THE Ms ) | - A AMERICAN NATURALIST A Bi-MowTHLY JOURNAL DEVOTED TO THE ADVANCEMENT OF THE BIOLOGICAL SCIENCES WITH SPECIAL REFERENCE TO THE FACTORS OF EVOLUTION VOLUME LVI "VN NEW YORK THE SCIENCE PRESS™” 1922 OCT 1 1923 G i hate n d IRDen \ BRP THE AMERICAN NATURALIST Vor. LVI. January-February No. 642 THE ORIGIN OF VARIATIONS SYMPOSIUM AT THE THIRTY-NINTH ANNUAL MEETING OF THE AMERICAN SOCIETY OF NATURALISTS, TORONTO, DECEMBER 29, 1921 VARIATION IN UNIPARENTAL REPRODUCTION PROFESSOR H. S. JENNINGS THE JOHNS HOPKINS UNIVERSITY Darwinism left the origin of variations the unsolved problem. Give us inherited variations, it said, and we can explain adaptation, by natural selection. But this was the omission of 99 per cent., if not 100 per cent., of the problem of evolution. Are we in better case to-day? Has the experimental study of geneties given us some solid knowledge of the origin, the causes, of variation? Have we learned that the obvious differences observable everywhere among individuals are the foundations of evolution? Or that they are not? Are slight quantita- tive fluetuations the material out of whieh evolution is made? Have we discovered that extensive saltations are the steps in evolution? Or that less extensive mutations, qualitative or chemical changes, that may be minute or large, are that by which evolution is constituted? Do we know the origin of such saltations, mutations? Have we found that the present constitution of the organism pre- determines in some way the course of further change; or that an élan vital is driving the organism to unfold in a definite way, like a flower; that evolution is orthogenesis? Do we comprehend the nature and causes of such a push to unfold, and of the direetion in whieh it tends? Have we found perhaps, as at least one investigator maintains, 5 6 THE AMERICAN NATURALIST [Vor. LVI that it is mixing of stocks, hybridization, that is the origin of organie diversity? Or do we know that the physieal and chemical conditions of the environment produce changes that are inherited and give us evolution? Or finally, do several or all of these methods of action con- cur? Such are the questions, I take it, on which we hope for light in the discussion this afternoon, and in the discus- sion of orthogenesis before the American Society of Zo- ologists, and of the species concept before the Botanical Section of the American Association. The lines of attack on the problem of variation are as manifold as are the questions to be answered. The basic idea in that attack whose results I shall try to summarize is this: In the reproduction from two parents familiar to us in higher animals and plants, there is a mixing of dif- ferent stocks, a formation of great numbers of diverse groups of the hereditary materials, with consequent pro- duction of a great variety of diverse offspring from a given pair of parents. This is the chief cause of the dif- ferences everywhere observable among individuals: dif- ferences formerly classed as variations and considered the material of evolutionary change. But such kaleido- scopic regrouping of materials, the units of which are not changing, has no obvious relation with evolutionary vari- ation; in the next generation a new grouping of the same - material occurs, and so on indefinitely. If there likewise occur progressive evolutionary changes, these are so lost, so hidden, in the multitude of kaleidoseopie recombina- tions that they ean not be distinguished; the literature of evolution is filled with eonfusion due to this difficulty. Therefore the idea suggests itself: Why not avoid at once all this, by studving evolutionary ehanges in those organisms where no mixing of stocks is oecurring; where there is no kaleidoscopie regrouping of the hereditary materials? "There are organisms that reproduce from a single parent, with no shifting or recombination of the germ plasm; in these, actual changes that persist from No. 642] VARIATION IN REPRODUCTION 7 generation to generation, such as evolution requires, should lie open before us, unconfused. We should see evolution occurring, as we see water flowing. This plain, simple and optimistic maxim has, I fear, like many another such, not proved so illuminating as its promise. But led by it, investigators set themselves at the study of the passage of generations, with selection and propagation of individuals showing diversities, in these creatures where seemingly all lasting change from parent to offspring must be evolutionary. Their hope was to see evolution occurring. And what did they see? I need not review the details; Johannsen, Barber, Hanel, the present writer, Lashley, Agar, and others, followed for long periods the passage of generations in many dif- ferent organisms during uniparental reproduction. Their report, after years of work, was astonishingly simple and clear. As to the origin of hereditary varia- tions, it resembled the famous chapter on the Snakes of Ireland. It summed itself, in effect, in the succinct, suffi- cient, exhaustive proposition that there is no inherited variation; hence no origin of such variation. There is nothing to find out about it, for it doesn’t occur. The in- dividuals produced in uniparental reproduction may in- deed differ, but these diversities are transitory effects of environmental differences; they are not inherited. All the descendants of a single individual are genetically and hereditarily alike; they form in effect a set of identical’ twins. And from this it could be concluded that in bi- parental reproduction all the observed diversities are due to the kaleidoseopie regrouping of hereditary ma- terials; nothing to evolutionary change. Outeries—objurgations and aeeclamations— greeted these propositions. Some reviled them for their mani- fest absurdity, others acclaimed them for their obvious truth and the clarification they wrought. Opponents tried to disprove them by investigating the matter them- selves; their evidence strengthened the propositions they 8 THE AMERICAN NATURALIST [Vor. LVI had thought to overthrow; who came to scoff remained to mourn. Such was, in the gross, the upshot of the first phase of the study of uniparental inheritance; of perhaps the first ten years. But the matter could not rest here. This work cleared the ground. It showed that 99 per cent. or more of what had been ealled variation had nothing to do with evolutionary change—a?' conclusion which Mende- lian study was reaching independently. Now it remained to aecept that faet, to take a new hold, to grapple with the more diffieult question: Is there yet an infinitesimal residuum of evolutionary change? If we select the most favorable organisms, and study them in most minute de- tail for suffieiently long series of generations, shall we indeed find that there are no persistent variations what- ever? Such is the work that has in this field occupied, with redoubled intensity, the last ten years. What are the results of this second phase of the work? | Some of the workers devoted themselves to observa- tional breeding work on the passage of many generations, accompanied by selection; others attempted to modify the inherited characters by physical and chemical agents. In the observational search for persisting alterations, with the attempt to accumulate their results by selection, we find, first, that many of the organisms studied have as yet defied all attempts to find any inherited variations. Such is the report of Ewing on his extended work with aphids ; such is the case with the fungi studied by Brierly (1920). Such is the case with most of the strains of the infusorian Paramecium, studied in detail for long periods by many different observers. Only in certain deformed strains, and possibly in one or two other instances, has the occurrence of persisting variation been observed in animals living under the usual conditions. Such is the case with the great majority of the strains of the Clado- cera studied with such extraordinary thoroughness for long periods by Banta (1921); out of 16 strains to which selection was applied for many generations, all but one No. 642] VARIATION IN REPRODUCTION 9 gave on the whole negative results; they did not change. Some of the investigators still insist that this is indeed the outeome of all this work; that all eases seeming to give other results are for one reason or another decep- tive; that no hereditary variations occur; that evolution- ary change has not been observed in this sort of repro- duetion— and presumably therefore in no other sort. Thus, for example, argue Brierly (1919), and, in effect, Vietor Jollos (1921). On the other hand, in some of the organisms studied, visible changes persisting from generation to generation of uniparental reproduction have been observed. Even in the first period of this sort of work, extremely rare ** mutations ". were reported by Barber in his work on baeteria, an apparent single one by Lashley in Hydra; a ** bud variation "" or two by Johannsen; and other iso- lated cases occurred. In the second period of the work, as a matter of observational faet, whatever the interpre- tation, it is certain that in the lowest Rhizopoda: in Dif- flugia, in Centropyxis, in Arcella; in the infusorian Stylo- nychia, and in certain abnormal strains of Paramecium, as studied in our laboratory at the Johns Hopkins Uni- versity, there arise in uniparental reproduction, changes affecting both physiological and structural characters; changes that may be very slight, or of great extent; that are passed on to later generations in uniparental repro- duction. By selection and breeding of the changed indi- viduals, stocks are isolated which differ persistently from the stock with which the work of breeding began. In this way might well arise the diverse biotypes found in nature to occur within a species, in these organisms. Something similar was found by Stout in the propagation of certain plants by cuttings. Again, among the 16 strains of Cladocera, subjected by Banta to selection for a physiological characteristic, one, and only one, showed persisting alterations, accumulated by the selective process, so that from the single strain, two continuously diverse strains were produced. Jollos 10 THE AMERICAN NATURALIST: [Vor.LVI too has observed a few cases in which strains of Parame- cium became differentiated in ways that could hardly be considered the result of environmental action. Doubtless some other cases might be collected. Here then we seem to have what we were searching for; here at last is some- thing solid; here by our presuppositions we have evolu- tion evolving; we have seen it! But as with so many of the seeming solid things of science—so these became sicklied o'er with a pale cast of thought, of doubt, of speculation. What, it is asked, is the cause, the funda- mental nature, of these persistent changes? And are they indeed of a sort to be considered steps in evolution? And when we look closely, the observational and selec- tional work has given us little information on these points. In Arcella Hegner found that certain of the in- herited structural changes are mere results of increase or decrease in number of nuclei, brought about in a simple manner. But most of the changes in the lower organisms studied can not be accounted for in this way. The work of Erdmann (1920) indicates that certain persistent changes occur in Paramecium as a result of the periodic nuclear reorganization called endomixis. These would perhaps have only a significance similar to that of the recombinations occurring in biparental inheritance. It is a favorite speculative idea with opposing speculators that most or all of the persisting changes we have men- tioned arise through irregularities in nuclear division, and hence are of little evolutionary significance, but this is thus far a mere possibility, without solid base; as the Germans say, it floats in the air. Another speculative notion is that the changes lack permanence; that if fol- lowed for a sufficiently great number of generations there would be reversion to the original condition. Whether this doubt can ever be resolved by observation ean not be predieted; it depends perhaps on the number of genera- tions demanded by the doubters. In the attempts to modify inherited charaeters by physical and chemical agents, more positive evidence as No. 642] VARIATION IN REPRODUCTION 11 to the cause of variation has perhaps resulted. Can we not, it is asked, by subjecting the hereditary material to chemicals, to physical agents, alter it, as we can alter practically everything else in nature? Of course we can; it is easy. But when we alter it we usually kill it, or pre- vent it from developing; our task is like that consumma- tion devoutly to be wished, of killing the pathogenic bac- teria in a man—which is easy—but it also kills the man! Have we succeeded in so altering the germ plasm, without killing it, that it now develops differently, and transmits the diversity to its progeny? It is easy by altering the chemical and physical condi- tions to change tremendously the development and char- acteristics of these creatures, and that without stopping life and reproduction. But in the infinitely greater pro- portion of cases such changes have no inherited effect; so soon as these particular conditions are removed, the progeny go back at once to the usual constitution. Such has been the result of extensive experiments of my own in modifying Paramecium with chemicals; and of Noyes in modifying Rotifera. Startling transformations of form, structure and function are readily produced and kept up for generations, but disappear when the offspring are reared under normal conditions. Once in our work the task seemed accomplished. After many generations of treatment with aleohol, Paramecium yielded monstros- ities and deformities, analogous to those Stockard ob- tained by the same method in guinea pigs, and these de- formities were transmitted after removal from alcohol, for generation after generation. This was stirring; all the energy of the laboratory was devoted to following the monstrous stock through long periods, leaving the formu- lation of pedigrees till time permitted. But when this could be done it appeared that all these abnormal indi- viduals came from one single ancestor, out of the hun- dreds with which the experiment began; the rest had all returned at once to normal. We know that such heredi- tarily abnormal stocks occur at times in Paramecium, 12 THE AMERICAN NATURALIST [Vor. LVI produced in some frequency by an agency which takes the matter at once out of the field with which I am dealing —by the recombinations occurring at conjugation, at bi- * parental reproduction. Our monstrous stock may have come from such an individual, included by accident in the experiment. Our spirit-stirring results faded into noth- ingness—a type of what has so often happened in promis- ing work in the inheritance of environmental effects, of what will probably often happen again. Other workers have been more successful. In ‘the bac- teria, if we can accept the accounts given by many investi- gators, and well summarized, for example, in Adami’s ** Medical Contributions to the Theory of Evolution," environmental conditions frequently alter, in an adaptive way, the persisting characteristics of the stocks, differen- tiating a single race into several. The difficulties of cer- tainly working with unmixed strains is very great in these minute creatures, a fact which leads many students of experimental evolution to reject generalizations based on these organisms. Further, the extraordinary work of Lóhnis (1921), recently published by the National Academy, tends, if substantiated, to so completely upset all supposed knowledge of life history in the bacteria that it will be best to omit these from consideration until the air is cleared. For similar reasons, and from considera- tions of space, I will not speak of the work on pathogenic Protozoa. Turning then to those larger organisms that are iso- lated with as much ease as are guinea pigs, Middleton has found that differences of vigor and of rate of reproduction are produced by subjection of infusoria for long periods to diverse temperatures, and are perpetuated, after equalizing the temperatures, from generation to genera- tion for long periods, and through the process of conju- gation. At this meeting he has reported similar results produced by subjection to diverse chemicals. How far this is comparable to change of other characteristics than reproductive vigor we do not know. No. 642] VARIATION IN REPRODUCTION 13 Vietor Jollos (1921) has just published in this field work which must make a deep impression on the study of experimental evolution, work which gives us more posi- tive results than have before been achieved. By experi- mentation extending over years he has, by subjection for long periods of time, altered the resistance of the infu- sorian Paramecium to certain chemicals, and to heat. After removal of the causative agent these physiological changes are passed on from generation to generation of uniparental reproduction, for longer or shorter periods. Of extreme interest is the fact that longer subjection to the altering agent causes longer persistence after the agent is removed. . The induced changes lasted in some eases for hundreds of generations, not yielding at the periodic nuclear reorganizations known as endomixis. But the acquired resistance in practically all cases finally disappeared if the organisms were continued sufficiently long in the normal conditions. Subjection to frequently varied environment hastened the disappearance of the persisting effect ; and it usually disappeared at once when there occurred the profound reorganization accompany- ing conjugation and biparental reproduction. But in some cases, as in Middleton’s results, the acquired re- sistance lasted through conjugation; even through several cycles of conjugation. But in all cases in which it was clear that he was dealing with resistance acquired through subjection to chemical or physical agents, it finally disappeared, after hundreds of generations, if the organisms were kept sufficiently long in an environment lacking the causative agent. Jollos is from this inclined to draw the conclusion that the changes are not com- parable to the (assumedly) permanent differences that . separate genotypes or species, and hence that they do not indicate a method by which such permanent differences may arise. . Here emerges an obvious logical difficulty involved in all work on the production of inherited change through environmental action. If we succeed in producing such 14 THE AMERICAN NATURALIST [Vor. LVI change, it is clear that the character altered was not a permanent one. And if after long re-subjection to the original environment the induced change disappears, it is equally clear that the new character was no more per- manent than the original one. If we now assume that there are other characters that are permanent, not alter- able by environmental action, of course we can obtain no. light on these by changing those characters that can be changed. ‘To me it appears that we have no right to as- sume, at the present stage in the game, that any such absolutely permanent characters exist. If this be true, then the production of changes persisting through many generations of uniparental, and even of biparental, re- production, with the further fact that the greater the number of generations the altering agent has acted, the greater the number of generations the change persists, seems of the greatest interest. It perhaps would, if action of the environmental agent continued sufficiently long, lead to production of inherited characteristics that are as permanent as any such characters are. It is cer- tainly, as Jollos agrees, capable of producing such di- versity of biotypes as we find within a species; and it might perhaps, if the results of diverse agents are cumu- lative, produce any of the inherited diversities found in organisms. This is the most promising lead that we have found in the study of uniparental production. In sum, the study of variation in uniparental repro- duction yields the following: The germinal or genotypic constitution in most organisms is extremely stable; in many stocks it changes not at all, so far as observation goes. To alter it by physical or chemical agents is usu- ally to kill it. In some of the lowest organisms—rhizo- pods, bacteria, some infusoria—it changes with some- what greater frequency, though still rarely. The nature of the changes, and whether they may be permanent, or must after many generations revert to the original condi- 1 The important observations and discussions of J ollos relating to changes producible by environmental action at the time of conjugation do not fall within the compass of a discussion of uniparental reproduction. No. 642] VARIATION IN REPRODUCTION 15 tion, is in some dispute. In these same organisms, en- vironmental agents may produce changes persisting through many generations of uniparental reproduction and even through biparental reproduction, the period of persistence depending partly, on the number of genera- tions through which the producing agent acted. This suggests that inherited characters as permanent as any that exist might in time be so produced. In spite of im- portant differences of opinion among investigators, to the reviewer the facts in uniparental reproduction seem to point more toward the production of evolutionary change by the action of the environment on the germ plasm than by any of the other methods. In this respect it takes its place in that modern revival of work on the inheritance of acquired characters, of which we had so striking an example this morning, in the account of the dizzy rats and of the inheritance of their dizziness; though in the study of uniparental reproduction nothing has appeared that indicates a transfer of somatic characters to the germ. REFERENCES (It has not seemed necessary to repeat here titles that are to be found generally in the literature of this subject; e.g., in the writer's book, ‘‘ Life, Death, Heredity and Evolution in Unieclulát- Organisms.’’ The following are not found in that work.) Banta, A. M, 1921. Selection in Cladocera on the Basis of a Physiological Character. Carnegie Institution Publ. No. 305. 170 pp Brierly, W. B. 1919. Some Concepts in Myeology—4An Attempt at a Synthesis, Trans. British Mycol, Soc., 6, 204-235. Brierly, W. B. 2 On a Form of Botrytis cinerea with Colourless Sclerotia. Phil. Trans. Roy. Soc., Ser. B, 210, 88-114. Erdmann, R. 1920. Endomixis and Size Variations in Pure Bred Lines of Parame- cium aurelia. Arch. f. Entw.-mech., 46, 85-148. Jollos, V. 1921. Experimentelle Protistenstudien. Arch. f. Protistenkunde, 43, Lóhnis, F, 1921. gsm upon the Life Cyeles of the Bacteria. Part I. Mem. Acad. Sci., 16, Second Memoir. 335 pp. VARIATIONS IN DATURA DUE TO CHANGES IN CHROMOSOME NUMBER - DR. ALBERT FRANCIS BLAKESLEE STATION FOR EXPERIMENTAL EvoLuTION, COLD SPRING Hanson, L: I, N- Y. Two forms with which we have recently carried on breeding experiments, the garden flower Portulaca and the jimson weed (Datura Stramonium), are strikingly different in the types of variations which they show. The Portulaca is procurable in a wide range of color va- rieties, and is apparently subject to relatively frequent mutations, both seminal and somatic, with sectorial and periclinal chimeras a common phenomenon. Sufficient breeding tests have been made to indicate that the varie- ties of Portulaca are due in large measure at least to gene mutations. In comparison with Portulaca, the jimson weed is relatively stable so far as gene mutations are concerned. Despite the large amount of breeding work with this species, both before and since the rediscovery of Mendel’s law, only the two allelomorphie pairs of char- acters, purple vs. white flowers, and spiny vs. smooth cap- sules, have been identified aside from the pair, tall vs. short stature recently determined by the writer and Avery (3 It is true that certain of our pure lines of Datura differ slightly from others when grown in comparable pedi- grees, but the fact remains that so far as sharply con- trasting Mendelian characters are concerned, the jimson weed is highly stable, while the Portulaca is highly mu- table. Our knowledge of changes in chromosome number in other forms is not sufficient to indicate if there is any significance for the present discussion in the difference just mentioned between Portulaca and Datura. Our interest in Datura began about 1910 or 1911, when the jimsons were used as demonstration Materi for students in genetics. In 1915 we found our first mutant which we called the Globe from the shape of its capsules. í 16 No. 642] VARIATION IN DATURA 17 The capsules of normal plants are ovate and the edges of the leaves somewhat toothed. Globe plants, on the con- trary, have depressed capsules and broader leaves with a more entire margin (cf. 3, figs. 7 and 9). Figure 1 shows sro | | /6 SEIS ' - 16780 Ch 1. Young plants in 3-inch pots. The normal 2n za ra in the middle, vin pis +1) Globe on the right, ut. e (2n + 2) Globe on the 1 young plants beginning to flower. In the center is a normal and on the right a Globe. The leaves of the latter are broader and more closely massed together. In the plant on the left, the Globe characters are more strongly developed. This plant represents an extreme type of the Globe mutant, and has been called the Round-leaf Globe. It is of considerable genetic interest and will be discussed later. It was at first thought that the Globe might be a tetraploid type like the Gigas GZnothera but a preliminary cytological investigation showed that such was not the ease. A peeuliarity in the inheritanee of the Globe (1, table 3) was found to be that the Globe complex is transmitted to only about one fourth of its offspring when a Globe parent is selfed; that about the same proportion of one fourth Globes only appears in the offspring when the Globe parent is erossed with pollen from a normal plant; and that the mutant character is transmitted to only a slight 18 THE AMERICAN NATURALIST [Vor. LVI extent or not at all through the pollen—to less than 2 per cent. in a large series of crosses. The next mutant found was Cocklebur (3, fig. 11) named from the resemblanee of its fruits to those of the cocklebur weed. The plant is weak and lopping and the leaves narrow and twisted. The Poinsettia mutant (3, fig. 14) was named from a fancied resemblance of its long clustered leaves to the hothouse plant of that name. The Poinsettia is of espe- cial interest, since this mutant was found to give curious ratios when heterozygous for color factors. As our eyes became better trained, other mutants were added to the list, largely through the keen discrimination of Mr. Avery and Mr. Farnham, until we now have 12 main mutants with some varieties, all of which transmit their mutant characters essentially in the same way in which the Globe complex was found to be transmitted. In addition we had a mutant which, unlike the 12 types just mentioned, was found to breed true, and since it is practically impossible to obtain crosses between it and the normal form from which it arose, it was called ‘‘ New Species "' (3, fig. 15). The capsules are somewhat spheri- cal and the leaves broad, although in a race of the same type later discovered the leaves are not greatly different from the normals. Heterozygous plants of the ** N. S." sometimes gave curious ratios in their offspring. Such was the situation up to the spring of 1920, when we were fortunate in securing the cooperation of Mr. Belling in a study of the nuclear condition of our mutants. On the basis of his work we are able to make the classi- fieation of types shown in Fig. 2. In the individual fig- ures—which of course are highly diagrammatie—the chromosomal constitution of somatie cells is represented. We have not attempted to represent the size differences determined by Mr. Belling and pietured in our paper in the morning session. A word of explanation of terms is desirable. The terms diploid, triploid and tetraploid are already current to indicate a balanced condition in which each chromosomal] set (we can not say chromosomal 1To be published shortly in the AMERICAN NATURALIST. No. 642] VARIATION IN DATURA 19 Balanced Types Unbalanced Types Diploid’ Moditied Diploids 2 m Sy NS S WE ANA ANZ DASA TRA TRAN [| AA SANTO [SANZISAVZISA SZ es Simple Tris Simple TetraSomic Double Trisomic Triploid Modified Triploids 4 M S AN uL VANI SY (3n) Tetraploid Modified Tetraploids eO |Z SiG. WMS [AS Au NOE RO Simple Pentasomic a Hexasomic jene) 4ntr2) Fic. 2, Diagrams illustrating the chromosomal types already found in Datura. pairs when there are more than 2 in a set) has respec- tively 2, 3, or 4 chromosomes. I have suggested (2) the terms disome, to indicate a set of 2 chromosomes, trisome a set of 3, and tetrasome a set of 4, ete., with the adjectives disomie, trisomie, tetrasomie, ete. Such terms may be found useful, but it seems impossible to devise a simple terminology that will adequately describe even the chro- mosomal irregularities at present known in Drosophila and Datura. Accordingly, after considerable discussion with Dr. Bridges, we have agreed upon a set of formule which is illustrated in the diagram and which we shall use in our present papers. A 20 THE AMERICAN NATURALIST [Vor. LVI Of the balanced forms there are even-balanced or stable, and odd-balanced or unstable types. In the even- balanced diploid, which is the normal jimson weed, the two chromosomes in each set go to opposite poles by the ordinary process of disomic reduction, and the plants breed true for chromosome number. Partly for the same reason, the even-balanced tetraploid, which is our *' New Species," breeds essentially true. The triploid, on the other hand, is odd-balaneed and therefore unstable, since in the trisomie disjunetion in each set two of the three chromosomes go to the one pole and one to the other, the process taking place at random. Through the operation of chance, therefore, gametes of different chromosomal number will be formed, and simple and double mutants as well as diploids will occur in the offspring. The rela- tion may be seen from the pollen of the three balanced types under the same magnification (Fig. 3), where the photograph at the left (a) shows a field of pollen from a diploid; that at the right, (c) with larger grains, pollen from a tetraploid; while that above (b) shows pollen from a triploid. Pollen from a triploid is not only character- ized by a large proportion of empty grains, but also by a great diversity in the size of the grains brought about by the differences in the number of ehromosomes which they eontain. ; The upper left-hand figure of the unbalanced types (Fig. 2) has one extra chromosome in the lower right-hand set, indieated by the arrow, giving 1 trisome, and 11 disomes in this nucleus, and its formula may be written (2n + 1). Such a simple mutant is the Globe—simple because only one set is affected. If another set has the extra chromo- some—say the set on the right—instead of the one with the arrow, this extra chromosome would cause the plant to assume the characters of, say, the Cocklebur mutant. It is obvious that since there are 12 sets in Datura and each set may have an extra chromosome, there are 12 mu- tants with the formula (2» +1) theoretically possibie. Through the process of disjunetion in these 12 mutants, half of the gametes should contain the extra chromosome, No. 649] VARIATION IN DATURA 21 and half should not. Differential mortality, affecting adversely zygotes with the extra chromosome, prevents the expected equality of (2n) and (2n + 1) individuals in the offspring from test erosses with diploids. Fic. 3. Photomicrographs of pollen grains: (a) from a diploid Datura; (b) from a triploid; (c) from a tetraploid. The magnification is indicated by the seale, each division of which equals 0.10 mm. The 12 mutants under discussion may best be repre- sented in a single figure by their capsules. In Figure 4 we have capsules of the 12 simple trisomic mutants viewed from the ovate side, each one of which represents the addition of a single extra chromosome presumably in a different set. There is the Globe with depressed capsules and stocky spines; the large long-spined Poinsettia; the narrow short-spined Cocklebur; the slender-spined Ilex; 22 THE AMERICAN NATURALIST [Vor. LVI Pointe fa Globe | Md lated : : Super doef a Glossy Reduced I Baek ling Fic. 4. Photographs of capsules of 12 mutants of Datura viewed from the ovate side, the Mutilated, usually mutilated with a dise: aged blotch; the short-spined Sugarloaf; the shiny capsule of Glossy, ete., with lastly the narrow, long-spined Wedge. I have No. 642] VARIATION IN DATURA 23 provisionally called these mutants the 12 apostles. Cer- tain of the 12 have varieties which may be called acolytes, and perhaps some of these in the figure may be reduced from the rank of apostles to that of acolytes when other forms are discovered. #66 Ba inset ts T fec Bodl %4 Uiiey Felled : Strawherey Fic. 5. Capsules representing 3 pairs of mutants. Those in the lower row are believed to represent varieties or ‘“‘ acolytes” of the ston rei types repre- sented abov In Figure 5, the mutants from which the capsules in the lower row were taken have been provisionally classed as acolytes of their respective apostles represented above. The evidence is best in regard to the mutants Wiry and Poinsettia which form the pair at the left in the figure. They both contain a single extra chromosome of approxi- mately the same size, and in both eases this extra chromo- some is shown, by peculiar color ratios in their offspring, to be in the set which carries the factors for purple and white flower color. The fact that though perfectly dis- tinet they are yet similar in appearance, and the fact that one has not infrequently given rise to the other in our 24 THE AMERICAN NATURALIST [Vor.LVI cultures, is a line of argument applicable not alone to the pair Wiry and Poinsettia. It also leads us to consider Rolled an acolyte of Sugarloaf, and Strawberry an aco- lyte of Buckling. The possibility of acolytes being caused by modifying Mendelian factors is being investigated. I have said that the Poinsettia mutant gave curious color ratios in its offspring (4 and 2, table 2). The evi- dence seems conclusive that Poinsettia has its extra chro- mosome in the set which carries the factors for purple and white flower color. A heterozygous Poinsettia may have one dose or two doses of the dominant purple flower color. The offspring of Poinsettia (like those of the Globe), it will be remembered, are part normals and part mutants. If a Poinsettia parent is duplex for purple, its normal off- spring show 8 purples to 1 white, while its Poinsettia offspring are all purples. If the Poinsettia parent is simplex for purple the ratio for the normal offspring is 5 purples to 4 whites, and for Poinsettia offspring is 7 purples to 2 whites. The back crosses are also distine- tive. By similar.reasoning we believe the Cocklebur mu- tant has its extra chromosome in the set which carries genes for presence or absence of spines on the capsules. _ The evidence is especially good for Poinsettia, since the color classes can be recognized in the seedpan. . Using a Poinsettia which arose in a purple line from Washington, D. C., we crossed it with a white line of similar appear- ance also from Washington, and, without going outside of these two lines, have synthesized Poinsettias of all the possible combinations of color factors and have made nearly all the possible combinations of crosses between them. The results with the Washington lines are in ac- cord with what would be expected from a random assort- ment of 3 chromosomes in the set containing the purple- white color factors. In a certain group of Poinsettias simplex for purple in which the 2 chromosomes bearing the white factor might have been brought in, so far as we knew, either from the white Washington stock, or from a distinct white line from Erfurt, Germany, the color ratios in the offspring of some parents were according to calcu- No. 642] VARIATION IN DATURA : 25 lation, but from other parents the whites were approxi- mately 6 times as frequent as would be expeeted. Later experiments seem to indieate: that we get the definite excess in white offspring from simplex parents when both the ** white ’’ chromosomes come from the German line; that we get the Poinsettia ratios typieal of random as- sortment when the two white chromosomes come from the Washington whites ; and that we get both of the two types of ratios from different individual F, parents when we make up an F, Poinsettia containing both a Washington white, and a German white chromosome. It is apparent that the peculiarity must be attributed to the German chromosomes. The question is receiving further experi- mental investigation but our provisional hypothesis to account for the difference in the ratios is that for some reason in trisomie disjunction the German white chromo- somes go to opposite poles rather than to the same pole 6 times as frequently as the laws of random assortment would dictate. Let us return to our diagrams in Fig. 2. Of the modi- fied diploids we may have 2 extra chromosomes in a single set forming a simple mutant of the formula (2n-4-2). An example is the round-leaf Globe (fig. 1) already mentioned. If two different sets are affeeted each with a single extra ehromosome we have a double mutant with the formula (2n+1-+1). Of the 66 different double tri- — somie mutants theoretically possible, we have a consider- able number now under cultivation. As an example, the double mutant Globe-Reduced is shown in Fig. 6. At the top is a eapsule of a normal diploid with its chromosomal diagram. At the left is a capsule of the Globe, and at the right a eapsule of Redueed. "Their diagrams indieate that the two mutants have different sets affected. "The plant represented by the capsules below, from the appearance of its leaves as well as from that of its fruit, is un- doubtedly a double mutant with the two sets affected as indieated in the diagram below. If the Globe-Reduced behaves like other double mutants we have bred, its off- E] 26 THE AMERICAN NATURALIST [Vor. LVI spring should contain normal diploids, both the Globe and the Reduced mutants, as well as the Sanne ‘toe Globe- Reduced, roughly in the proportion of 6: 2: 2:1 Peipioida (fig. 2) have been discussed in Tes morning's session. Our prediction at last year's meeting has been | TEN, lode Globe- Reduced ( dover i) 4 ł 1> A li WG! LT i Trisomie T nmm hashes eet ha nettes Sirs art Capsule of € diploid (25) Move: capsule of Reduced (2n +1 Rd) aa ^ : bim of Glo (2n -1 Gl) at left; and capsule of double mutant e-Reduced (2n +1 x Rd) below. Below peor capsule is given its ch mosomal diagram, T poc No. 642] VARIATION IN DATURA 2: fulfilled and we have obtained, in the offspring of a tri- ploid, practically the full range of (2n +1) mutants as well as double mutants of the formula (25 + 12-1). No modified triploids have as yet been identified, but even if we found them we could not expect to be able to propa- gate them by seed. Heterozygous tetraploid plants. also show curious ratios, according to whether there are 1, 2, or 3 doses of the dominant factor. Duplex plants give a 35: 1 ratio when selfed and the different types in the offspring segre- gate in a characteristic fashion. In the tetraploids we may have a single extra chromo- some in one set making a simple (4n +1) mutant, or 2 chromosomes in a set making a simple (4n-+ 2) mutant. We have two cases of a tetraploid with a deficiency in one set, producing a (4n — 1) mutant. Up to the present time, except for Gregory’s work on | tetraploid Primulas (5) which was correctly interpreted by Muller (6), Mendelian research has dealt almost ex- elusively with disomic inheritance. Our work with the jimsons and the recent investigations of Bridges on tri- ploid Drosophilas offer an opportunity for the rather novel study of trisomie, tetrasomie and pentasomie in- heritanee. We do not believe, however, that the jimson weed is peeuliar among plants in giving rise to chromo- somal mutants. The unbalancing effect of the extra chromosomes can best be illustrated by extra chromosomes in the Globe set. The (2n-+2) Globe has two extra chromosomes in the Globe set and hence should show a greater divergence from normal than the Globe with only one extra chromo- some. Such is the case. The simple (2n +1) Globe (like other mutants of this type) is less vigorous in growth than normals. The (2n +2) Globe is still less vigorous than the more common (2» 4- 1) Globe. From fig. 1 it will be seen, further, that the Globe characters i in the (2n + 2) Globe on the left, such as broadness of leaves, fatness of bud, and density of foliage, are much further developed 28 THE AMERICAN NATURALIST [Vor. LVI than in the (2n +1) Globe at the right, which