction between the response to such stimuli and the reaction to a beam of light. The difference lies only in the source of the stimulus itself, i.e., in the distance from which it comes, and in the continuousness of its effect. In the case of a negative reaction to pressure the stimulus must cease as soon as the recoil is made. In case of a positive reaction the stimulus may continue but there can be no further visible effect beyond the change in shape of the organism unless its parts are differenti- ally sensitive. The similarity to phototropism becomes closer in the case of rheotropism or reaction to currents in the sur- rounding medium, which, as Verworn pointed out,® is to be classed under the group of responses to mechanical stimuli. A word remains to be said regarding the significance of the facts brought out in this study with reference to the interpretation of some of the phenomena of normal development. The re- semblance between the arrangement of cells on the web fibers (especially as seen in figure 12) and the sheath cells of an embry- onic nerve suggests that stereotropism may have something to do with the latter, though not necessarily to the exclusion of chemo- tropic influences. Similarly the close application of mesenchyme cells to such structures as blood vessels, muscles, and various other organs, resulting in the formation of a cellular sheath, which afterwards becomes skleratized, may be due in the first instance to a stereotropic response. In fact the conditions with- in the embryo at the time when the tissues are taken for the experiments are such as to suggest that this stereotropic reaction is an important factor in the behavior of many kinds of cells. There is a very active movement of mesenchyme at the time, during which cells stream from certain regions and fill in the ft 32 Allgemeine Physiologie. 544 ROSS G. HARRISON spaces between the main organ systems of the body which are already laid down. In this process the surface of structures such as the medullary cord, notochord, alimentary canal, muscle plates and the inner surface of the epidermis would serve as a solid base upon which the cells might creep. It is found that sooner or later all of these surfaces become ensheathed or lined by connective tissue cells. In the encystment of foreign bodies within the organism a similar phenomenon is observed. With regard to the movements of the growing nerve fiber the evidence, as pointed out above, is not quite so varied, but it is sufficient to warrant the conclusion that also this protoplasm is stereotropic. No free outgrowth of nerves in a fluid medium has ever been observed, while such solids as the fibrin clot and smooth glass surfaces serve readily to support them, as do the surfaces of the larger cell masses and the interstitial protoplasmic network inside the embryo. Most of these points were discussed some years ago by Herbst,** who, however, did not claim to have reached a definite conclusion as to the exact kind of reaction involved. The experiments here described do not of course settle the question either, but since it has been shown that most embryonic cells are stereotropic, and that such arrangements as they assume in the embryo may often be induced under cultural conditions by reactions to solids, there is a presumption in favor of the view that this type of re- action isa potent factor in normal development also. Inferences as to what goes on in the embryo which are not based upon exact information regarding the physiological properties of the tissue elements are likely to prove erroneous. On the other hand, if we know the actual properties of individual cells in detail, it will be possible to form, on the basis of the observation of normal development, an accurate conception of the influences actually at work in shaping the embryonic body. 33 Biol. Centralbl., Bd. 14, pp. 746. THE INFLUENCE OF FOOD IN CONTROLLING SEX IN HYDATINA SENTA DAVID DAY WHITNEY From the Biological Laboratory, Wesleyan University Considerable interest has been manifested concerning the sex ratio in the parthenogenetically produced male and female individuals in the rotifer Hydatina senta. At one time in general cultures of rotifers only females are found, at another time fe- males and males are found in equal numbers, and at still other times very few females but from 80 to 90 per cent of males are found. The problem has been to determine the causes that regulate the production of the two sexes in this rotifer—why at one time there is produced an excess of females and at another time an excess of males. The results of the work of earlier investigators on this problem having been reviewed so thoroughly and frequently recently, only references to them in the bibliography will be made at this time. Whitney made an extensive series of experiments and obser- vations in regard to temperature and starvation and came to the conclusion that neither factor was influential in the regu- lation of the production of the two sexes. He, moreover, found that the so-called female sex strains of Punnett could be made to produce many males by changing the environment. How- ever, he was unable to discover the real factor that changed the female sex strain into one that produced many males, but he was of the opinion that whatever the potent factor was that some- times caused only females to be produced and at other times caused nearly all males to be produced, it must be an external factor. Moreover, he made observations on one strain through 289 generations, for about two years, in which no males were 545 546 