has only one extra chromosome. Photographs of eapsules (Fig. 7) will further illustrate the idea of unbalance. Unfortunately the (2n + 2) Globe just mentioned fruits poorly and none of its capsules were available when the fruits of the other types were photographed. Later a photograph of a capsule was made to the same scale, and inserted in the proper place in the series. It will be evident that the Globe char- acters of relative stockiness of spines and depression of capsules are more marked in the (2n-+ 2) Globe where there are 2 extra chromosomes in the Globe set than in the (2n +1) Globe on the left where there is only one extra chromosome in this particular set. Likewise in the modified tetraploids the (plus 2) Globe on the right is more Globe-like than the (plus 1) Globe beside it. The degree of unbalanee of chromosomes in the nuclei may be given a quantitative expression. Thus in the (2n +1) Globe, the extra chromosome produces an excess of one over the balanced 2n condition. The nucleus is overbalanced by the active factors in a single Globe chro- mosome. This unbalance may be said to be 1 over 2n. In a similar way the (2n + 2) Globe with 2 extra chromo- somes has an unbalance of 2 over 2n. Having in mind these quantitative differences one would expect the (4n +1) Globe with an unbalance of 1 over 4n to show a less marked expression of the Globe charaeters than the (2n +1) Globe with an unbalance of 1 over 2n, They are, in faet, less readily recognized in recording our pedi- grees. The relation of unbalance enabled us to prediet the possibility of finding (4n +2) Globes with an unbal- ance of 2 over 4» which one would expect to be as distinct in appearance as (2n +1) Globes with an equivalent un- balance of 1 over 2». The predietion has been fulfilled and we are led to expect the appearance of Globes with 3, and Globes with 4 extra chromosomes in the Globe set, if tetraploid plants ean endure the extreme unbalanee of 4 over 4», the equivalent of 2 over 2n obtained in the (2n + 2) Globe. : No. 642] VARIATION IN DATURA 29 It must be emphasized that our quantitative expres- sions of unbalance hold strictly only for the chromosomal numbers in reference to a single set, and not necessarily for the somatic characters conditioned by them, although the nuclear unbalance seems to be reflected in the somatie 2 4ALS 4 MW aes 2) <= PA FA A n vH SUG an 4n 1 | sb Sour (nie ai) (4n*1&() (4nt26) Fia. Above, capsules with diagrams of a diploid (2n) and a tetraploid (4n). Blov, capsules with diagrams of the different Globe mutants appearance, at least in the Globe series just discussed. In double mutants, moreover, somatic effects may be in- tensified or largely neutralized by individual genes in the two extra ehromosomes, and an easy expression of the combined unbalance which they exert will therefore be impossible. The structural characters have been taken for illustra- 30 THE AMERICAN NATURALIST [Vor.LVI tion from a particular part of a single mutant, the Globe. A more detailed study of changes in external and internal morphology brought about by the presence of specific extra chromosomes in the several mutants is being under- taken in eooperation with Dr. Sinnott. : The unbalancing effect of an extra chromosome is shown in the lessened vigor of mutant plants. Thus from Globe parents as an example of (2n 4- 1) mutants, ordi- narily only one quarter of the offspring to reach record- able size are Globes, instead of the 50 per cent. expected. Moreover, when the plants are crowded the proportion of Globes surviving is considerably lessened. We have been discussing the unbalance as affecting the sporophytic generation. In the gametophyte, the un- balance is doubled. Thus from (2x+1) Globe plants with an unbalance of 1 over 2» the pollen grains with the extra chromosome have an unbalance of 1 over ». This extreme unbalanee hinders their funetioning and brings it about that the Globe character is transmitted to only a slight extent through the pollen (under 2 per cent. in a considerable series of crosses). It is of interest in this connection to note the results of selfing and erossing Globes of the tetraploid series. The unbalance in a (4n +1) Globe is 1 over 4n, while the unbalance in its pol- len grains which carry the extra chromosome is 1 over 2n. Due to this lessened unbalance in comparison with pollen of (2n +1) Globes, the pollen of the (4» +1) Globe trans- mits the Globe character to a higher percentage of its progeny (14 per cent. in the single pedigree tested), and partially for the same reason we have obtained higher proportions of Globes in the offspring from selfing such (4n +1) Globes (a total of about 60 per cent. in a single experiment). A more specific study of the effect of ex- tra chromosomes upon the gametophyte is being under- taken in cooperation with Dr. Buchholz. It will not be advisable at the present stage of our in- vestigations to discuss the possible external and internal factors which may induce the chromosomal aberrations which form the basis of our common mutations in Datura. No. 642] VARIATION IN DATURA 3l A study of the effects of radium rays undertaken in co- ` operation with Dr. Gager has given results which, al- though in an early stage of the experiment, appear sug- gestive in this connection. Other stimuli are being tested which appear to induce irregularities in the distribution of chromosomes to the pollen grains. It will be a matter of theoretical interest to be able to control experimentally the production of chromosomal mutations. It might also prove to be of considerable economie importance to be able to produce at will the full range of chromosomal mu- tants in any plants, especially in those which are propa- gated by vegetative means. To us, one of the most interesting features of the Da- tura work is the possibility afforded of analyzing the in- fluence of individual chromosomes upon both the mor- phology and physiology of the plant without waiting for gene mutations. Evidence is at hand which indicates that every chromosome in Datura carries factors which influence the expression of the so-called unit character purple pigmentation. Our work so far we believe adds evidence to the conclusion that the mature organism— plant or animal—is not a structure like a child’s house of blocks, made up of separate unit characters, nor is it de- termined by separate and unrelated unit factors. It is rather the resultant of a whole series of interacting and more or less conflicting forces contained in the individual chromosomes. LITERATURE CITED 1. Blakeslee, A. F. 1921. ata Globe Mutant in the Jimson Weed. Genetics, 6: 241—264, fi . Blakeslee, A. F. 1921. ea of Mutations and Their Possible Signifi- cance in Evolution. AMER, NAT., 55: 254-267 . Blakeslee, A. F., and B. T. Avery, Jr. 1919. Mutations in the Jimson Weed. ‘Jour, Heredity. 10: 111-120, fig. 5-15. . Blakeslee, A, F., John Belling and M. E. Farnham. 1920. Chromo- vind Duplication i Mendelian Phenomena in Datura Mutants. Science, N. S., 8-390. : Gregory, R. P, SA On the pope s Tetraploid Plants in Primula ` Proc. Roya: Society, B 87: 484—492. ; Matas, = J. 1914. A New Mode x: ‘Segregation in Gregory’s Tetra- ploid Primulas. AMER. NAT., 48: 508-512 ix) we) T e o VARIATION DUE TO CHANGE IN THE INDIVIDUAL GENE: DR. H. J. MULLER DrPARTMENT OF ZOOLOGY, UNIVERSITY OF TEXAS I. Tue RELATION BETWEEN THE GENES AND THE CHAR- ACTERS OF THE ÜRGANISM Tue present paper will be concerned rather with prob- lems, and the possible means of attacking them, than with the details of cases and data. The opening up of these new problems is due to the fundamental contribution which genetics has made to cell physiology within the last decade. This contribution, which has so far scarcely been assimilated by the general physiologists themselves, consists in the demonstration that, besides the ordinary proteins, carbohydrates, lipoids, and extractives, of their several types, there are present within the cell thousands of distinct substances—the ‘‘ genes "; these genes exist as ultramicroscopic particles; their influences neverthe- less permeate the entire cell, and they play a fundamental role in determining the nature of all cell substances, cell structures, and cell activities. Through these cell effects, in turn, the genes affect the entire organism. It is not mere guesswork to say that the genes are ultra-mieroscopie bodies. For the work on Drosophila has not only proved that the genes are in the chromo- somes, in definite positions, but it has shown that there must be hundreds of such genes within each of the larger chromosomes, although the length of these chromosomes is not over a few mierons. If, then, we divide the size of the chromosome by the minimum number of its genes, we find that the latter are partieles too small to give a visible image. The chemieal doni Moe of the genes, and the for- mule of their reactions, remain as yet quite unknown. We do know, for example, that in certain cases a given 1 Contribution No. 156, 32 No. 642] VARIATION IN INDIVIDUAL GENE 33 pair of genes will determine the existence of a particular enzyme (concerned in pigment production), that another pair of genes will determine whether or not a certain agglutinin shall exist in the blood, a third pair will deter- mine whether homogentisie acid is secreted into the urine (* alkaptonuria "), and so forth. But it would be absurd, in the third case, to conclude that on this account the gene itself consists of homogentisie acid, or any related substance, and it would be similarly absurd, there- fore, to regard cases of the former kind as giving any evidence that the gene is an enzyme, or an agglutinin-like body. The reactions whereby the genes produce their ultimate effects are too complex for such inferences. Each of these effects, which we call a ‘‘ character "' of the organism, is the product of a highly complex, intri- cate, and delicately balanced system of reactions, caused by the interaction of countless genes, and every organic structure and activity is therefore liable to become in- creased, diminished, abolished, or altered in some other way, when the balance of the reaction system is disturbed by an alteration in the nature or the relative quantities of any of the component genes of the system. To return now to these genes themselves. II. Tue PROBLEM or GENE MUTABILITY The most distinctive characteristic of each of these ultra-mieroseopie particles—that characteristic whereby we identify it as a gene—is its property of self-propaga- tion: the fact that, within the complicated environment of the cell protoplasm, it reacts in such a way as to convert some of the common surrounding material into an end-product identical in kind with the original gene itself. This action fulfills the chemist’s definition of ** autocatalysis ’’; it is what the physiologist would call ** growth °’; and when it passes through more than one generation it becomes ‘‘ heredity.” It may be observed that this reaction is in each instance a rather highly localized one, since the new material is laid down by the side of the original gene. 34 THE AMERICAN NATURALIST [Vou. LVI The fact that the genes have this autocatalytic power is in itself sufficiently striking, for they are undoubtedly complex substances, and it is difficult to understand by -what strange coincidence of chemistry a gene can happen to have just that very special series of physico-chemical effects upon its surroundings which produces—of all pos- sible end-products—just this particular one, which is identical with its own complex structure. But the most remarkable feature of the situation is not this oft-noted autocatalytic action in itself—it is the fact that, when the structure of the gene becomes changed, through some ** chance variation,’’ the catalytic property of the gene may * become correspondingly changed, in such a way as to leave it still autocatalytie. In other words, the change in gene structure—accidental though it was—has some- how resulted in a change of exactly appropriate nature in the catalytic reactions, so that the new reactions are now accurately adapted to produce more material just like that in the new changed gene itself. It is this para- doxical phenomenon which is implied in the expression ‘* variation due to change in the individual gene,” or, as it is often called, ** mutation." What sort of strueture must the gene possess to permit it to mutate in this way? Since, through change after change in the gene, this same phenomenon persists, it is evident that it must depend upon some general feature of gene construction—common to all genes— which gives each one a general autocatalytic power—a “carte blanche ’’—to build material of whatever speeifie sort it itself happens to be composed of. This general prineiple of gene strueture might, on the one hand, mean nothing more than the possession by each gene of some very simple character, such as a particular radicle or ** side- chain ’’—alike in them all—which enables each gene to enter into combination with certain highly organized materials in the outer protoplasm, in such a way as to result in the formation, ** by ” the protoplasm, of more material like this gene which is in combination with it. In 2 It is of course conceivable, and even unavoidable, that some types of changes do d j i i pianot : a estroy the gene’s autoeatalytie power, and thus result in its No. 642] VARIATION IN INDIVIDUAL GENE 35 that case the gene itself would only initiate and guide the direction of the reaction. On the other hand, the extreme alternative to such a conception has been generally as- sumed, perhaps gratuitously, in nearly all previous theories concerning hereditary units; this postulates that the chief feature of the autocatalytie mechanism resides in the structure of the genes themselves, and that the outer protoplasm does little more than provide the build- ing material. In either case, the question as to what the general principle of gene construction is, that permits this phenomenon of mutable autocatalysis, is the. most fundamental question of genetics. The subject of gene variation is an important one, however, not only on account of the apparent problem that is thus inherent in it, but also because this same peeuliar phenomenon that it involves lies at the root of organic evolution, and hence of all the vital phenomena which have resulted from evolution. It is commonly sald that evolution rests upon two foundations—inher- itanee and variation; but there is a subtle and important error here. Inheritance by itself leads to no change, and variation leads to no permanent change, unless the varia- tions themselves are heritable. Thus it is not inheritance and variation which bring about evolution, but the in- heritance of variation, and this in turn is due to the general principle of gene construction which causes the persistence of autocatalysis despite the alteration in structure of the gene itself. Given, now, any material or colleetion of materials having this one unusual char- acteristic, and evolution would automatically follow, for this material would, after a time, through the accumula- tion, competition and seleetive spreading of the self- propagated variations, come to differ from ordinary in- organic matter in innumerable respects, in addition to the original difference in its mode of catalysis. There would thus result a wide gap between this matter and other matter, which would keep growing wider, with the Increasing complexity, diversity and so-called ‘‘ adapta- tion ” of the selected mutable material. 36 : THE AMERICAN NATURALIST [Vor.LV1 III. A POSSIBLE ATTACK THROUGH CHROMOSOME BEHAVIOR In thus recognizing the nature and the importance of the problem involved in gene mutability have we now entered into a cul de sac, or is there some way of pro- ceeding further so as to get at the physical basis of this peculiar property of the gene? The problems of growth, variation and related processes seemed difficult enough to attack even when we thought of them as inherent in the organism as a whole or the cell as a whole—how now can we get at them when they have been driven back, to some extent at least, within the limits of an invisible particle? A gene can not effectively be ground in a mortar, or distilled in a retort, and although the physico- chemical investigation of other biological substances may conceivably help us, by analogy, to understand its struc- ture, there seems at present no method of approach along this line. ; There is, however, another possible method of approach available: that is, to study the behavior of the chromo- somes, as influenced by their contained genes, in their various physical reactions of segregation, crossing. over, division, synapsis, ete. This may at first sight seem very remote from the problem of getting at the structural principle that allows mutability in the gene, but I am in- clined to think that such studies of synaptic attraction be- tween chromosomes may be especially enlightening in this connection, because the most remarkable thing we know about genes—besides their mutable autocatalytic power— is the highly specific attraction which like genes (or local products formed by them) show for each other. As in the case of the autocatalytic forces, so here the attractive forces of the gene are somehow exactly adjusted so as to react in relation to more material of the same com- plicated kind. Moreover, when the gene mutates, the forces become readjusted, so that they may now attract material of the new kind; this shows that the attractive or synaptic property of the gene, as well as its eatalytie property, is not primarily dependent on its specific struc- ture, but on some general principle of its make-up, that No. 642] VARIATION IN INDIVIDUAL GENE 31 causes whatever specific structure it has to be auto-attrac- tive (and autocatalytic). This auto-attraction is evidently a strong force, exert- ing an appreciable effect against the non-specific mutual repulsions of the chromosomes, over measurable micro- scopic distances much larger than in the case of the ordi- nary forces of so-called cohesion, adhesion and adsorp- tion known to physical science. In this sense, then, the physicist has no parallel for this force. There seems, however, to be no way of escaping the conclusion that in the last analysis it must be of the same nature as these other forces which cause inorganic substances to have specific attractions for each other, according to their chemical composition. These inorganic forces, according to the newer physics, depend upon the arrangement and mode of motion of the electrons constituting the molecules, - which set up electro-magnetic fields of force of specific patterns. To find the principle peculiar to the construc- tion of the force-field pattern of genes would accordingly be requisite for solving the problem of their tremendous auto-attraction. Now, according to Troland (1917), the growth of erys- tals from a solution is due to an attraction between the solid erystal and the molecules in solution caused by the similarity of their force. field patterns, somewhat as similarly shaped magnets might attract each other— north to south poles—and Troland maintains that essen- tially the same mechanism must operate in the auto- catalysis of the hereditary particles. If he is right, each different portion of the gene structure must—like a erystal—attract to itself from the protoplasm materials of a similar kind, thus moulding next to the original gene another structure with similar parts, identically arranged, which then become bound together to form another gene, a replica of the first. This does not solve the question of what the general principle of gene construction is, which permits it to retain, like a crystal, these properties of auto-attraetion,? but if the main point is correct, that 3It ean hardly be true, as Troland intimates, that all similar fields at- traet each other more than they do dissimilar fields, otherwise all substances would be autoeatalytie, and, in fact, no substances would be soluble. More- 38 THE AMERICAN NATURALIST [Vor. LVI the autocatalysis is an expression of specific attractions between portions of the gene and similar protoplasmic building blocks (dependent on their force-field patterns), it is evident that the very same forces which cause the genes to grow should also cause like genes to attract each other, but much more strongly, since here all the indi- vidual attractive forces of the different parts of the gene are summated. If the two phenomena are thus really dependent on a common principle in the make-up of the gene, progress made in the study of one of them should help in the solution of the other. Great opportunities are now open for the study of the nature of the synaptic attraction, especially through the discovery of various races having abnormal numbers of _ chromosomes. ‘Here we have already the finding by Belling, that where three like chromosomes are present, the close union of any two tends to exclude their close union with the third. This is very suggestive, because the same thing is found in the cases of specific attractions between inorganic particles, that are due to their force field patterns. And through Bridges’ finding of triploid Drosophila, the attraction phenomena can now be brought down to a definitely genic basis, by the introduction of specific genes—especially those known to influence chro- mosome behavior—into one of the chromosomes of a triad. The amount of influence of this gene on attraction may then be tested quantitatively, by genetic determina- tion of the frequencies of the various possible types of Segregation. By extending such studies to include. the effect of various conditions of the environment— such as temperature, electrostatic stresses, ete.