D. D. WHITNEY produced. This was accomplished by using a continuous and uniform diet of the flagellate, Polytoma, reared in a solution of cooked horse manure. Throughout this time in side experi- ments on this strain males were produced in large numbers many times by transferring the rotifers to a new mixed food culture. Shull carried on many experiments with temperature, starva- tion, and various chemicals and finally corroborated the general conclusions of Punnett and Whitney that temperature and starvation were negative factors in the regulation of the two sexes. He, however, was able to demonstrate that the unknown factor must act upon the grandmother and not upon the mother in order to cause abrupt changes in the sex ratio among the pro- geny. In this point he was in agreement with Maupas’ work. However, Shull finally concluded that this influential factor that controlled the production of the two sexes was an internal one. Mitchell in working with the rotifer, Asplanchna, recently has found that a continuous diet of a uniform food culture will cause only females to be produced, but if the diet is changed to a new food, males are produced in the third generation (grand- children). He suggested that a change of food might be found to be the controlling factor in regulating the production of the two sexes in Hydatina senta. In the experiments of the previous workers upon Hydatina senta very little attention has been given to any particular kind of food. Usually any mixed Protozoa culture upon which the rotifers would thrive was used. However, Shull and Whitney used a more or less pure culture of the colorless flagellate, Poly- toma, but obtained varying results. This may be due to the fact that these cultures in some instances were mixed cultures to a considerable extent although not suspected to be mixed. Whitney’s cultures were probably the same throughout the two- year period in which he obtained 289 generations of females. Uniform conditions produced uniform results. At that time this point was not fully appreciated. INFLUENCE OF FOOD IN CONTROLLING SEX 547 About five years ago the author had a general culture of rotifers in a jar which was placed in a south window in the labora- tory and remained there during the spring months. Some time in May during a period of a few days the jar was swarming with countless numbers of the green flagellate, Dunaliella, Teodor (Chlamydomonas, Cohn). On one day thirty-two female eggs of the rotifers were taken from the surface of the culture water and placed in watch-glasses and allowed to hatch. Each young female matured and thirty (93-+per cent) of these females pro- duced male offspring. This Dunaliella culture soon disappeared and no other good one ever appeared again. However, the author saved a little of the culture and spent the entire summer in the attempt to grow pure or even mixed cultures of it. No medium was found in which it would grow. This year another attempt has been made to grow cultures of Dunaliella and has been quite successful although the method is not yet completely perfected. They were raised in countless numbers in direct sunlight and a solution of bouillon. An Armour’s bouillon cube was dissolved and boiled in 400 cc. of tap water and equal parts of this bouillon solution and of steri- lized water were used for the culture medium. The medium was inoculated with a few individuals of Dunaliella and the jar placed in a south window. Within 7 to 14 days the surface of the culture water was swarming with the active Dunaliella during the sunny part of the day. Toward night the animals became inactive and motionless, in which state they remained until the following morning and then only became thoroughly active again provided there was direct sunlight. After the culture became 2 to 3 weeks of age the Dunaliella would remain continually in the motionless state. At this time the old culture water was siphoned off and some new culture medium was added to the quiescent Dunaliella. Within 2 to 3 days, if in sunlight, practically all the Dunaliella would be active. The surface was removed with a pipette and placed in a test-tube and by centrifuging, all the Dunaliella were collected at the bot- tom of the test-tube. Then nearly all of the liquid in the test- tube was poured off, thus leaving the Dunaliella in enormous dD48 D. D. WHITNEY numbers in a small amount of the culture water. Three or four drops of this material was added to about 8 cc. of old filtered culture water in which a general mixed culture of Protozoa and rotifers were living and placed in watch-glasses. These glasses were placed in sunlight at a temperature of about 25°C. and the Dunaliella remained active as long as they were in the sunlight. It, however, usually happened that when these watch-glasses were left in the sunlight for any appreciable length of time, the temperature would rise to 31 to 40°C., which would be fatal to the rotifers. To obviate this difficulty, a smaller battery jar was set and fastened into a larger battery jar which had padding in the bottom of it. This allowed the top of the smaller jar to protrude above the top of the larger jar. A space was thus made between the two jars into which running water was con- ducted, and by reguiating the flow the temperature could be kept at almost any point desired while this water-jacket jar was in direct sunlight. The rotifers in the Dunaliella culture were placed in the inner jar in watch-glasses and the jar was covered with a glass plate. By this method a constant temperature of about 25 to 26°C. was maintained, at which degrees the Dun- aliella were the most active in swimming about in the culture water. The rotifers eat them only when they are active, chiefly because when they are motionless they form a scum at the sur- face of the culture water or become fastened to the sides of the dish and in either place they are inaccessible to the rotifers. At the beginning of the experiments a female rotifer was taken from a jar containing a general culture of rotifers which had developed from fertilized eggs a few months previously. This female was fed upon pure cultures of the colorless flagellate, Polytoma, and became the progenitor of a parthenogenetic strain of rotifers which produced almost entirely female offspring through twenty-five generations. The method of causing a strain to produce only female offspring has been described in a former paper, but as the method of making the food cultures has been improved it will be briefly described again. Pure cultures of Polytoma were made in the following manner: 800 grams of fresh horse manure were put into 1200 ec. of water and INFLUENCE OF FOOD IN CONTROLLING SEX 549 cooked in a steam sterilizer at 10 to 15 pounds pressure for one hour. The liquid part was then pressed out and usually equalled 1000 cc. This was put into a flask with a cotton plug and sterilized. This could be kept indefinitely and used as a stock supply provided it was sterilized each time after it was opened. One part (100 ec.) of this hquid was added to three parts (300 ec.) of sterilized water, inoculated with a few Polytoma and placed in a large flat dish, thus giving a large surface of the culture water exposed to the air. Within 2 to 3 days the culture water would be teeming with the Polytoma but after a day or two the culture became spent and only a few Polytoma would be found. When this culture was about three days old one-half of it was poured off and 100 ce. of the stock supply of the manure Jiquid and 300 ce. of sterilized water were added. Eighteen to twenty-four hours later countless numbers of the Polytoma could be taken off from the surface with a pipette. This process of pouring off one- half of the old culture water and adding new medium was re- peated every day and in this way a vigorous food culture of Polytoma was continuously maintained for the rotifers at room temperature. Female rotifers of various ages were taken from the controlled female strain, which produced nearly all females through twenty- five generations by being fed upon a continuous diet of Poly- toma, and fed the green flagellate, Dunaliella, by the method as described earlier in this paper. Tables 1 to 5 show the important results obtained in a long series of observations. Experiments 1 to 23 have been omitted because they were made when the methods of feeding were being perfected and consequently did not show as favorable results as the later experiments. Experiments 24 to 63 include all the other experi- ments which were made in a successive series. A large number of individuals, considering the amount of work that each one entailed, has been observed, as recorded in tables 1 to 3. 341 mothers produced 5562 daughters, of which 3 per cent were male- producing daughters when reared under the influence of Poly- toma diet, and of which 57% were male-producing daughters 550 D. D. WHITNEY when the diet of their mothers was changed from Polytoma to Dunaliella. In the general control experiments in which a continuous diet of Polytoma was maintained 177 mothers produced 2500 daughters, of which 2+ per cent were male- producing daughters. The highest percentage of male-producing daughters which was produced by a change of the diet in the experiments in tables 1 to 3 is 83+ but as high a percentage as 87+ was produced in the experiment in table 4. If twenty-five of the best experiments in tables 1 to 3 are selected and the percentage computed for the male-producing daughters, it is seen that a change of diet from the Polytoma to TABLE 1 Showing that adult female-producing females of various ages, selected at random from a rotifer culture reared on a Polytoma diet through several generations, can be forced to produce a high percentage of male-producing daughters by being trans- ferred from a Polytoma diet to a Dunaliella diet | CONTROL REARED AND CONTINUED ADULT fe) os REARED ON A POLYTOMA DIET ON A POLYTOMA DIET TRANSFERRED TO A DUNALIELLA DIET EXPERIMENT a © a ios sue Offspring oO Offspring Ors |= = OS = 22 ofc) %SQ = O29 ae? GMS? 24 5 30 2 ee 5 23 19? 4 acaba 25 eh eee 2 6+ 5 27 12 30+ 26 5 36 0 0 18 130 AB |) 2524 27 2 20 0 0 2 12 So 40 28 5 17 1 5+ 5 29 | 15 | 344 29 5 30 0 0 5 Di 7 | 25+ 30 5 44 0 0 8 24 | 56 | 70 31 4 30 0 0 8 37 a see 32 4 24 0 0 20 100; © 160 i 432 33 5 40 0 0 9 32 32 | 50 34 5 33 1 one 12 ae 49 | 38+ 35 5 AT 5 9+ 22 54 | 1 | Ore 36 Br 458 2 25 e UP A 20 42 172 80+ 37 TW Go 0 0 15 50 122 | 70+ 38 5 65 3 4+ 16 84 156 65 39 5 | 104 2 1+ 10 64 | 837s e56s= 805 1086 | 57+ fa (o) > 2 ~I Te o ba | (op) — (0/0) [o) — [0 6) j=) INFLUENCE OF FOOD IN CONTROLLING SEX 551 the Dunaliella causes the percentage of male-producing daughters to change from almost zero to 63+. The average size of a family of daughters from each mother was 16+. The size of the families in tables 1 and 2 were inten- tionally made small because of the labor entailed had they been made larger. In table 3 the mothers seemed unable to bear the strain of being transferred from the Polytoma diet to the Dunaliella diet for the second time and usually soon died. It is possible that some of them were producers of small families, as Shull has