—in the pres- ence of the different genetic situations, a considerable field is opened up. This suggested connection between chromosome behav- ior and gene structure is as yet, however, only sibility. It must not be forgotten that at present over, a pos- Wwe can if the parts of a molecule are in any kind of ** solid," three dimen- sional formation, it would seem that those in the middle would scarcely have opportunity to exert the moulding effect above mentioned. It therefore appears that a ial manner of construction must be necessary, in order that a complicated structure like a gene may exert such an effect, No. 642] VARIATION IN INDIVIDUAL GENE 39 not be sure that the synaptic attraction is exerted by the genes themselves rather than by local products of them, and it is also problematieal whether the chief part of the mechanism of autocatalysis resides within the genes rather than in the ‘‘ protoplasm.’’ Meanwhile, the method is worth following up, simply because it is one of our few conceivable modes of approach to an all-im- portant problem. It may also be recalled in this connection that besides the genes in the chromosomes there is at least one sim- ilarly autoeatalytie material in the chloroplastids, which likewise may become permanently changed, or else lost, as has been shown by various studies on chlorophyll inher- itance. Whether this plastid substance is similar to the genes in the chromosomes we can not say, but of course it can not be seen to show synaptic attraction, and could not be studied by the method suggested above.* IV. THE Arrack THROUGH STUDIES OF MUTATION There is, however, another method of attack, in a sense more direct, and not open to the above criticisms. That is the method of investigating the individual gene, and the structure that permits it to change, through a study of the ehanges themselves that occur in it, as observed by the test of breeding and development. It was through the investigation of the changes in the chromosomes— eaused by erossing over—that the structure of the chro- mosomes was analyzed into their constituent genes in line formation ; it was through study of molecular changes that molecules were analyzed into atoms tied together in definite ways, and it has been finally the rather recent finding of changes in atoms and investigation of the resulting pieces, that has led us to the present analysis of atomie strueture into positive and negative electrons having eharaeteristie arrangements. Similarly, to under- . stand the properties and possibilities of the individual gene, we must study the mutations as directly as possible, and bring the results to bear upon our problem. 4It may be that there are still other elements in the cell which have the nature of genes, but as no eritieal evidenee has ever been addueed for their existence, it would be highly hazardous to postulate them. 40 THE AMERICAN NATURALIST [Vou LVI (a) The Quality and Quantity of the Change In spite of the fact that the drawing of inferences concerning the gene is very much hindered, in this method, on account of the remoteness of the gene-cause from its character-effect, one salient point stands out already. It is that the change is not always a mere loss of material, because clear-cut reverse mutations have been obtained in corn, Drosophila, Portulaca, and prob- ably elsewhere. If the original mutation was a loss, the reverse must be a gain. Secondly, the mutations in many cases seem not to be quantitative at all, since the different allelomorphs formed by mutations of one original gene . often fail to form a single linear series. One case, in fact, is known in which the allelomorphs even affect totally different characters : this is the case of the truncate series, in which I have found that different mutant genes at the same locus may cause either a shortening of the wing, an eruption on the thorax, a lethal effect, or any combina- tion of two or three of these characters. In such a ease we may be dealing either with changes of different types occurring in the same material or with changes (possibly quantitative changes, similar in type) occurring in dif- ferent component parts of one gene. Owing to the uni- versal applicability of the latter interpretation, even where allelomorphs do not form a linear series, it can not be categorically denied, in any individual ease, that the changes may be merely quantitative changes of some part of the gene. If all changes were thus quantitative, even in this limited sense of a loss or gain of part of the gene, our problem of why the changed gene still seems to be autocatalytic would in the main disappear, but such a situation is excluded a priori since in that case the thousands of genes now existing could never have evolved. Although a given gene may thus change in various ways, it is important to note that there is a strong tend- ency for any given gene to have its changes of a particular kind, and to mutate in one direction rather than in another. And although mutation certainly does not always consist of loss, it often gives effects that might be termed losses. In the case of the mutant genes for No. 642] VARIATION IN INDIVIDUAL GENE 41 bent and eyeless in the fourth chromosome of Drosophila it has even been proved, by Bridges, that the effects are of exactly the same kind, although of lesser intensity, than those produced by the entire loss of the chromosome in which they lie, for flies having bent or eyeless in one chromosome and lacking the homologous chromosome are even more bent, or more eyeless, than those having a homologous chromosome that also contains the gene in question. The fact that mutations are usually recessive might be taken as pointing in the same direction, since it has been found in several cases that the loss of genes— as evidenced by the absence of an entire chromosome of one pair—tends to be much more nearly recessive than dominant in its effect. The effect of mutations in causing a loss in the char- acters of the organism should, however, be sharply distin- guished from the question of whether the gene has undergone any loss. It is generally true that mutations are much more apt to cause an apparent loss in character than a gain, but the obvious explanation for that is, not because the gene tends to lose something, but because most characters require for proper development a nicely adjusted train of processes, and so any change in the genes—no matter whether loss, gain, substitution or rear- rangement—is more likely to throw the developmental mechanism out of gear, and give a ‘‘ weaker ” result, than to. intensify it. For this reason, too, the most fre- quent kind of mutation of all is the lethal, which leads to the loss of the entire organism, but we do not conclude from this that all the genes had been lost at the time of the mutation. The explanation for this tendency for most changes to be degenerative, and also for the fact that certain other kinds of changes—like that from red to pink eye in Drosophila—are more frequent than others— such as red to brown or green eye—lies rather in develop- mental mechanies than in geneties. It is because the developmental processes are more unstable in one diree- tion than another, and easier to push ‘‘ downhill ” than up, and so any mutations that oceur—no matter what the gene change is like—are more apt to have these effects 42 THE AMERICAN NATURALIST [Vor.LVI than the other effects. If now selection is removed in regard to any particular character, these character changes which occur more readily must accumulate, giv- ing apparent orthogenesis, disappearance of unused ` - organs, of unused physiological capabilities, and so forth. As we shall see later, however, the changes are not so frequent or numerous that they could ordinarily push evolution in such a direction against selection and against the immediate interests of the organism. In regard to the magnitude of the somatic effect pro- duced by the gene variation, the Drosophila results show that there the smaller character changes occur oftener than large ones. The reason for this is again probably to be found in developmental mechanics, owing to the fact that there are usually more genes slightly affecting a given character than those playing an essential róle in its formation. The evidence proves that there are still more genes whose change does not affect the given char- acter at all—no matter what this character may be, unless it is life itself—and this raises the question as to how many mutations are absolutely unnoticed, affecting no character, or no detectable character, to any appreciable extent at all. Certainly there must be many such muta- tions, judging by the frequency with which *' modifying factors ’’ arise, which produce an effect only in the presence of a special genetic complex not ordinarily present. ; (b) The Localization of the Change Certain evidence concerning the causation of mutations has also been obtained by studying the relations of their _ occurrence to one another. Hitherto it has nearly always been found that only one mutation has occurred at a time, restricted to a single gene in the cell. I must omit from consideration here the two interesting cases of deficiency, found by Bridges and by Mohr, in each of which it seems certain that an entire region of a chromosome, with its whole cargo of genes, changed or was lost, and also a certain peculiar case, not yet cleared up, which has re- cently been reported by Nilson-Ehle; these important No. 642] VARIATION IN INDIVIDUAL GENE 43 cases stand alone. Aside from them, there are only two instances in which two (or more) new mutant genes have been proved to have been present in the same gamete. Both of these are cases in Drosophila—reported by Muller and Altenburg (1921)—in which a gamete con- tained two new sex-linked lethals; two cases are not a greater number than was to have been expected from a random distribution of mutations, judging by the fre- quency with which single mutant lethals were found in the same experiments. Ordinarily, then, the event that causes the mutation is specific, affecting just one par- tieular kind of gene of all the thousands present in the cell. That this specificity is due to a spatial limitation rather than a chemical one is shown by the fact that when the single gene changes the other one, of identical com- position, located near by in the homologous chromosome of the same cell, remains unaffected. This has been proved by Emerson in corn, by Blakeslee in Portulaca, - and I have shown there is strong evidence for it in Dro- sophila. Hence these mutations are not caused by some general pervasive influence, but are due to ‘‘ accidents ”’ occurring on a molecular scale. When the molecular or atomic motions chance to take a particular form, to which the gene is vulnerable, then the mutation occurs. It will even be possible to determine whether the entire gene changes at once, or whether the gene consists of several molecules or particles, one of which may change at atime. This point can be settled in organisms having determinate cleavage, by studies of the distribution of the mutant character in somatically mosaic mutants. If there is a group of particles in the gene, then when one par- ticle changes it will be distributed irregularly among the descendant cells, owing to the random orientation of the two halves of the chromosome on the mitotic spindles of succeeding divisions, but if there is only one particle to 5 This depends on the assumption that if the gene does consist of several particles, the halves of the chromosomes, at each division, receive a random sample of these particles, That is almost a necessary assumption, since a gene formed of particles each one of which was separately partitioned at division would tend not to persist as such, for the occurrence of mutation in one partiele after the other would in time differentiate the gene into a num- ber of different genes consisting of one particle each. 44 THE AMERICAN NATURALIST — [Vor.LVI change, its mutation must affect all.of the cells in a bloc, that are descended from the mutant cell. (c) The Conditions under which the Change occurs But the method that appears to have most scope and promise is the experimental one of investigating the con- ditions under which mutations occur. This requires studies of mutation frequency under various methods of handling the organisms. As yet, extremely little has been done along this line. That is because, in the past, a muta- tion was considered a windfall, and the expression ** mu- tation frequency ’’ would have seemed a contradiction in terms. To attempt to study it would have seemed as absurd as to study the conditions affecting the distribu- tion of dollar bills on the sidewalk. You were simply fortunate if you found one. Not even controls, giving the ** normal ’’ rate of mutation—if indeed there is such a thing—were attempted. ^ Of late, however, we may say that certain very exceptional banking houses have been found, in front of which the dollars fall more frequently— in other words, specially mutable genes have been dis- eovered, that are beginning to yield abundant data at the hands of Nilsson-Ehle, Zeleny, Emerson, Anderson and others. For some of these mutable genes the rate of change is found to be so rapid that at the end of a few decades half of the genes descended from those originally present would have become changed. After these genes have once mutated, however, their previous mutability no longer holds. In addition to this ‘‘ banking house method "' there are also methods, employed by Altenburg and myself, for—as it were—automatieally sweeping up wide areas of the streets and sifting the collections for the valuables. By these special genetie methods of reaping mutations we have recently shown that the ordinary genes of Drosophila—unlike the mutable genes above— would usually require at least a thousand years—prob- € Studies of ‘‘ mutation frequency '' had of course been made in the CEnotheras, but as we now know that these were not studies of the rate of gene change but of the frequencies of erossing over and of chromosome aberrations they may be neglected for our present purposes No. 642] VARIATION IN INDIVIDUAL GENE 45` ably very much more—before half of them became changed. This puts their stability about on a par with, if not much higher than, that of atoms of radium—to use a fairly familiar analogy. Since, even in these latter ex- periments, many of the mutations probably occurred within a relatively few rather highly mutable genes, it is likely that most of the genes have a stability far higher than this result suggests. The above mutation rates are mere first gleanings—we have yet to find how different conditions affect the occur- . rence of mutations. There had so far been only the negative findings that mutation is not confined to one sex (Muller and Altenburg, 1919 ; Zeleny, 1921), or to any one stage in the life cycle (Bridges, 1919; Muller, 1920; Zeleny, 1921), Zeleny’s finding that bar-mutation is not influenced by recency of origin of the gene (1921), and the as yet inconclusive differences found by Altenburg and myself for mutation rate at different temperatures (1919), until at this year’s meeting of the botanists Emerson announced the definite discovery of the influence of a genetic factor in corn upon the mutation rate in its allelomorph, and Anderson the finding of an influence upon mutation in this same gene, caused by developmental conditions—the mutations from white to red of the mu- table gene studied occurring far more frequently in the cells of the more mature ear than in those of the younger ear. These two results at least tell us decisively that mutation is not a sacred, inviolable, unapproachable process: it may be altered. These are the first steps; the way now lies open broad for exploration. It is true that I have left out of account here the re- ported findings by several investigators, of genetic vari- ations caused by treatments with various toxic substances and with certain other unusual conditions. In most of these cases, however, the claim has not been made that ` actual gene changes have been caused: the results have usually not been analyzed genetically and were in fact not analyzable genetically; they could just as well be interpreted to be due to abnormalities in the distribution of genes—for instance, chromosome abnormalities like 46 THE AMERICAN NATURALIST . [Vor.LVI those which Mavor has recently produced with X-rays— as to be due to actual gene mutations. But even if they were due to real genie differences, the possibility has in most cases by no means been excluded (1) that these genic differences were present in the stock to begin with, and merely became sorted out unequally, through random segregation; or (2) that other, invisible genie differences were present which, after random sorting out, themselves caused differenees in mutation rate between the different lines. Certain recent results by Altenburg and myself suggest that genie differences, affecting mutation rate, may be not uncommon. To guard against either of these possibilities it would have been necessary to test the stocks out by a thorough course of inbreeding beforehand, or else to have run at least half a dozen different pairs of parallel lines of the control and treated series, and to have obtained a definite difference in the same direction between the two lines of each pair; otherwise it can be proved by the theory of ** probable error ’’ that the dif- ferences observed may have been a mere matter of ran- dom sampling among genic differences originally present. Accumulating large numbers of abnormal or inferior individuals by selective propagation of one or two of the treated lines—as has been done in some cases—adds noth- ing to the significance of the results. At best, however, these genetically unrefined methods would be quite insensitive to mutations occurring at any- thing like ordinary frequency, or to such differences in mutation rate as have already been found in the analytical experiments on mutation frequency. And it seems quite - possible that larger differences than these will not easily be hit upon, at least not in the early stages of our investi- gations, in view of the evidence that mutation is ordi- narily due to an accident on an ultramieroscopie scale, rather than directly caused by influences pervading the organism. For the present, then, it appears most prom- ising to employ organisms in whieh the genetie composi- tion ean be controlled and analyzed, and to use genetie methods that are sensitive enough to disclose mutations occurring in the control as well as in the treated individ- No. 642] VARIATION IN INDIVIDUAL GENE 47 uals. In this way relatively slight variations in muta- tion frequency, caused by the special treatments, can be determined, and from the conditions found to alter the mutation rate slightly we might finally work up to those which affect it most markedly. The only methods now meeting this requirement are those in which a ‘particular mutable gene is followed, and those in which many homozygous or else genetically controlled lines can be run in parallel, either by parthenogenesis, self-fertiliza- tion, balanced lethals or other special genetic means, and later analyzed, through sexual reproduction, segrega- tion and erossirig over. V. OTHER POSSIBILITIES We can not, however, set fixed limits to the possibilities of research. We should not wish to deny that some new and unusual method may at any time be found of directly producing mutations. For example, the phenomena now being worked out by Guyer may be a case in point. There is a curious analogy between the reactions of immunity and the phenomena of heredity, in apparently funda- mental respects,’ and any results that seem to connect the two are worth following to the limit. Finally, there is a phenomenon related to immunity, of still more striking nature, which must not be neglected by geneticists. Thisis the d’Hérelle phenomenon. D'Hérelle found in 1917 that the presence of dysentery bacilli in the body caused the production there of a filterable sub- stance, emitted in the stools, which had a lethal and in fact dissolving action on the corresponding type of bac- teria, if a drop of it were applied to a colony of the bac- teria that were under cultivation. So far, there would be nothing to distinguish this phenomenon from im- 7I refer here to the remarkable specificity with which a particular com- plex antigen calls forth processes that construct for it an antibody that is attracted to it and fits it ** like lock and key,’’ followed by further proc- esses that cause more and more of the antibody to be reproduced. If the antigen were a gene, which could be slightly altered by the cell to form the antibody that neutralized it—as some enzymes can be slightly changed by heating so that they counteract the previous active enzyme—and if this antibody-gene then became implanted in the cell so as to keep on growing, all the phenomena of immunity would be produced. 