shown in some of his experiments. TABLE 2 Showing that young adult female-producing females reared on a Polytoma diet will continue to produce female-producing daughters but can be forced to produce a high percentage of male-producing daughters by being put upon a Dunaliella diet and later can be forced to produce again female-producing daughters by being put back upon the Polytoma diet YOUNG ADULT 9 9s CONTROL REARED REARED ON A SRA ae SNP z = SAME ADULT 9 9S SAME ADULT 9 9S GN Apaursol a | cxaesensnnnb ae’|| NE AS A es EXPERI- DIET MENT SE = —-S = (oa a = = > —= = jor & Offspring o£ Offspring Zz Offspring z Offspring 0s] 8 | e.3)— eRe eau $35 OF |e EIS 56 ats) = eats) LS BP S/F 2) % re Lo 8/2 PIMP aD =12 Pic? 7S | S19 910191 GP 40 5/54 OF OD A} 5 640/0)A! 5 26 18) 40+ 41 5,58 0| O |A/ 5 50/0/0/A| 4] 18) 12) 40 42 5| 88} 0 0 {A} 5 440)0/A} 5) 22) 38 63+ B| 5) 44.0/0)|B) 5) 36 32) 47+ 43° | 5) 48) 0) OA) 8) 45) 07) OAL) 5) 85) 13) 27-5 | | 44 5122 32 20+ |A} 4 160'0/A'! 4 25 45 644+ /A!| 4 664) 5+ 45 | 51106 2 1+ ])A) 5) 260|/0/A| 5) 41) 25 374+ |A| 5 32/0] 0 46 5) 84 0 0 A} 5 160|)0/A) 5) 10) 26) 72-- |A)| 5 240) O PB 2 ON On sa OR has eB Te GON TO 47 5) 99) 9) 8+ /A} 5) 30;0/0)AJ 5 15) 55) 78+ | A) 5) 14) 2) 12+ 48 S100), OF O. | Al 5 36 0;0;A| 5 34 30) 464+ |A| 5 160; 0 | Bye 528) 0) 209) BaP 5 |iS5) 35) 50 Bal ol 26070 49 51106, 0| O |A| 518 0/0/A| 5] 34 14 204+ /A| 5] 42/0] 0 Bay LZ ON OR a ONG) Sif oul cite 50 5| 72) O| 0 A} 5 26,0|0)/A)} 5) 28 36 56-4 |A| 5 260) O B| 5 28'0/0)B) 5] 42) 26 38+ |B} 5) 240) O Total | 55937 43) 4+ 71485 0 0 70420418 49+- 46303 7) 2+ | az D. D. WHITNEY The temperature was uniform for all the individuals of each experiment. When the individuals were put upon a Dunaliella diet they were kept at 25 to 26°C. and the control individuals were also kept at this same temperature. During the other time both before and after the change in diet all individuals were kept at room temperature of 18 to 20°C. The results of Maupas are very easily explained in the light that is given by these experiments. Maupas probably used a TABLE 3 Showing that young adult female-producing females reared on a Polytoma diet can be forced to produce a high percentage of male-producing daughters among the first children of each family by being put upon a Dunaliella diet and later these same adult females can be forced to produce a high percentage of female-producing daughters by being changed back to a Polytoma diet and still later they can be forced to produce again male-producing daughters by being put back upon a Dunaliella diel ean aioe Bei a er SAME ADULT 2? 9S Rept ce ae Se | CONTINUED ON POLYTOMA DIET TRANS- TRANSFERRED BACK aa) x UN ALTETOA A POLYTOMA | FERRED TO A DUNA- TO A POLYTOMA DIET | DIET SSE DIET LIELLA DIET MENT : = = Sel ~ uo Offspring A oO, Offspring . . Offspring 3 a Offspring oe ro (4 eS | et [bes cE 7 BIO 2 9| 45| Pa/P Pic" P| % oP |g |2 21791 % we | 3 |S2P21% AD | | | | | | | <_< | 53 | 5/98] 22 |A| 5] 9] 39 814+ |A| 5] 28) 2 6+ | | IB} 5, 9| 47; 88+ |B] 5) 17) 7 29+ 54 5106, 43-+/A| 5] 21) 17) 444+ |A| 5] 15) 7] 81+ | |B 5] 11) 19} 63+ /B| 5 41} 9 18 55 6©| 5130] OO |A| 5) 17/17/50 | A) 5] 83) 7 11+ | A) 5) 75) 16) trp B| 5] 10 18 64+ |B/| 5] 51) 7) 12+ |B{ 5) 85) 13) 18+ 56 5| 75, OO |A| 5] 26 22) 45+ |A| 5) 36 4:10 | A} 5 19) 6 23+ B| 5] 29) 15) 34+ |B] 5) 34) 2) 5+ 57 5| 78 Ii+|A| 5] 19 37) 66+ |A]| 5) 14 4) 22+ |A| 1 2 2) 50 | |B| 5) 27 31) 53+ |A| 5] 28 4 124+ |B] 4) 25) 17) 40+ 58 5142) 21+) A| 5] 15) 39) 724 5} 35) 38} 7+] | B| 5| 36 50\ 57+ |B} 5) 61) 3) 4+ 59 5100 43+|A| 5| 36 4857+ A! 5) 60 0 O | | ‘By 5) 56 56 50 B| 5 64 0 0 60 5|L04 OO | A) 5| 25] 41) 62+ | A) 5) 47) 1) 2+ | |B| 5] 22) 16 424+ |B 5| 64! 0 0 61 | 5/80) 00 |A| 5] 30.30/50 |A| 5 60) O| O BI 5 16 12, 42+ |B} 5 42| 0| 0 Total | 45913 131+! | 901414554 57+ | | 90/750 60, 7+ 20206 54, 20+ INFLUENCE OF FOOD IN CONTROLLING SEX bao mixed food culture of Protozoa. When this food culture was placed at a low temperature of 12 to 14°C. only certain species of Protozoa were active and consequently could be used as food by the rotifers. However, when the temperature of this same food culture was raised to 26 +°C., other species of Protozoa which had been in a quiescent stage while the culture was at 12 to 14°C., now became active and were used as food by the rotifers instead of the Protozoa that were used at the low temperature. Thus by changing the temperature the diet was so markedly changed that it constituted the necessary stimulus upon the mothers for the production of male-producing daughters. TABLE 4 Showing that the influence of the diet acts upon the mother and not wpon the male- producing daughter that causes the daughter to produce males. In other words, the influence of the change of diet acts solely upon the grandmother and causes her to beget male grandchildren. Young 9 Qs reared on a Polytoma diet were trans- ferred to a Dunaliella diet for 27 to 86 hours. At the end of this period they” were transferred to filtered culture water and allowed to produce eggs. These eggs were transferred to Polytoma culture in which they hatched and the young females grew to maturity and reproduced while being fed exclusively upon Polytoma. | OFFSPRING FROM EGGS WHICH DEVELOPED ON A DUNALIELLA DIET BUT FED POLYTOMA 9 ie} MOTHERS REARED ON A POLYTOMA DIET | AFTER HATCHING ome) o'? %SF DEVE ASH S54 See war cee Hay OP RR SEES oe ie 8 16 66+ Several Wy Geta ee cae acre ae and an) Wet ee 9 63 87+ TABLE 5 As a control for table 4. Young 2 9s reared on Polytoma diet were transferred to a Dunaliella diet for 28 hours. At the end of this period they were transferred to filtered culture water and allowed to produce eggs. These eggs were trans- ferred to fresh Dunaliella culture in which they hatched and the young females grew to maturity and réproduced. OFFSPRING FROM EGGS WHICH DEVELOPED ON A DUNALIELLA DIET AND FED DUNA- 2 2 MOTHERS REARED ON A POLYTOMA DIET | LIELLA AFTER HATCHING : MeL: ohio) 9 | %SP Severali acu con tee are tee ase net a aes 35 25 41-+- Several ia it Waa hate nie Ook Rieke ean meats om ce m0) 10 | 20 THE JOURNAL OF EXPERIMENLAL ZOOLOGY, VOL. 17, No. 4 554. D. D. WHITNEY The author has several records of epidemics of males occurring in his experiments during the last eight years at periods when accidental rises of temperatures from 20+ to 26 or 27°C. took place and while the food used was a mixed culture of green and colorless Protozoa. These epidemics of males can be explained in the same manner as the results of Maupas. At room tempera- ture certain species of green Protoza were more or less quiescent but when the temperature rose suddenly to 26+°C. all the individuals of these somewhat quiescent green Protozoa became very active and furnished a new diet for the rotifers. The stimulation by this new diet caused the mothers to produce male-producing daughters. After a few hours the temperature sank back to the normal temperature of 20+°C. and the remain- ing green Protozoa again became quiescent and the rotifers were forced to eat the other Protozoa that were normally active at . this temperature and which was their regular diet. It has been previously observed that rotifers in a newly made general culture of manure medium produce a much higher per- centage of male individuals than rotifers in an old culture of manure medium. It is generally known that in a newly made hay infusion—and the same is true in a newly made manure infusion—the protozoan fauna fluctuates greatly. At first individuals of certain species may be very abundant and later individuals of other species became very numerous, while the individuals of the earlier-appearing species in the culture become relatively few. Thus there is a never-ending change in the protozoan fauna in a new culture of water and manure. Certain species flourish and are very abundant for a short period and then they disappear and new forms replace them. When rotifers are in such culture water with its varying protozoan fauna they are, of course, subjected to many changes of diet. Some of these changes of diet probably act as a stimulus upon the female roti- fers so as to cause them to produce male-producing daughters which produce males in the following generation. The sporadic production of males in the numerous experiments of various workers who have used mixed protozoan cultures as INFLUENCE OF FOOD IN CONTROLLING SEX BOD food for the rotifers can thus be simply explained. Under certain conditions some of the Protozoa are active and others are more or less quiescent, although they may be reproducing in this quiescent stage. When the conditions are changed, .possible in other ways than temperature, the quiescent Protozoa become very active, thus constituting a new diet for the rotifers which eventually causes males to be produced. It would be interesting to know the effect of a continuous feeding of a new diet upon the rotifers, but unfortunately it has been impossible to use the green Dunaliella continuously as a diet. They are the most active in sunlight but during the night they become more or less quiescent and consequently can not be used as a food by the rotifers in any appreciable numbers. * It is very probable that other forms of Protozoa as well as Dunaliella have the same stimulating effect upon female-pro- ducing females of Hydatina senta in causing them to produce male grandchildren because in mixed colorless protozoan cultures epidemics of males often occur. It may be possible to find and cultivate a colorless flagellate which will be even more effective than the green one, Dunaliella, which has already been used and caused the female-producing females to yield as high as 87+ per cent of male-producing daughters. The sex strains of Punnett may be more or less due to the diet used. Punnutt states that they were fed upon Euglena most of the time but does not state whether the Euglena were in pure cultures. In some of the food cultures made this year an undetermined species of Euglena was cultivated in the same kind of bouillon solution as was Dunaliella. This was an ex- cellent food for the rotifers but it did not stimulate them to produce males, as the Dunaliella always did when the rotifers were suddenly transferred to it from a Polytoma diet. In a previous paper it has been shown that a constant and uniform food supply caused one family to produce only female offspring through 289 generations, although males were produced in side experiments when the food was changed. Punnett unwittingly may have used a pure Euglena culture as food for some of his 556 D. D. WHITNEY strains and mixed cultures for other strains, or he may have used a mixed food culture whose protozoan fauna was stable or varied according to the light conditions or possibly other conditions in the laboratory. In order to cause rotifer mothers to produce male-producing daughters by changing the diet from Polytoma to Dunaliella the latter diet must be very abundant so that the mothers may consume enormous quantities of it. A change to a meagre diet of Dunaliella causes no male-producing daughters to be produced. This fact that the Dunaliella diet must be very copious in order that male-producing daughters may be produced, rather indicates that a sudden ¢éhange in metabolism in which the processes are carried on at their maximumrate, may be a necessary accompany- ing stimulus, coupled with the new diet stimulus that causes male-producing daughters to be produced. SUMMARY 1. In the parthenogenetic reproduction of Hydatina senta the influence of the diet acting upon the grandmother determines the sex of the grandchildren. 