48 THE AMERICAN NATURALIST [Vor. LVI munity. But he further found that when à drop of the affected colony was applied to a second living colony, the ` second colony would be killed; a drop from the second would kill a third colony, and so on indefinitely. In other words, the substance, when applied to colonies of bacteria, became multiplied or increased, and could be so increased indefinitely ; it was self-propagable. It fulfills, then, the definition of an autocatalytic substance, and although it may really be of very different composition and work by a totally different mechanism from the genes in the chromosomes, it also fulfills our definition of a gene But the resemblance goes further—it has been found by Gratia that the substance may, through appro- priate treatments on other bacteria, become changed (so as to produce a somewhat different effect than before, and attack different bacteria) and still retain its self- ` propagable nature. That two distinct kinds of substances—the d'Hérelle substances and the genes—should both possess this most remarkable property of heritable variation or ‘‘ muta- bility," each working by a totally different mechanism, is quite conceivable, considering the complexity of proto- plasm, yet it would seem a curious coincidence indeed. It would open up the possibility of two totally different kinds of life, working by different mechanisms. [ On the other hand, if these d'Hérelle bodies were really genes, fundamentally like our ehromosome genes, they would give us an utterly new angle from which to attack-the gene problem. They are filterable, to some extent isol- able, ean be handled in test-tubes, and their properties, as shown by their effects on the baeteria, ean then be studied after treatment. It would be very rash to eall these bodies genes, and yet at present we must confess that there is no distinetion known between the genes and them. Hence we can not categorically deny that perhaps we may be able to grind genes in a mortar and cook them in a beaker after all. Must we geneticists become bae- 8 D'Hérelle himself thought that the substance was a filterable virus para- sitie on the bacterium, called forth by the host body. It has since been found that various bactcria each cause the production of D’Hérelle sub- stances which are to some extent specifie for the respective bacteria. No. 642] VARIATION IN INDIVIDUAL GENE 49 teriologists, physiological chemists and physicists, simul- taneously with being zoologists and botanists? Let us hope so. I have purposely tried to paint things in the rosiest possible colors. Actually, the work on the individual gene, and its mutation, is beset with tremendous difficulty. Such progress in it as has been made has been by minute steps and at the cost of infinite labor. Where results are thus meager, all thinking becomes almost equivalent to speculation. But we can not give up thinking on that account, and thereby give up the intellectual incentive to our work. In fact, a wide, unhampered treatment of all possibilities is, in such cases, all the more imperative, in order that we may direct these labors of ours where they have most chance to count. We must provide eyes for action. "The real trouble comes when speculation masquerades as empirical fact. For those who cry out most loudly against ‘‘ theories ’’ and ‘‘ hypotheses ’’—whether these latter be the chromosome theory, the factorial ‘‘ hypoth- esis," the theory of crossing over, or any other—are often the very ones most guilty of stating their results in terms that make illegitimate implicit assumptions, which they themselves are scarcely aware of simply because they are opposed to dragging ‘‘ speculation "' intothe open. Thus they may be finally led into the worst blunders of all. Let us, then, frankly admit the uncer- tainty of many of the possibilities we have dealt with, us- ing them as a spur to the real work. LITERATURE CITED Blakeslee, A. F, 1920. A Dwarf Mutation in Portulaca showing Vegetative Reversions. oura, Vol. 5, pp. 419-433. Bridges, C. B. 1917. Deficiency. Genetics, Vol. 2, pp. 445—465. Bridges, C, B. 1919. The Developmental Stages at which Mutations Oeeur in the Germ Tract. Proc. Soc. Exp, Biol. and Med., Vol. 17, pp. 1-2. Bridges, C, B. 1921. Genetical and Cytological Proof of Non-disjunetion-of the 50 THE AMERICAN NATURALIST [Vor. LVI Fourth Chromosome of Drosophila me'anogaster. Proc. Nat. Acad. Sci., Vol. 7, pp. 186-192. D'Hérelle, F. 1917. Compt. rend. Acad., Vol, 165, p. 373. 1920. Compt. rend. Soc. Biol., Vol, ise pp. 52, 97, 247. Emerson, R. 1911. The Inheritance of a Recurring Somatic Variation in Variegated Ears of Maize. Amer. NAT., Vol. 48, pp. 87-115. Gratia, A. 1921. Studies on the D'Hérelle Phenomenon. Jour, Ezp. Med., Vol. 34, pp. 115-126, à; Mavor, J. W, 1921. On the Elimination of the X-chromosome from the egg of Dro- sophila melanogaster by X-rays. Science, N. S., 54, pp. 277- 279, Mohr, O. L. 1919. Charaeter Changes eaused by Mutation of an Entire Region of a Chromosome in Drosophi'a. Genetics, Vol. 4, pp. 275-282. Muller, H. J. 1920 Mine Changes in the White-eye Series of Drosophi'a and their aring on the Manner of Occurrence 9t Mutation. Jour. Rep . Zool., Vol. 31, pp. 443-473. Muller, H. J., and E Altenburg. 1919. The Rate of Change of Hereditary Faetors in Drosophila. Proc. Soc. Exp. Bio’. and Med., Vol, 17, pp. 10-14 Muller, H, J. and E. Altenburg. 1921. A Study of the Character and Mode of Origin of 18 Mutations in the X-ehromosome of Drosophila. Anat, Rec., Vol. 20, p. 213. Nilsson-Ehle, H, 1911. Ueber Fälle spontanen Wegfallens eines Hemmungsfaktors beim a Zeit. f. Ind. Abst. u. Vererb., Vol. 5, pp. 1-37. Nilsson-Ehle, H. 1920. — Allelomorphe und oo beim Weizen. Hereditas, Vol. 1, pp. 277-312 Troland, L. T. 1917. Biological Enigmas and the Theory of Enzyme Action. AMER. Nart., Vol. 51, pp. 321-350, Wollstein, M. 1921. gone on the Phenomenon of D'Hérelle with Bacillus dysen- tarie. Jour. Exp. Med., Vol. 34, pp. 467-477. Zeleny, C. _ 1920, The Direction and Frequency of Mutation in a Series of Multiple. Allelomorphs. Anat, Rec., Vol. 20, p. 21 Zeleny, C. 1921. The Direction and Frequency of Mutation in the Bar-eye Series of Mu ultiple Allelomorphs of Drosophi'a. Jour. Exp. Zool., Vol, 34, pp. 203-233. THE ORIGIN OF VARIATIONS IN SEXUAL AND | SEX-LIMITED CHARACTERS DR. CALVIN B. BRIDGES COLUMBIA UNIVERSITY Ix dealing with sex and its determination, attention has been most sharply foeused upon forms with separate sexes and upon the visible differences between the chro- mosome groups of the two sexes. The result has been that the formulation of sex-determination has remained in terms of chromosomes, while the modern unit of deter- mination is the gene; and also the subject of sex has been rather separated off from the main body of heredity. My diseussion will be largely a process of resolving chromo- somes into component genes, and showing that the con- ception of the nature and action of genes as gained from the study of non-sexual characters is valid in interpreting sex phenomena. The facts of mutation and of linkage have given us the conception of a gene as a distinct chemical entity having a definite location in a particular chromosome. Each gene is essentially a factory, which is manufacturing a characteristic set of chemical products that are delivered to the common cytoplasm, and that produce development through interaction with each other and with materials from outside. But since the chemicals produced by the different genes are different, some genes will have muck effect upon one character and little effect upon another, so that a relatively small proportion of the genes will be actively concerned in producing any given character. Some of these genes tend to make the character more pro- nounced, and others tend to make it less pronounced, so that the grade of development actually realized by each particular character will be determined by the equilibrium - between its modifying genes. The forms into which a given character can be modified are in general quite di- verse, but for the sake of simplicity we may call them all plus or minus modifications. If the effectiveness of a given plus or minus modifier is changed by mutation, the grade of the character will shift correspondingly. We can conceive of the evolution of the sexual 51 52 THE AMERICAN NATURALIST [Vou. LVI characters of hermaphrodites in terms of successive simple mutations in genes. But to interpret male ` and female forms with observed differences in number or size of chromosomes and with sex-linked inheritaiee re- quires comparison with mutations in which the unit of change is a whole chromosome or section of chromosome instead of a single gene. Such mutations can be under- stood in terms of the action of component genes as fol- lows. Linkage experiments show that the various kinds of genes are distributed pretty much at random among the various chromosomes and along each chromosome. But since the number of genes with a given tendency is relatively small, any particular small section of chromo- some might not contain these genes in the same propor- — tion as they exist in the entire complement, and still less would the normal proportion of every kind be present. The loss of a section of ehromosome (a eondition known as deficiency) would ordinarily remove more minus than plus modifiers (or vice-versa), and since in that case more plus than minus modifiers would remain in action, the grade of the corresponding character would be shifted in a minus direction. This is the interpretation of the fact that a deficiency may cause many character changes, the complex of altered characters being inherited as a domi- nant. When a whole chromosome is lost through non- disjunction, the effects are similar to those in deficiency for a section except that they are greater in degree. The way in which genes act together in producing a character, and the relation of the balance of plus and minus modifiers to deficiency or to the absence of a chro- mosome may perhaps be made clearer by an analogy. Let us suppose that a man is an ardent stamp collector, and has accumulated a lot of stamps. These stamps are to represent genes, so their number may be put at 5,000 to correspond roughly to the number of genes in Drosophila. Among the Russian stamps, especially those of recent issue, there is a very large number of reds, but also a fair number of pinks, and even a few whites. These differ- ences in tint correspond to the plus and minus modifiers of a certain character, namely, the redness of Russian stamps. Now the stamps of different tints are in some definite ratio, whatever that ratio is, and we will call it the No. 642] VARIATION IN SEX CHARACTERS 53 normal ratio or balance. This stamp collector carries his collection around with him, and it fills two big, coat pockets, a trousers pocket, and there are even a few in his vest pocket. But unlike most collectors this one has never taken the trouble to sort over more than a few of his stamps. Meanwhile he strings them together pretty much hit or miss. This stringing stamps together is rather disapproved of by some other stamp collectors, who think that is no way to treat stamps, and each of whom has his own favorite method of arranging them. Because of this hit or miss method of making the strings of stamps the ratio among the different grades of redness of Russian stamps is different in different parts of the strings, and so if some other ardent collector should snip off a piece of one of the strings and earry it away, the re- mainder of the Russian stamps might have a eonsiderably redder tone, while at the same time the Polish stamps might become bluer. If a whole string were lost, then many of the sets of stamps might have quite different complexions. i Now I have been recently studying the effects of the loss of one of the ehromosomes of Drosophila, namely, the small round fourth-chromosome, and the phenomena offer striking parallels to those of dicecious sex, including sex-linked inheritance and sex-limited characters. Indi- viduals having only one fourth-chromosome show a change in many characters, among which may be men- tioned smaller size, smaller bristles, later hatching, poorer viability, paler body-color, darker trident pattern, shorter blunter wings, ete. Each of these differences corre- sponds to a character for which the fourth-chromosome was internally unbalanced, that is, for which the ratio of plus to minus modifiers was different from that of the whole group. For all of the characters in which there was an internal preponderance of plus modifiers the grade will be shifted in a minus direction by the loss of the fourth-chromosome, for example, the shorter wings and paler body-color. Likewise the characters that shift in a plus direction, as the darker trident pattern and the large eyes, are characters for which the fourth-chromo- some possesses an excess of minus modifiers. In the 1 Proc. Nat’l Acad, of Sci., 7: 186-192. 54 THE AMERICAN NATURALIST [Vor.LVI male of Drosophila there is only one X-chromosome, though there is present a Y-chromosome that ean be dis- regarded, since the evidence from non-disjunetion of the X-chromosome shows that it has very little effect upon sex or characters. These individuals with only one X- i Fic. 1. Wild-type (2n) female, with normal chromosome group. chromosome likewise show a complex of characters that are different from those shown by the individuals with the normal two X-chromosomes. Among these characters are gonads and genitalia of a type that we call male. The haplo-X individual is also smaller, has smaller bristles, is less viable, hatches later, and differs in other details from the 2-X type that we call female. Each of these dif- ferences likewise corresponds to a character for whieh the balance of the genes in the X is different from that in the’ group as a whole. The absence of one X leaves in action an unbalanced set of genes which produces male characters. The X-chromosome is a chromosome that is internally unbalanced by an excess of genes that we may eall female-producing. * No. 642] VARIATION IN SEX CHARACTERS 55 In an outeross of a haplo-IV individual to a normal, the entire complex of characters is inherited as a simple dominant and gives a 1:1 ratio, except that the haplo- IV's are less viable. Likewise in outerosses of haplo-X individuals the entire complex of male characters is in- fic. 2. Haplo-IV (2n-1 IV) female, with chromosome group. herited as a simple dominant and gives a 1 : 1 ratio except that the haplo-X's are somewhat less viable. When a haplo-IV individual is mated to a recessive whose gene is in the fourth-chromosome, all the haplo- fourth offspring show this recessive—a behavior that is strictly parallel to sex-linked inheritance; for if a haplo-X individual, that is, a male, is mated to a recessive whose gene is in the X, all the haplo-X offspring show this re- cessive. The fourth-chromosome recessive characters present in haplo-IV individuals from the cross of a haplo-fourth to the recessive show a grade of development that is differ- ent from their grade as homozygous characters in diplo- IV's. This phenomenon is known as ‘‘ exaggeration,”’ and is interpreted as the effect of an unbalance within the 56 THE AMERICAN NATURALIST [Vor. LVI normal fourth-chromosome. With respect to a character that is exaggerated in a plus direction the fourth-chromo- some has an unbalance in the minus direction. But since the whole complement is in balance, this unbalance within the fourth-chromosome is neutralized by a reciprocal un- balance in the other chromosomes. So the removal of one fourth-chromosome with its excess of minus modifiers leaves the remainder of the genes with an excess of plus modifiers, and these plus modifiers are free to work in the same direction as the recessive gene that is present, and thus to give an even greater effect than the homo- zygous recesive. Corresponding to these exaggerated fourth-chromosome characters there is a class of sex- linked characters that are exaggerated in the absence of one X-chromosome. These mutant characters show a different grade of development in the male from that which they show in the female. A good example is the race called eosin, in which the male has a much paler eye- color than the eosin female. These characters exagger- ated by the absence of an X are called sex-limited. Some of them, like eosin, are exaggerated in a plus direction, corresponding to an excess of minus modifiers within the X-chromosome, while others, such as bobbed, are exag- gerated in a minus direction. Thus bobbed, which shows searcely at all in the males, corresponds to an excess of genes within the X tending to make bristles short, and two X-chromosomes can outweigh the genes in the autosomes that tend to make the bristles long, but one X is not enough to do so. When haploidy for the fourth-chromosome is combined with mutants whose genes are outside the fourth-chromo- some there is of course no effect corresponding to sex- linkage, but there is ‘‘ exaggeration.” Thus, haploidy for the fourth-chromosome exaggerates the third-chromo- some mutant Hairless in a plus direction. This type of exaggeration finds its parallel in the 20 or so sex- limited mutations that are not sex-linked. These are mutations whose differential genes are in the autosomes and not in the X and which nevertheless show a different grade of development in the male from that in the female. In these cases also the modifiers of each character are of different weights in the X from the general collection, No. 642] VARIATION IN SEX CHARACTERS 57 and absence of one X leaves a surplus of genes that work in the same or in the opposite direction from that of the mutant in question. Thus, by studying three kinds of effects, first, the character complexes that result directly, secondly, the exaggerations of the mutant characters whose genes are in the same section or chromosome as that involved in the loss, and thirdly the exaggerations of mutant characters whose genes are in other regions, we can analyse roughly the kinds and the signs of the genes that are in the region in question. Since sexual and sex-limited characters are shown to rest on the same genetic basis, namely, a preponderance within the X of the plus or the minus modifiers of those characters, it may be questioned whether there js any real difference between these two categories. If the race of the mutant eosin were to become established in nature, a systematist would certainly include this difference in eye-color among his sexual differences. I am of the opinion that there is no difference between these two cate- gories except that we call those sexuał that are most closely connected with reproduction. There is one striking difference between haploidy for X and haploidy for an autosome—namely, that the changes connected with haploidy for autosomes are rela- tively more numerous and extreme. Haploidy for the second or third autosomes probably produces changes so great as to be lethal, while haploidy for the very small © fourth-chromosome produces changes comparable in ex- tent to all those of the male aside from the reproductive organs. The proportion of sex-limited mutant charac- ters is only about a tenth of the total, while X contains about a quarter of the genes. Since the changes in char- acter produced by absence of an X are relatively small, the internal balance of the X must be relatively high. For a high proportion of the characters of the animal, the plus and minus modifiers in the X must be in about the same ratio as in the group as a whole. The comparison just made between the effects of hap- loidy for an autosome and the effects normally present m dicecious sex shows that they have similar genic bases— namely, each is due to differences in the ratio between two 58 THE AMERICAN NATURALIST [Vor. LVI aggregates of genes; and that the X produces its char- acteristic effects because it contains a preponderance of genes tending to produce the characters that we call fe- male. This point of view receives even stronger and more direct support from a study of cases in which the f KY Tre vun mm f Fic. 3. Triploid (3n) female, with chromosome group. ratio of X-chromosome to autosomes has been changed, and in which new sex relations are present. These new types of chromosome combinations and of sex take their origin in the occurrence of triploidy in Drosophila, for which there is full genetical and cytological proof. The first point is that individuals having three full sets of chromosomes (3n) are females not to be distinguished from normal females except for slight differences in size and proportion that may well be due simply to the greater 2 Science, N. S., 54: 252-954. No. 642] | VARIATION IN SEX CHARACTERS 59 amount of chromatin. The nearly complete identity be- tween the triploid and diploid forms both as to sex and as to non-sexual characters is a splendid evidence that these characters owe their grade to the ratios among the genes, for those ratios are identical in the 3n and 2n forms. Among the offspring of triploid females are individuals that are neither males nor females but are sex-interme- diates, or rather, are mixtures of male and female char- acters, very similar in type to the intersexes of Lyman- iria? Genetical and cytological proof was obtained that these intersexes in Drosophila possess two X-chromo- somes and three sets of autosomes. The old formulation of 2X equals 9 is at once seen to be inadequate, for here we have individuals that have two X-chromosomes and yet are not females. They are shifted out of the female class by the presence of an extra set of autosomes, and thereby the autosomes are proved to play a positive role in the production of sex. Since the intersexes differ from females by the assumption of certain male charac- ters this effect of the autosomes is due to an internal pre- ponderance of ‘‘ male-tendency "' genes. We may now formulate the sex-relations as follows: both sexes are due to the simultaneous action of two op- posed sets of genes, one set tending to produce the char- acters called female and the other to produce the char- acters called male. These two sets of genes are not equally effective, for in the complement as a whole the female-tendency genes outweigh the male-tendency genes and the diploid (or triploid) form is a female. When the relative number of the female-tendency genes is low- ered by the absence of one X, the male-tendency genes outweigh the female and the result is the normal haplo-X male. When the two sets of genes are acting in a ratio between these two extremes, as is the case in the ratio of 2X: 3 sets autosomes, the result is a sex intermediate— the intersex. The intersexes as a class can always be easily distin- guished from normal males and females by reason of their large size, large coarse-textured eyes and by certain other characters such as scalloped wing-margins. Some - of these characters are probably non-sexual effects of the 3 R. Goldschmidt, Zeit. f. ind. Abst. u. Vererb., 23: 1-199. 60 . THE AMERICAN NATURALIST [Vor. LVI triploidy for the autosomes, others are sex-limited. Within the elass of intersexes there is a very wide range of fluetuation, on the one hand to flies that are nearly fe- male and on the other to flies that are entirely male in appearance. In an intersex of a given grade the several ji ^ N ' Un. » L Y N à X 4. Dorsal and ventral views of extreme Peces eel intersex. Ventral view of mid-grade intersex. Chromosome group of intersex showing 2X and 3 sets autosomes. characters do not all present the same intermediate step between male and female, but, apparently just as in the intersexes of Lymantria, some characters are completely male, some completely female, while others are complex mixtures of male and female parts. When the intersexes are classified according to a system of grades, they are seen to be a bimodal class consisting of more ‘‘ female- type " and more ‘‘ male-type "' intersexes, both of which fluctuate widely and overlap considerably. The cytological investigation of the intersexes had shown that there are four sub-types of intersexes that differ in the presence or absence of a Y and in having three or only two fourth-chromosomes. It is possible, and there is some slight cytological and genetical evi- dence in support, that the male- and female-types of No. 642] VARIATION IN SEX CHARACTERS 61 intersexes correspond to the presence of three or of two fourth-chromosomes respectively. There is another connection in which the wide fluctua- tions of the intersexes are interesting, namely, the aetion ( H R f D. I p 1 7/ m NBS AA TUA nma m my WE Fic. 5. Dorsal and nai ipe of extreme female-type intersex. Two chromosome groups, the left with two IV-chromosomes and a Y, the right with two IV's but no Y of environmental factors. The slight range of fluctua- tion in such a character as miniature-wings in Drosophila probably means that there is a critical balance or ratio of plus to minus modifiers beyond which all balances give miniature, at least until the overbalance proceeds so far that a new critical ratio is passed and a new super- miniature character is realized. The balance in minia- ture is so far beyond the critical balance that only rarely are the environmental factors strong enough to outweigh this overbalance and thus cause fluctuation. In mutants in which the overbalanee is slight there will be both wide fluetuation due to environmental interference and a high susceptibility to modification by other genes, as 1S no- toriously the case with Beaded and with Truncate. In normal males and females there are high overbal- ances beyond the critical points, and consequently only slight genetical or fluctuating variations. But in the in- 62 THE AMERICAN NATURALIST [Vor. LVI tersexes these two overbalances in opposite directions cancel each other, and since the two sets of genes are now of almost exaetly the same weight the point of balance is between the two eritieal balances. d Accordingly the char- Fic. 6. *''Superfemale" (2n+X), with two chromosome groups. acters of the intersex fluctuate widely with slight environ- mental differences, and fall into two modes corresponding to the slight difference in balance between two and three of the tiny fourth-chromosomes. RELATION OF SEX TO CHROMOSOMES IN Drosophila me anogaster X-chromo- Sex somes Buperiemale. 5s ei es 3 aage le Tt PEPEE E ee wee 3 Female diio a OE pepe EN 2 —bDUDO Leve LS TERR EVEN 2 Intersex HVE is Lu IRE CIR 2 Male.) eov our TEES 1 Buüpermalé >i iaooank a ae E Era oe 1 Sets Autosomes| Sex Index B. 1.5 3 1 2 1 3(—IV) a 67+ 3 .67 2 5 3 33 The phenomenon of intersexuality might be expected to have a reciprocal phase—namely, supersexes. If the No. 642] © VARIATION IN SEX CHARACTERS 63 intersexes result from an intermediate ratio of X to auto- somes because the X has a net female tendency, then it might be expected that by increasing the ratio of X to autosomes a superfemale would be produced, and con- versely, a supermale by increasing the relative number of rA DANS u DANT, tA n Pig; i Supermale.” No cytological SAIS genetical tests show 1X and 3 sets autosomes. autosomes. Diploid individuals with an extra X-chromo- . some (2n plus X) have now been identified among the progeny of certain strains of high non-disjunction, among the offspring of triploid females and elsewhere. These flies resemble females but are very inviable and form a distinet character type. They are sterile and sections of the gonads show abnormal ovaries. These differences all result from the unbalance within the X, and are there- fore of the sexual-sex-limited category. That these dif- ferences are not greater is partly due to the same high internal balance of the X that we met with in analysing males and intersexes, and is partly to be explained on the ground that for many of the characters the overbalance is not yet great enough to pass a second critical point. Conversely, individuals with one X-chromosome and an extra set of autosomes have been identified among the offspring of triploid females. These are males distinctly different from normal males and sterile. If there were time, it would be interesting to supple- ment and modify the view just presented by comparisons with the rich materials elsewhere, and perhaps to specu- late as to how this machinery was evolved, and how the genes involved come to expression physiologically. THE NATURE OF BUD VARIATIONS AS INDICATED BY THEIR MODE OF INHERITANCE' PROFESSOR R. A. EMERSON CoRNELL UNIVERSITY Tue title limits this account to such bud variations as have been studied critically with respect to their inheri- tance in sexual reproduction. The further limitation of time makes it necessary that I choose from among such studies certain cases to serve as illustrations of the sev- eral types of bud variation. I shall, therefore, attempt no complete review of the researches bearing on the prob- lem at hand. A survey of published accounts of bud-variation studies shows that as yet comparatively little is definitely known of the real nature of these vegetative sports. It seems not unlikely, however, that to point out some of the prob- lems suggested by these studies and, where possible, to note modes of attack may serve the purpose of this sym- posium quite as well as a rehearsal of known facts and their interpretation. As here used, the term bud variation is synonymous with vegetative as contrasted with seminal variation. The term somatic variation may also be employed to the same effect, provided it is not thereby intended to exclude cases in which the germ tract as well as the soma is in- volved. At the outset, however, there must be imposed on any of these terms, for the purpose of this discussion at least, the limitation that the variation involves a change in the genetic constitution of the parts affected. The expressions somatic mutation and somatic segre- gation are specific terms and as such are not to be used interchangeably with the more general terms somatic, vegetative, or bud variations. Moreover, to speak of a particular vegetative variation as a case of somatic muta- 1 Paper No. 94, Department of Plant Breeding, Cornell University, Ithaca, New York. 64 No. 642] . NATURE OF BUD VARIATIONS 65 tion or of somatic segregation without basis from critical inheritance or cytological studies is to prejudge the nature of the observed modifieation. FREQUENCY or Somatic VARIATIONS Attempts have been made to estimate the relative fre- quency of vegetative and seminal variations in plants, but little definite information has been gained. The problem is beset with grave diffieulties inherent in most attempts to determine coefficients of mutability. The possibility of overlooking even prominent variations until they have once been noted, together with the readiness with which they are found after one’s attention has been focused on them, will hardly be questioned by anyone who has given attention to the discovery of new variations in almost any organism. One may attempt with some assurance an estimation of the frequency of recurrence of a particular mutation, for instance, whether it appears in vegetative parts of individuals or in sexually produced progenies, but it is a hazardous undertaking to estimate the fre- quency of variations in general. Until some one can de- vise a scheme for estimating the frequency of bud varia- tions as Muller has done for determining mutation fre- quencies in Drosophila, little progress can be looked for other than through investigations of the somatic muta- tion or segregation of specific genes. The problem of the relative frequency of occurrence of somatie and gametie variations meets the further diffi- culty that it is often impossible to determine the ontoge- netie stage at which particular variations have arisen—a faet that has been noted for plants by various writers (deVries, 1910; Emerson, 1913; East, 1917). Both Bridges (1919) and Muller (1920) have discussed this problem from the standpoint of studies of partieular mu- tations in Drosophila. The prevalent opinion that varia- tions arise in the gametes or at about the time of their formation may have come in part from a belief that aber- rant chromosome behavior is most likely to occur at the time of the reduetion division. It seems likely, however, that the situation has been confused by failure to realize 66 THE AMERICAN NATURALIST . [Vou.LVI that recessive mutations—the most frequent kind—can not be expressed in the individual in which they occur except when the dominant allelomorph is simplex, while such mutations may appear in a later generation of sexu- ally produced progeny (East, 1917). Somatic Mutation OF GENES Several cases of vegetative variation in plants have been studied with sufficient thoroughness to leave little doubt that they are mutations in the strict sense, in- volving the modification of particular genes. Most of them are concerned with variegated color patterns of flowers, leaves, or fruits, and they are more or less regu- larly recurrent, a fact that makes them especially well suited to quantitative studies, for it is obvious that a quantitative study can be made only of variations that occur. with considerable frequency. For the most part also these somatic mutations are dominant to the type from which they spring, appearing frequently in material homozygous for their recessive allelomorphs, facts that exclude the possibility of their being due to any sort of somatic segregation of unlike genes. Blakeslee’s (1920) case of a somatic variation in Portulaca is one of the few examples not involving variegation. Other cases have been reported by Baur (1918). One of the earliest cases of somatic mutation was re- ported by deVries in variegated flowers of Antirrhinum. Though the work was done prior to the rediscovery of Mendelism and not discussed from the standpoint of re- cent genetic interpretation, there is little doubt, as I have noted elsewhere (Emerson, 1913), that the results can best be interpreted as due to a somatie gene mutation. Correns's (1910) results with respect to the occurrence and behavior in inheritance of green-leaved variations on variegated-leaved Mirabilis and of self-colored flowers on variegated flowered strains of the same species were among the first to be subjected to critical genetic analysis. The behavior in inheritanee of green branches of varie- gated Mirabilis shows this vegetative variation to be a simple dominant mutation affecting ordinarily only one . of the duplex recessive allelomorphs. A mutated branch No. 642] NATURE OF BUD VARIATIONS 67 is, therefore, as truly a heterozygote as if it had arisen through hybridization of green and variegated strains. Self-colored branches on variegated-flowered plants of Mirabilis usually do not transmit the self-color character to their seed progenies in greater percentages than do variegated-flowered branches of the same plants. They are thought by Correns to be fundamentally of the same nature as the green branches of variegated-leaved plants, their failure to transmit the self-color character being due presumably to the accident that the mutation occurs in epidermal cells from which no gametes arise. The fre- quent occurrence of self-colored plants in seed progenies ` both of self-colored and of variegated flowers is consid- ered evidence of their origin as vegetative rather than as gametic mutations, their failure of expression in the soma being thought due to their origin in sub-epidermal cells in which these flower colors do not develop. Studies of variations in variegated pericarp of maize by myself (Emerson, 1914, 1917) and by Anderson, Eyster, and Demeree;? involve practically the same results as those so far reported in investigations of other species and afford in addition quantitative data on certain as- pects of the somatic-mutation problem not included in other investigations. The genes for variegated pericarp have been shown to belong to a comparatively large series of multiple allelomorphs including those for colorlessness (white seeds), self color of different intensities, and cer- tain definite color patterns of both the pericarp of the seeds and the glumes and pales of the cobs. Variegation is known to be a simple recessive to self color and a domi- nant to white. Self-colored seeds whether occurring singly or in groups in variegated ears produce progenies consisting of approximately 50 per cent. self-colored ears, the other 50 per cent. being either all variegated or all white de- pending on whether the parent was homozygous varie- gated, V V, or heterozygous variegated, V W, from a pre- vious cross with white. Seeds that are less than wholly self colored throw a correspondingly smaller per cent. of 2 Unpublished data by W. H. Eyster and E. G. Anderson, and by E. G. Anderson and M, Demeree, 68 THE AMERICAN NATURALIST . [Vor.LVI self-colored ears. Self-colored seeds thus produced have, so far as tested, proved to be heterozygous for self color, behaving in later generations exactly as if produced by crosses of self-colored with variegated or with white races. Certain cultures of self-colored maize produce a few variegated seeds. Such seeds have been observed only on ears that are heterozygous from previous crosses with variegated strains, S V, or with white strains, S W, never from ears that are homozygous for self color, S S. From such variegated seeds, new variegated races have been produeed. These facts are regarded as indicating (1) that the oe- currence of self-colored or partly self-colored seeds on va- riegated ears is due to somatic mutations of the recessive variegation gene to the dominant self-color allelomorph; ` (2) that only one of the two variegation genes of homo- zygous variegated maize mutates at a given time; (3) that it is always the variegation gene, never the white one, of heterozygous material that mutates; (4) that the oe- currence of variegated seeds on otherwise self-colored ears is due to reverse mutations from the dominant self- color gene to the recessive variegation allelomorph; and (5) that only one of the duplex genes of self-color strains so mutates at any one time, for otherwise there would re- main no dominant self-color gene to prevent the expres- sion of the mutation as variegated seeds in Hungop. self-colored material. Another type of somatie variation, quite distinet from the self-color mutations discussed above and often termed dark-erown variation, also occurs frequently in varie- gated maize pericarp (Emerson, 1917). It is quite as striking in appearanee as the self-color mutation, but is not inherited, the progenies of the aberrant seeds being in no way different from those of the normal seeds of the same ears. Microscopic examination of dark-erown and of self-color seeds indieates that in the former the epi- dermis alone is colored while in the latter the epidermis alone remains colorless. The conclusion seems war- ranted, therefore, that the two types of variation are fun- damentally the same, both being true gene mutations, and No. 642] NATURE OF BUD VARIATIONS 69 that the non-inheritance of the dark-erown type is due to the accident that it occurs in epidermal tissue outside the germ traet. Recent