2. A continuous diet of the colorless flagellate, Polytoma, causes female grandchildren to be produced. 3. A sudden change of the diet from Polytoma to an abundant supply of the active green Dunaliella causes male grandchildren to be produced. 4. The regulation of the sex ratio in the parthenogenetic reproduction of Hydatina senta therefore can be controlled by food conditions. Middletown, Conn. August 14, 1914 INFLUENCE OF FOOD IN CONTROLLING SEX SOT BIBLIOGRAPHY Mavpas, M. 1890 a Sur la multiplication et la fécondation de | Hydatina senta Ehr. Comp. Rend. Acad. Sci., Paris, T. 111, pp. 310-312. 1890 b Sur las fécondation de l’Hydatina senta Ehr. Comp. Rend. Acad. Sci., Paris, T. 111, pp. 505-507. 1891 Sur la déterminisme de la sexualité chez ’ Hydatina senta. Comp. Rend. Acad. Sci., Paris, T. 113, pp. 388-390. Mircuety, C. W. 1913a Experimentally induced transitions in the morpho- logical characters of Asplanchna amphora Hudson, together with remarks on sexual reproduction. Jour. Exp. Zodl., vol. 15, no. 1, pp. 91-130. 1913 b Sex-determination in Asplanchna amphora. Jour. Exp. Zool., vol. 15, no. 2, pp. 225-255. Nussspaum, M. 1897 Die Entstehung des Geschlechts bei Hydatina senta. Archiv f. Mikr. Anat. u. Entw., Bd. 49, pp. 227-308. Punnett, R. C. 1906 Sex-determination in Hydatina, with some remarks on parthenogenesis. Proc. Roy. Soc., B, vol. 78, pp. 223-231, 1 plate. Suutu, A. F. 1910a The artificial production of the parthenogenetic and sexual phases of the life cvcle of Hydatina senta. Amer. Nat., vol. 44, pp. 146-150. 1910 b Studies in the life cycle of Hydatina senta. I. Artificial control of the transition from the parthenogenetic to the sexual method ef reproduction. Jour. Exp. Zodl., vol. 8, no. 3, pp. 311-354. 1911 Studies in the life cycle of Hydatina senta. IJ. The réle of tem- perature, of the chemical composition of the medium, and of internal factors upon the ratio of parthenogenetic to sexual forms. Jour. Exp. Zo6l., vol. 10, no. 2, February, pp. 117-166. 1912a III. Internal factors influencing the proportion of male- producers. Ibid., vol. 12, no. 2, February, pp. 283-317. 1912 b The influence of inbreeding on vigor in Hydatinasenta. Biol. Bull., vol. 24, no. 1, December, pp. 1-13. 1913 Inheritance in Hydatina senta. I. Viability of the resting eggs and the sex-ratio. Jour. Exp. Zo6l., vol. 15, no. 1, July, pp. 49-89. Wuitney, D. D. 1907 Determination of sex in Hydatina senta. Jour. Exp. Zool., vol. 5. 1909 a Observations on the maturation stages of the parthenogenetic _ and sexual eggs of Hydatina senta. Jour. Exp. Zoél., vol. 6, no. 1, January, pp. 137-146. 1909 b_ The effect of a centrifugal foree upon the development and sex of parthenogenetic eggs of Hydatina senta. Jour. Exp. Zodl., vol. 6, January, pp. 125-136. 558 D. D. WHITNEY Wuitney, D.D. 1910 The influence of external conditions upon the life cycle of Hydatina senta. Science, N.S8., vol. 32, no. 819, September 9. 1912a Strains in Hydatina senta. Biol. Bull., vol. 22. 1912 b Reinvigoration produced by cross-fertilization in Hydatina senta. Jour. Exp. Zodél., vol. 12. 1912 c Weak parthenogenetic races of Hydatina senta subjected to a varied environment. Biol. Bul., vol. 28. 1913 An explanation of the non-production of fertilized eggs by adult male-producing females in a species of Asplanchna. Biol. Bul., vol. 25. 1914 The production of males and females controlled by food con- ditions in Hydatina senta. Science, N. 8., vol. 39, no. 1014, June 5, pp. 832-833. SUBJECT AND MPHIBIANS when cultivated outside the body. The behavior of epidermis of 281 LOWEFLY larva (Calliphora erythroceph- ala Meigen). A quantitative determi- nation of the orienting reaction of the. . Body movements. On the early pulsations of the posterior lymph hearts in chick em- bryos: their relation to the.............. Browne, Eruet NicuHorson. The effects of centrifuging the spermatocyte cells of No- tonecta, with special reference to the MUMPLOCHON GMA meaner eee. cee eee Bursaria to food. II. Digestion and resorp- tion in the food vacuole, and further analysis of the process of extrusion. The TelAtlons Ole. Cente eee cee ELL fusion in Paramaecium. A normal periodic reorganization process without. Cells of Notonecta, with special reference to the mitochondria. The effects of centri- fuging the spermatocyte.................. Centrifuging the spermatocyte cells of Noto- necta, with special reference to the mito- chondria- ‘Theieffectsiof..-7;............. Chick embryos: their relation to the body movements. On the early pulsations of the posterior lymph hearts in Cuitp, C. M. Studies on the dynamics of morphogenesis and inheritance in experi- mental reproduction. VIII. Dynamic factors in head-determination in Planaria. 61 Chromosome of Drosophila. A gene for the fOUrbh ser eee eee es oe eee Chromosome studies in the Diptera. I. A preliminary survey of five different types of chromosome groups in the genus Dro- BOpHlas 25 ee sb saints na Ree crane CLARK, ELEANOR L., and CLarK, Exiot R. On the early pulsations of the posterior lymph hearts in chick embryos: their re- lation to the body movements , Eviot R., CLark, ELEANOR L. and. On the early pulsations of the posterior lymph hearts in chick embryos: their rela- tion to the body movements Cluster formation of spermatozoa caused by specific substances from eggs............. Controlling sex in Hydatina senta. The in- HuenceonLoodmneerra = re enactee eat: Crossing within and without an inconstant stock of Drosophila. Fertility and ster- ility in Drosophila ampelophila. III. Effects of crossing on fertility in Droso- phila. IV. Effects on fertility of......... Curtis, MAYNIE R., PEARL, RAYMOND and. Studies on the physiology of reproduc- tion in the domestic fowl. VIII. On some physiological effects of ligation, sec- tion, or removal of the oviduct 337 373 343 AUTHOR INDEX IGESTION and resorption in the food vacuole,and further analysis of the pro- cess of extrusion. The relations of Bursariatouoods: “lineman eres Diptera. I. A preliminary survey of five different types of chromosome groups in the genus Drosophila. Chromosome StuUdiesduttheks-cs5s.0nis cee oe eee Division rate of Paramaecium. The effect of thyroid onthe. nce ao eee eee Domestic fowl. VIII. On some physiological effects of ligation, section, or removal of the oviduct. Studies on the physiology of re- production in the Drosophila. A gene for the fourth chromo- SOME ONG. acon here te ee ete ee ; ampelophila. 1. Sterility in Droso- phila with especial reference to a defect in the female and its behavior in heredity. Hertilitivaena: stenlity. ile cee nee anne lon ampelophila. ITI. Fertility in Droso- phila and its behavior in heredity. Fer- tilitveand sterility 1... sales eee wees ampelophila. III. Effects of crossing on fertility in Drosophila. IV. Effects on fertility of crossing within and with- out an inconstant stock of Drosophila. Heriilipyuandustenilitysir. cen eee een nen and their influence on the sex-ratio. Two sex-linked lethal factors in Chromosome studies in the Diptera. I. A preliminary survey of five different types of chromosome groups in the genus. Dynamics of morphogenesis and inheritance in experimental reproduction. VIII. Dy- namic factors in head-determination in Planaria. Studies on the FFECTS of ligation, section, or removal of the oviduct. Studies on the physiol- ogy of reproduction in the domestic fowl. VIII. On some physiological... Embryonic cells to solid structures. The re- ACUONIOL eee,- selepe tae eatin is ciate eae Ee Embryos: their relation to the body move- ments. On the early pulsations of the posterior lymph hearts in chick Epidermis of amphibians when cultivated outside the body. The behavior of ERDMANN, RHopA, WoopRuFFr, LORANDE Loss and. A normal periodic reorganization process without cell fusion in Paramae- CLUE ES eee: ERA IEEE TUE EEC Cia ened Sere Extrusion. The relations of Bursaria to food. II. Digestion and resorption in the food vacuole, and further analysis of the pro- cess of BYERTILITY and sterility in Drosophila ampelophila. I, Sterility in Drosophila with especial reference to a defect in the female and its behavior in heredity... . 559 A third sex-linked lethal factor in... : 61 373 281 or 141 560 Fertility and sterility in Drosophila ampelo- phila. II. Fertility in Drosophila and its behavior in heredity..................-.. 173 and sterility in Drosophila ampelo- phila. III. Effects of crossing on fertility in Drosophila. IV. Effects on fertility of crossing within and without an inconstant stock? of MDrosophila ens. 2 ae ee 343 Food. Il. Digestion and resorption in the food vacuole, and further analysis of the process of extrusion. The relations of Bilrsariatoles oe ae eee eee. 1 in controlling sex in Hydatina senta. ‘Dheanfluence oles ca eae fee eee 545 Fourth chromosome of Drosophila. A gene boy eA eL (eee BAe Ta lk ot Late aan ae ob Se a 325 ENE for the fourth chromosome of Dro- Sophilae SU te edtescete sin te aes 325 ARRISON, Ross G. The reaction of embryonic cells to solid structures .... 521 Head-determination in Planaria. Studies on the dy namics of morphogenesis and in- heritance in experimental TE PPOCHEA VIII. Dynamic factors in...... 61 Heredity. Fertility and sterili Vy ‘in Droso- phila ampelophila. I. Sterility in Droso- phila with especial reference to a defect in the female and its behavior in............ 141 Fertility and sterility in Drosophila ampelophila. IT. Fertility in Drosophila andhits behavior mesee eeae ee eee 173 Hommes, 8. J. The behavior of epidermis of amphibians when cultivated outside the Odi s\. S87 5 os a cet ee em, ce ee hs Reon a ae 281 Hydatinasenta. The influence of food in con- Grol lamp sex INL Sioa pes rc san eee ee 545 Hype, Roscoe R. Fertility and sterility in Drosophila ampelophila. I. Sterility in Drosophila with especial reference to a defect in the female and its behavior in MEROGLG Yi, le hee oe eRe 141 IJ. Fertility in Drosophila and its be- havior.in hereditvas eee cee eee 173 III. Effects of crossing on fertility in Drosophila. IV. Effects on fertility of crossing within and without an inconstant stock/of Drosophila sean 2. akeenee 343 eae of food in controlling sex in tydstina senta:, “heys-5 oes cde. 545 Inheritance in experimental reproduction. VIII. Dynamic factors in head-determi- nationin Planaria. Studies on the dynam- ics of morphogenesis and................. 61 ARVA (Calliphora erythrocephala Mei- gen). A quantitative determination of the orienting reaction of the blowfly... .213 Lethal factor in Drosophila. A third sex- Jinkeahcre, Se BN tO Nee St he Rae oa en 315 factors in Drosophila and their influ- ence on the sex-ratio. Two sex-linked... 81 Logs, JAcqurEs. Cluster formation of spermat- o0zoa caused by specific substances from ORES so Satnc eeia'e Mtoe koe ected Cette etna ar 128 Lunp, EK. J. The relations of Bursaria to food. II. Digestion and resorption in the food vacuole, and further analysis of the process of extrusion.c: ease sates cee 1 Lymph hearts in chick embryos: their rela- tion to the body movements. On the early pulsations of the posterior........ . ale INDEX ETZ, Cuartes W. Chromosome studies in the Diptera. JI. A preliminary sur- vey of five different types of chromo- some groupsin the genus Drosophila. 45 Mitochondria. The effects of centrifuging the spermatocyte cells of Notonecta, Mae special reference to the................... Morean, T. H. A third sex-linked ae factor sin YDrosophilac.-8 2 earns 315 Two sex-linked lethal factors in Dro- sophila and their influence on the sex- Morphogenesis and inheritance in experi- mental reproduction. VUI. Dynamic factors in head-determination in Planaria. Studies on the dynamics of............... 61 Mutuer, HerMANN J. A gene for the fourth chromosome of Drosophila. . ee ee OTONECTA, with special reference to the m itochondria. The effects of centrifug- ing the spermatocyte cells of .......... 337 RIENTING reaction of the blowfly larva (Calliphora erythrocephala Meigen). A quantitative determination of the..... 213 Oviduct. Studies on the physiology of repro- duction in the domestic fowl. VIII. On some physiological effects of ligation, sec- tion, or removal of the YS A normal periodic re- ore process without cell fusion “The effect of thyroid on the division VALCLOLn 2 5 ae oe estes Es re ee ee 297 Patten, BRapLEY M. A quantitative deter- mination of the orienting reaction of the blowfly larva (Calliphora erythrocephala Meieen) i Me sme ae tee Se oe ome PEARL, RayMonp, and Curtis, MayYniE R. Studies on the physiology of reproduction in the domestic fowl. VIII. On some physiological effects of ligation, section, or removal of the oviduct................ 395 Periodic reorganization process without cell fusion in Paramaecium. A normal...... 425 Physiology of reproduction in the domestic fowl. VI1l. On some physiological ef- fecis of ligation, section, or removal of the oviduct. Studies on the................. 395 Planaria. Studies on the dynamics of mor- phogenesis and inheritance in experi- mental reproduction. VIII. Dynamic fac- tors in head-determination in............ 61 Pulsations of the posterior lymph hearts in chick embryos: their relation to the body movements. On the early................ 373 UANTITATIVE determination of the orienting reaction of the blowfly larva (Calliphora erythrocephala Meigen).A 213 EACTION of embryonic cells to solid Abructtines. WeNe ef 2.keiern cc tue ee 521 Removal of the oviduct. Studies on the physiology of reproduction in the domes- tic fowl. VIIll. On some _ physiological effects of ligation, section, or............. 395 Reorganization process without cel] fusion in Paramaecium. A normal periodic....... 425 Reproduction in the domestic fowl. VII]. On some physiological effects of ligation, settion, or removal of the oviduct. Studies onthe physiolepy of. sso toeeee eae 395 Reproduction. VIII. Dynamic _ factors in head-determination in Planaria. Studies on the dynamics of morphogenesis and inheritance i in experimental. . Resorption in the food vacuole, ‘and ‘further analysis of the process of extrusion. The relations of Busaria to food. II. Diges- CIOMMATG. 2 hoe Re ee noses omic ECTION, or removal of the oviduct. Studies onthe physiology of reproduc- tion in the domestic fowl. VIII. On some physiological effects of ligation. . Senta. The influence of food in controlling INDEX 61 395 sexin Hydatina..........0...-..+-2..-05. 545 Sex in Hydatina senta. The influence of food RI CONETOMAN Bey eee torte ee late tess rei Laious Bertieeed lethal factor in Drosophila. A Sex linked lethal factors in Drosophila and their influence on the sex-ratio. Two Sex-ratio. Two sex-linked lethal factors in Drosophila and their influence on the. . SHumway, Waupo. The effect of thyroid on the division rate of Paramaecium........ Solid structures. The reaction of embryonic 81 297 Cells ts seer rn ae See ere hae oe eee 521 Spermatocyte cells of Notonecta, with special referencetothe mitochondria. The effects Ofcentrihupine thes. -cee.ce tec eae 337 Spermatozoa caused by specific substances from eggs. Cluster formation of......... Sterility in Drosophila ampelophila. ility in Drosophila with especial reference to a defect in the female and its behavior in heredity. Fertility and................ Il. Fertility in Drosophila and its behavior in heredity. Fertility and..... . III. Effects of crossing on fertility in Drosophila. IV. Effects on fertility of crossing within and without an inconstant stock of Drosophila. Fertility and...... HYROID on the division rate of Para- maecium. The effect of ............. ACUOLE, andfurtheranalysisof the proc- ess of extrusion. The relations of Bur- saria to food. II. Digestion and resorp- GiOnwin ENS fOOd) - see cess see eee HITNEY, Davin D. The influence of food in controlling sex in Hydatina OEE AS RAIS Som Ati TO SOME SOE S Wooprurr, LORANDE Loss, and ERDMANN, Ruopa. Anormal periodic reorganization process without cell fusion in Paramae- 561 123 MONTANE a Bi: E 02028 Kone ep SIRE: pale seta cs Se ehe His sf) eee: yt 4 vend: if ater eere see ~ baie 7 t Fit tad gap aT eC: ‘ zi % AL SD PA AIGA : 4] eh i Bre fie Eo Bie a ¥ Bit uf on i cetera tu rere eset es ee rete ae tis ‘) a ate acs ape esate bate tasend Merete ee es Pi? bry ’ ‘ ee ae Sata ral ‘et Lt od aE pS Ne ea Hh AEE. ees es pious