investigations of variegated maize by Eyster and Anderson have established the fact that somatic mu- tations affecting small areas occur much more frequently than those affecting large areas. Since a mutation aris- ing in a single cell late in development obviously could not affect so large an area as one originating earlier, it follows that mutations in variegated maize occur with increasing frequency in the later stages of ontogeny. It is true, as pointed out by Muller (1920), that given a con- stant rate of mutation throughout all stages of ontogeny and granting that one cell is as likely as another to mu- tate, mutations should appear more frequently in the later stages of development because of the fact that there are then many more cells in which mutations may arise. But Eyster and Anderson have found that the increase in the frequency of occurrence of mutations during the progress of development is accelerated far beyond ex- pectation based on the increase in number of cells. This behavior is strongly suggestive of a progressive acceleration in the mutability of the variegation gene as development proceeds. It is much too early to say whether this progressive change, if such it be, is inherent in the organization of the gene itself, as suggested by Anderson and Demeree, or whether it is a response to progressive changes in physiological and environmental relations. Perhaps the assumption of an equal chance of mutation as between any two cells is without sufficient warrant. Possibly there is a time element to be taken into account, as noted by Muller (1920). As cell division becomes progressively retarded in the late growth stages, may not each cell be exposed for an increasingly longer period of time to the chance of mutation? Perhaps it may be possible to test this assumption in favorable ma- terial by a comparison of the frequency of mutation in the very early slow-growth, the later rapid-growth, and the final slow-growth periods of the life cycle; but the relatively few cells present in the very early growth period seems likely to place serious limitations on the 70 THE AMERICAN NATURALIST [Vou.LVI practicability of such a test. An observation of possible importance in connection with the question of a time ele- ment in mutation and with the problem of environmental and physiological influences is that made by Eyster and Anderson concerning the greater frequency of the non- heritable (epidermal) mutations than. of the heritable (sub-epidermal) ones in variegated pericarp of maize. I have recently obtained results bearing on another phase of the somatie-mutation problem as related to variegated maize pericarp, namely, the relative frequency of mutation of homozygous, V V, and of heterozygous, V W, material. It has been shown above that the W gene for colorless (white) pericarp does not mutate, so far as known, when paired either with itself, W W, with the va-* riegation. gene, V W, or with the self-color gene, S W. It will be recalled further that only one of the two homolo- gous genes in homozygous variegated, V V, material mu- tates at any one time. If it could be assumed that the mutability of either allelomorph is uninfluenced by the presence of the other, it should follow that somatic muta- tions will occur with approximately twice the frequency in homozygous, V V, as in heterozygous, V W, material. But this expectation has not been realized. On the con- trary, both heritable (self-color) and non-heritable (dark- crown) mutations have appeared throughout all my cul- tures with somewhat greater frequency in heterozygous than in homozygous variegated ears. The difference has been especially pronounced in very light variegated strains, where mutations have appeared about two and one half times as often in heterozygous as in homozygous material. Even if mutations appeared with equal fre- queney in heterozygous and in homozygous ears, the simplex gene of the former must have a mutability of about twice that of either of the,duplex genes of the latter. In the very light variegated strains, therefore, a simplex gene must have a mutability of about five times that of a duplex gene. What appears to be a similar result in Mirabilis has been reported by Correns (1903, 1904). Crosses of a supposedly pure white race with several self-colored pink yellow, and pale yellow races resulted in every case in No. 642] NATURE OF BUD VARIATIONS 71 plants with strongly red-striped flowers and with numer- ous self red flowers or even whole branches of such flowers. Intererosses of the pink and yellow races gave only self-eolored progeny, from which fact it was con- cluded that the white-flowered race carried a latent factor for striping. It was later discovered that about three per cent. of the flowers of the white race showed minute flecks of red. It was evidently an extremely light, varie- gated race, rarely if ever throwing somatic self-color mutations when the variegation gene was duplex (homo- zygous material) but producing such mutations with con- siderable frequency when that gene was simplex (hetero- zygous material). Correns concluded that red variega- tion of Mirabilis flowers is a character that, with self- fertilization or inbreeding, remains almost completely latent, but which, through the entrance of foreign germ plasms, is brought to full expression. If the mutability of a gene can be increased through the influence of some modifying factor or factors brought into combination with it by crossing, as suggested by Correns, it should be possible to discover crosses that = would not produce the effects so far observed in Zea and Mirabilis. While the problem deserves much more study from this viewpoint, it seems unlikely that results with maize can be explained on any such basis, unless the postulated modifying factor is the allelomorph of the variegation gene or some factor very closely linked with it. It must be noted in this connection that the compari- son in maize was made between homozygous and hetero- zygous variegated ears of the same F, progenies grown from self-pollinated F, heterozygotes—a circumstance that would afford abundant opportunity for recombina- tions of independently inherited modifying factors. That the differences in mutability noted in maize may be due to differences in the interaction of like as contrasted with that of unlike allelomorphs, as suggested by Anderson and Demeree, is a somewhat novel conception worth care- ful consideration if means can be devised for subjecting it to a.erucial test. Before the topie of somatie mutation is dismissed, it should be noted that the pups is not limited to 72 THE AMERICAN NATURALIST [Vor. LVI plants. Among animals, Drosophila (Morgan and Bridges, 1919) has furnished several examples of un- doubted somatic mutation resulting in mosaic individuals other than gynandromorphs. SoMATIC SEGREGATION Bud variations have probably been ascribed to somatic segregation more frequently than to any one other cause. Perhaps the opinion commonly held that bud variations occur more frequently in hybrids than in other material and the long known fact that seed-grown offspring of hybrids exhibit segregation, is chiefly responsible for this usage. It is, of course, possible that most vegetative variations are of this nature, but the fact that the indi- viduals in which they arise are frequently found to be heterozygous for the genes concerned is no conclusive evidence that segregation is involved. Mutations also, as noted by several writers, are most likely to appear in heterozygous material because most of them are recessive and the unmutated dominant allelomorphs prevent their expression in the individuals in which they originate if the latter are homozygous. Chromosome Elimination.—The best examples of so- matic segregation that have been subjected to critical genetic analysis are afforded by the work with Droso- phila. lt has been shown by Morgan and Bridges (1919) that, of the relatively numerous gynandromorphs which have appeared in the course of investigations with Dro- sophila, nearly all have resulted from the elimination of the sex chromosome at some early cleavage division. If a fertilized egg starts as a female, XX, and one X chro- mosome is eliminated at an early segmentation that part of the individual developing from the cell that receives but.one X chromosome should be male, XO, while the re- maining part should be female, XX. The evidence in support of this view was obtained from erosses the parents of which had different sex-linked and different autosomal characters, that is, characters whose genes are carried by the sex chromosomes and by the autosomes, respectively. The male, as well as the female, No. 642] NATURE OF BUD VARIATIONS 13 side of gynandromorphs appearing in such crosses ex- hibited all the dominant autosomal characters whether they eame from the maternal or the paternal parent. When the mother had a recessive, mutant gene in one of her autosomes and the father had its dominant, normal allelomorph, the faet that the male side of gynandro- morphs did not have the maternal, recessive autosomal charaeter effeetively disposed of Boveri's hypothesis of partial fertilization. On the other hand, when a recessive autosomal gene entered from the father's side and its dominant allelomorph from the mother's side, the fact that the male side of the gynandromorphs did not show the paternal, recessive character likewise eliminated Morgan's earlier hypothesis of polyspermie fertilization. It has been shown, further, from erosses, the parents of whieh differed in sex-linked charaeters, that maternal and paternal X ehromosomes are eliminated with about equal frequency. i In certain experiments with Drosophila, in which a de- termination of the frequency of sex-chromosome elimi- nation was undertaken, it was found that one gynandro- morph appeared in about every 2,200 individuals. Since only those individuals that start as females give the kind of gynandromorphs observed in these tests, it was con- cluded that one ease of chromosome elimination occurs in about 1,100 individuals. Of the evidenee from plant material there is the recent account by Frost (1921) of the occurrence of a bud sport in Matthiola in which presumably linked genes have segregated out simultaneously in one or more branches. While this ease will require further investigation before the manner of its origin ean be positively established, it seems probable that it belongs to the category of somatie segregation by chromosome elimination or non-disjune- tion. Studies of mosaie endosperm of maize afford perhaps the most definite evidence available in plants that certain somatie variations are due to aberrant chromosome be- havior such as non-disjunetion or elimination ( Emerson, 1921). The genetic evidence that I have been able to ob- tain in support of this interpretation is of much the same 74 THE AMERICAN NATURALIST [Von LVI. nature as that noted above for Drosophila gynandro- morphs. In crosses in which recessive aleurone and endosperm characters are contributed by the female parent and their dominant allelomorphs by the male parent, spots of the recessive (maternal) aleurone color are underlaid by the recessive (maternal) type of endo- sperm when the genes for these aleurone and endosperm characters are genetically linked, that is, when they are carried in the same chromosome. On the contrary, simi- lar recessive (maternal) aleurone-color spots are always underlaid by the dominant (paternal) type of endosperm when the genes are not linked, that is, when they are carried in non-homologous chromosomes. The fact that linked genes separate out simultaneously while non-linked ones do not do so supports the view that mosaic seeds are the result of some chromosome aberration such as elimi- nation or non-disjunetion, and renders untenable the earlier hypotheses of incomplete fusion of endosperm nuclei suggested by Correns and by Webber and also that of gene mutation proposed by myself. The work with aberrant maize endosperm has fur- nished an opportunity to study the frequency of chromo- some aberrations in a specialized tissue. The available data show that when a single chromosome alone is con- cerned, about one mosaic seed occurs in every 420 seeds. If the other two homologous chromosomes of any one set are involved as frequently and if any one of the ten trip- loid chromosome sets is as likely to be involved as any other one, one ease of aberrant ehromosome behavior should occur in about every fourteen seeds. There 1s some evidence, though not convineing as yet, that in dif- ferent strains of maize chromosome aberrations may occur with strikingly different frequencies. In one cul- ture in which only a single chromosome could have been involved in the origin of mosaic seeds, as many as twenty- five such seeds have been observed on a single ear of ap- proximately 500 seeds, or one for each 20 seeds. If this behavior proves to be a constant one in this strain and if the other 29 chromosomes behave in like manner, it should furnish excellent material for cytological investi- gation. Moreover, the possibility of the existenee of No. 642] NATURE OF BUD VARIATIONS 75 strains of maize differing so widely in the frequency of chromosome elimination or non-disjunction raises inter- esting questions concerning the causes of such aberrations. It would seem possible to determine by appropriate tests something as to the relative influence of maternal and of paternal contributions on the rate of chromosome elimi- nation. There are circumstances connected with these results from Drosophila and Zea that may raise some doubt of their general applicability to cases of bud variation. The Drosophila evidence is limited almost exclusively to the .sex chromosomes, though there is no positive evidence that elimination may not occur among autosomes and re- sult in non-viable individuals. The data from Zea re- lates to endosperm alone, a specialized, nutritive, sterile, triploid tissue. There is perhaps justification for a be- lief that the sex chromosomes of animals and the triploid chromosomes of the endosperm of angiosperms may be subject to irregularities in behavior not commonly found in other material. The only answer to such a contention is (1) that gynandromorphs and endosperm mosaics are the materials that have been critically studied and (2) that there is, or should be, no presumption in favor of vegetative segregation through chromosome elimination or through other means as against vegetative mutation or any other mechanism as a possible explanation of bud variations that have not been subjected to cytological in- vestigation or to critical genetic analysis. Cytoplasmic Segregation.—Numerous cases of ap- parent segregation of cytoplasmic elements have been reported in plants. Of these, examples from Mirabilis, Pelargonium, Primula, and Zea may be noted. All of them involve visible effects on chlorophyll development and all show non-Mendelian inheritance. ` Correns (1909a, b) working with a white-spotted-leaved type of Mirabilis observed a very irregular distribution - of the white and green areas, each varying from small spots to whole branches. These white and green char- aeters were found to be inherited through the mother only. The situation with respect to Pelargonium, re- w 4 THE AMERICAN NATURALIST [Vot.LVI ported by Baur (1909), differs from that in Mirabilis in that the spotting is transmitted through the pollen as well as through the egg cells. Spotting appeared in F, in crosses of white with green without respect to which way the cross was made. As in Mirabilis, wholly white and wholly green, as well as mosaic, branches were ob- served. Examples of maternally inherited chlorophyll variega- ~ tion have been investigated by Gregory (1915) in Primula, and by Anderson? in Zea. The genetic behavior of these materials is quite the same as that of Correns’s Mirabilis variegation. The apparent difference in the cytological : basis of their behavior, however, must not be overlooked. Evidently these plants of Mirabilis, Pelargonium, Primula, and Zea are sectorial chimeras. Their main interest in connection with this discussion lies in the fact that, starting with a single fertilized egg cell, certain chlorophyll deficiencies are apparently separated out into certain vegetative cells and handed on through definite cell lines, while normal chlorophyll develops in other cell lines, with the result that areas of varying extent have one or the other of these characters. In what the mechanism of this segregation consists—if segregation it be—is not in all cases certainly known. It may even be that some eases of variegated chlorophyll are to be re- garded as recurrent variations arising de novo after the manner of somatic mutations but effecting changes in the cytoplasm, or some of its inclusions, rather than in the chromosomes. Baur is inclined to the view that in mosaic plants of Pelargonium deformed chloroplasts are respon- sible for the chlorophyll deficiencies and that these are segregated out by chance in cell division. This view is supported by Gregory, who noted in the young leaves of variegated plants of Primula the existence of normal and chlorotic plastids in the same cells. Correns does not commit himself to any particular element or inclusion of cytoplasm as the seat of the cause of chlorophyll defi- ciency. Randolph (1922), from cytological examination of Anderson’s striped leaved maize, found that, in the 3 Unpublished data, No. 642] NATURE OF BUD VARIATIONS ? 77 transition regions between the green and the pale-green areas, the cells contain not only green and colorless plas- tids, but all intermediate conditions as well. Since the green and the white plastids are not two sharply differen- tiated kinds, but are the end members of a continuous series arising from minute primordia which, so far as ean be seen, are of one kind, he regards any simple form of segregation hypothesis as inadequate. It seems possible, however, that these primordia may be functionally, even though not morphologically, of two more or less distinct classes. Graft-hybrids and Other Chimeras.—' The well-known graft-hybrids of Solanum reported by Winkler are of interest from the standpoint of this discussion because of the bud variations commonly exhibited by them. Sec- torial chimeras, produced by adventitious buds arising from the point of union of stock and scion of grafts of tomato and nightshade, and having one side of the one species and the other side of the other, have not infre- quently later produced branches that were periclinal chimeras having tissues of one species enclosed within an envelope of the other. That these branches are really periclinal chimeras has been established by chromosome counts and by the fact that seedlings produced by them are always of the species of the subepidermal tissue from which gametes arise. These periclinal chimeras in turn have been observed to produce branches wholly of one or other of the parent species. The marked difference in appearance between the sectorial and periclinal chimeras and between the latter and either parent species places this behavior clearly in the class of bud variation and, since the production of branches of the parent species from periclinal chimeras is the result of a separation of genotypes that were closely united previously, the phe- nomenon is perhaps rightly classed as a form of vegeta- tive segregation. It is obvious, however, that the separa- tion of tissues that are merely closely associated in the graft hybrid is a fundamentally different type of segre- gation from that by which the chromosomes or even the ° E 78 THE AMERICAN NATURALIST. [Vor.LVI plastids or other cytoplasmic elements of a single cell are dissociated. The behavior of ‘‘ natural”? periclinal chimeras of Pelargonium, noted by Baur (1909), and of Pelargonium and several other forms, described by Bateson (1919), all of which involve green and white regions of the plants and some of whieh produce reverse periclinal chimeras, is fundamentally the same as that of graft-hybrids. The manner of origin of these natural chimeras is unknown, but it is quite possible that they arose as somatic muta- tions. = The ease of Bouvardia also, as reported by Bateson (1916), is presumably of quite the same order as the examples noted above, though its behavior is strikingly different in detail. Varieties of Bouvardia that are maintained true to type by propagation from stem cut- tings produce plants with very different flower form, size, and color when propagated by root cuttings. While this behavior is not to be taken as positive proof that these varieties are natural periclinal chimeras, it is quite in keeping with such an assumption. Since in normally produced buds of the stem both the epidermis and the deeper lying tissues are maintained through direct cell lineage, while the roots produced by stem cuttings arise from the plerome and break through the periblem and dermatogen, forming these parts anew, sprouts that de- velop from the roots must have the genotype of the stele rather than that of the cortex or epidermis. . From the results of critical investigations cited in this account, it is evident that vegetative variations are due to diverse causes. Some are certainly due to somatic mutation of genes; others are as certainly due to chromo- some aberrations; and still others have been somewhat definitely shown to involve a vegetative segregation of plastids or other cytoplasmic elements. There are many problems relating to these several types of behavior that are in great need of further critical study both genetic and cytological. The results of future research will de- ' pend in large measure on the choice of favorable ma- terial. Quantitative data are of the greatest importance No. 642] NATURE OF BUD VARIATIONS 79 and from this standpoint no material gives more promise of fruitful results than that involving variegation. LITERATURE CITED Bateson, W. Root-euttings, Chimeras and ‘‘ Sports.’’ Jour. Genetics, 6: 16. Studies in Variegation. Jour. Genetics, 8: 93-98. 19. Baur, Erwin. Das Wesen ugs die Erblichkeitsverhiiltnisse der ** Varietates PRE ESY hort.’’ Von Pelargonium zonale. Zeitschr. indukt, Ab- . Vererb., 1: 330-351. 09. pae i von ette majus. Zeitschr. indukt. Abstamm. Vererb., 19: 177—193. sap ones A. F. A Dwarf Mutation in Portulaca Showing Vegetative Re- ns. Genetics, 5: 418-4 920 Bridges, udi B. The Heine Stages at which Mutations Occur in the Germ Tract. Proc. Soc. Exp. Biol. and Med., 17: 1-2. 1919. Correns, C. Ueber Basta —M mit Mirabilis-Sippen. Ber. : 594-608. 1903. ——— Zur Kenntnis der schei werd neuen TAN der Bastarde. Ber. Deutsch. Bot. Gesellsch., 23: —— Vererbungsversuche mit a ie ae und buntblattrigen = pe ei Mirabilis Jalapa, Urtica pilulifera, und Lunaria annua Zeitschr. indukt, ps Format Vererb., 1: 291-329. 1909a. Zur Kenntnis der Rolle von Kern und drome bei der Vererbung. Zeitschr. indukt, Abstamm. Fore, 2: 331-3 909b. —— Der Übergang aus dem homozygotische is einen heterozygotischen Zustand im selben Individuum bei buntblättrigen und gestreiftbluhen- den Mirabilis-Sippen. Ber. Deutsch. Bot, Gesellsch., 28: 418—434. 1910. . deVries, Hugo. The Mutation Theory (translation by J. B. Farmer and A. D. Darbi 1910. Emerson, R. A. The Possible Origin as Mutations i in Somatic Cells. AMER. NAT., 47: 375-377. 1913. The Inheritance of a Recurring Somatic Variation in Variegated Ears of Maize. AMER. NAT., 48: 87—115. 1914 Genetical Studies of Variegated Pericarp in Maize. Genetics, 2: 9 curro drohen. fividence of Aberrant TaS Behavior in Maize Endo- sperm. Amer. Jour. Bot., 8: 411-4 1921. Frost, Howard B. An TTE e a Somatie ee involving Two Linked Factors. AMER. Nat., 55: 461-464, Gregory, R. P. On Variegation in Prima sinensis. e. Genetics, 4: Pig" : Morgan, T. H. and Bridges, C. B. The Origin of Gynandromorphs. Car- negie Inst. Wash., Pub, 278: 1-122. 1919. Muller, H. J. Further Changes in the White-eye Series of Drosophi'a and their Bearing on the Manner of Occurrence of Mutation. Jour, Ezp. Zool., 31: 443-473. 1920. Randolph, LF E Oytological Study of Some Chlorophyll Types in Maize, Bot. Gaz., 1992. (In press.) SEROLOGICAL REACTIONS AS A PROBABLE CAUSE OF VARIATIONS PROFESSOR M. F. GUYER UNIVERSITY OF WISCONSIN WrirH an insight that has never been surpassed even to this day Claude Bernard,' more than forty years ago, remarked that ‘‘ Organic synthesis, generation, regenera- tion, maintenanee, and healing of wounds, are different aspects of an identical phenomenon," the phenomenon alluded to being the constructive activity manifested in ordinary nutritive processes. In a recent thoughtful paper, R. S. Lillie? reiterates and expands this point of view. That synthetic metabolism constitutes the very essence of embryonic development and therefore of the expression of heredity, scarcely admits of a doubt. To- day it is a truism to say that the visible ‘‘ characters "' we deal with in heredity are but the effects—by-products as it were—of far-reaching metabolic reactions. And since the metabolism of the actual living protoplasm cen- ters, if not exclusively, at least principally, in the pro- teins, the problems of metabolism, growth, reproduction and heredity become largely the problem of why and how a given kind of living protoplasm builds up proteins of its own specific type. The molecules of the ordinary native proteins are, as is well known, huge polymeric structures of extremely complex constitution. By appropriate chemical treat- ment they may be broken down into successively smaller and smaller units, each of which, however, still responds to the ordinary qualitative tests for proteins. There finally comes a point below which further reduction of the molecule results in the loss of the distinctive protein reaction, and the outcome is a series of ultimate char- acteristic units, the amino-acids. Thus native proteins seem to be built up of two different categories of units: first, combinations of various amino-acids which consti- tute the simplest protein blocks; and second, the combina- tion of these into the much in molecules eharaeteristie of the native proteins. 1**Tecons sur les phénomènes de la vie,” Vol. IL p. 514, 2 Biol. Bull., XXXIV, 2, 1918. 80 No. 642] SEROLOGICAL REACTIONS 81 In the process of digestion different proteins are broken down into their amino-acid units and these are then re- built into the tissue-proteins of the living organism, each tissue selecting such amino-acids as are required to re- construct its own peculiar complex. That is, the architec- ture of the new proteins into which the individual building units are regrouped is determined by the specific con- stitution of the tissue-proteins themselves. In different proteins the different amino-acids may exist in very dif- ferent ratios, and certain of them necessary for the meta- bolic repair of protoplasm may be lacking in some, such as gelatin, but the amino-acids in any particular protein are constant in nature and proportion and each probably has a definite position in the molecule. While kinds and proportions of amino-acid units determine in large meas- ure the characteristics of individual proteins, it may well be that configurational differences in molecules of the same chemical composition are responsible for the specjfi- cities of corresponding proteins in related species of animals. One estimate, for instance, assigns to the serum- albumin molecule alone the capability of having as many as ten thousand million stereoisomers. One may perhaps picture mentally, in a much simplified form, the simplest protein molecule as a main chain or ring, of which the representative links are amino-acid ‘‘ nuclei.” More- over, to each such link (e.g., in simplest form, dicere H a side-chain, differing in constitution in different cases, is attached or is attachable by replacement of a hydrogen atom. It is, then, with such complex molecular configurations that we have to do as a chemical basis for the phenomena of life; and in them, as I have stated elsewhere,’ we have ** ample basis for that peculiar handing on of metabolic energies already established which we term heredity.” Although habitually when speaking of heredity we think of the multifarious ** characters ”’ displayed by the adult . organism as the things inherited, and strive to picture in our minds how they are represented in the germ, it is 3 Amer, Nat., XLV, May, 1911. 82 THE AMERICAN NATURALIST [Vor. LVI clear, in the light of modern geneties, embryology and cytology, that what actually happens is a reduplieation generation after generation of germinal protoplasm. That is, the proteins already present in the germ-cell not only determine what will be built up in growth, but also the composition of that overgrowth which, as a detached individual, eonstitutes the physieal basis of inheritance. If the initial protoplasmic substance is chemically specific then inevitably the anatomical and physiological com- plexities which arise out of it must likewise be specific. Before we can grapple with the problem of the possible induction of changes in this fundamental mechanism through influences emanating from the body, we must consider some matters concerned with embryonic develop- ment and the fundamental chemical nature of the somatic cells. As to how the constitution of the egg becomes trans- formed into that of the adult, the most consistent and reasonable hypothesis to date, in my opinion, is that proposed by Child, based on axial or metabolic gradients. A full exposition of his hypothesis must be sought in his books, ‘‘ Individuality in Organisms "' and ‘‘ The Origin and Development of the Nervous System from a Physio- logical Viewpoint." I can sketch only such aspects of it as pertain to my present subject. Starting with the universally aecepted biological’ axiom that excitability followed by some degree of transmissibility is a funda- mental property of all living matter, Child believes, as I understand him, that the establishment of polarity in the fundamental organismie proteins of the germ-cell is the beginning of development. 'The eggs of many species al- ready show polarity (animal and vegetal pole) at the time of ovulation; in other forms polarity is not established until later. In the former case the polarity may have been determined in earlier cell-generations by extrinsic faetors or it may be due to the original position of the ovum in the ovary with reference to the nutritive stream. Yolk apparently accumulates in the region of least oxida- . tion and thus marks the vegetal pole. In the second type of egg, polarity, at first lacking, is soon established beeause differential environmental exposure (difference in oxygen supply, light, contaet, general surface exposure, No. 642] SEROLOGICAL REACTIONS 83 or other external factor or factors) causes one region to have the greatest metabolie aetivity. As does any stim- ulated part with reference to a resting part (muscle, nerve, ete.), this point of heightened activity sets the pace, as it were, for the other parts. Thus an excitation gradient is established which Child terms an avial gra- dient. Since the excitation initiates transmission— changes and thereby determines what shall happen at successive levels along the path of transmission, the region of highest excitation dominating and controlling the rest, the axis in question may also be called a met- abolic gradient. In this way a physiological unity is established and maintained by bringing the different regions within range of the gradient into definite physio- logical relations. Since the chief source of energy in protoplasm is oxida- tion, and inasmuch as many different tests have shown that the rate of oxidation gradually diminishes along the gradient from the region of highest activity, Child infers that differences in oxygen supply play a very important part in the local metabolic differences which arise. In any event, a differential axis of activity arises in the egg as the result of differences in environmental conditions (either before or after ovulation) and determines the axiate pattern of the developing organism. In such a complex system of chemical and physical activities, where unquestionably many associated simultaneous reactions and interactions are going on, different rates or condi- tions of reaction in different regions must result in unlike end-products. Thus, at different levels of a gradient which was quantitative in origin, qualitative differences arise. For once a gradient is established, any one of several purely quantitative changes in the system, such as increased oxidation which acts differentially on sub- stances at a given point, changes in temperature, in water content, or in colloidal state, or changes in the concen- tration of the reacting substances, may alter certain com- ponent factors of a given region more than it does the - corresponding factors in other regions, with the result that out of the same initial constituents the respective . organ- or tissue-stuffs that characterize the organism are gradually built up. For example, as Child points out, 84 THE AMERICAN NATURALIST [Vor. LVI ïn a region of rapid oxidation certain substances might be entirely oxidized as rapidly as they are formed, while in a region of slower oxidation they might accumulate as part of the structure. Bilateral organisms follow a law of antero-posterior development. Differential exposure having determined which region shalllead and having thus set the rate of activity of the successive regions, continuing differential relations of the environment maintain the various levels of the gradient, under the domination of the head-end, at their respective rates of activity. While this is the normal course of development, it may be greatly altered by experimental methods; the original gradient, partic- ularly in lower organisms, may be obliterated and a new one engendered. The latter under certain conditions may even be made to arise at right angles to the original axis. Change of gradient may be readily observed, for example, in the regulatory development of isolated pieces of many planarians. Moreover, the remarkable capacity for self- differentiation possessed by isolated parts taken at dif- ferent levels of the body in such forms, shows that posi- tional relations of the constituents of the regenerating mass, rather than cellular specificity, determine what structures shall arise in a given location. Since many species, including representatives from all the chief phyla of animals, show differential susceptibility at some stage of their development, when subjected to the action of various external agents such as alcohol, anesthetics or potassium cyanide, these agents may be used to bring about shifts of gradient, under-development or over-development of certain parts, or a remodelling of the organism or of various regions of it. Inasmuch as any one of several agents may bring about the same result, it is manifest, again, that quantitative external conditions are the factors which initiate and thereby determine the fundamental orientations and specializa- tions of the parts of an organism. As development progresses in the more complex organ- isms, again through differential stimulation, secondary or * symmetry ’’ gradients may also become established. `. For example, each limb region of a vertebrate becomes a subordinate system with its own internal correlations. No. 649] SEROLOGICAL REACTIONS 85 In organisms with radial symmetry, special centers of growth occur, and only at a certain distance from a given center ean another arise. After differences in protoplasmic constitution have arisen at different levels of the gradient, a system of chemical or transportative correlation probably begins to operate, and out of it all finally comes the various sup- portive and mechanieal tissues, vascular tissues, tissues of excretion and secretion, nervous tissues, etc., which constitute the underlying mechanisms of correlation and integration in the finished organism. The hypothesis says little of chromosomes, or of genes, for its objective differs somewhat from that of the genet- . jeist, but it in no wise denies the existence of such entities. Child argues, however, that since each cell of the body is a descendant of the original zygote and therefore presumably possesses the full complement of chromo- somes of that zygote, something other than the nuclear pattern per se must be responsible for the fact that cells become different in different parts of the body. And this something, he would say, is the metabolic gradient initiated by differential excitation, since the very estab- lishment of such a gradient means the concomitant es- tablishment of local differences along its path. As he says, ** if all the cells are originally alike they cannot of themselves become different.’? The specifice character of the differentiation, the kind of organ or organism pro- duced, he reiterates, is determined ‘‘ by the specific in- herited constitution of the protoplasm.’’ So much for Child's hypothesis of axial or metabolic gradients, in its bearings on embryogeny and differentia- tion. Ihave reviewed it at some length because it shows more clearly than any other theory of development with which I am aequainted that there is no necessity for believing that as cells become specialized they lose part of their original constituents. Due to local conditions, strue- tural modifications and special activities have appeared, but these are changes rung on what is fundamentally the same type of protoplasm in every cell, In higher organ- isms, in some cells possibly irreversible changes have oceurred—the cell may be incapable of dedifferentiation —but nothing constitutional, call it gene or what you will, 86 THE AMERICAN NATURALIST [Vor. LVI has necessarily been lost from the cell. The inheritance complex of the germ is like goods in the piece; it is only as development progresses that the garment becomes specified ; but above all, be it remembered that the finished garment is of the same fundamental constitution as the goods in the piece. It is a commonplace of experimental embryology and experimental morphology, in faet, that the same initial materials may yield very different end-products in dif-: ferent environments. The phenomena of heteromor- phosis, metaplasia, regeneration and regulation all attest this. Blastomeres originally directed toward becoming one part of an organism may be switched about to become another part; tissues originally subserving one function may be turned to other uses; ectodermal cells which by no possible chance could have been predestined to form erystalline lens will, nevertheless, form a lens similar to that of the normal eye if stimulated by a transplanted optic cup. Or, if the lens is removed from the eye of a salamander, a new lens may develop from the edge of the old iris, a part from which the lens never normally develops in the embryo. In short, it is a well-established fact in many organisms that cells occupied with the specializations of one part of the individual still retain the potentialities which would fit them to the functions of some different part, and may, in fact, under exper- imental conditions be made to redifferentiate into the structures of another part. Such facts, together with the exactitude of chromosome distribution in mitosis, indieate clearly that many, possibly all, eells of an or- ganism retain the hereditary tendencies that existed in the original zygote. Because of limitations due to its location in the organism, however, a given cell realizes only a small proportion of its inherent possibilities. And after all, this is no more remarkable than the fact that the genes of recessive characters may slumber indefinitely in germ and soma, generation after generation, until conditions suitable for their expression as characters occur. But now, regarding heredity as in its simplest expres- sion merely the passing on of metabolic activities already established, and conceding that the distinctive structural No. 642] 3 SEROLOGICAL REACTIONS 87 effects and functions which characterize the respective tissues are probably the outcome of unequal activities among the same kinds of fundamental protoplasmic con- stituents in differing local environments, the question of prime importanee to the student of evolution is how the properties of these constituents have come to be changed from what they were initially, how they may be altered in the future—in short, the question of the nature and origin of variations. For whatever we may believe about the degree of preformation which exists to-day in the mechanism of heredity, it is absurd to assume that in the simpler primitive protoplasm from whieh modern forms have evolved there could have been genes of the char- acteristies of all the organisms now in existence. What- ever individual development may be, we must assume that racial evolution was epigenetie. While doubtless in a sense man lived potentially in some primitive protozoan- like creature, actual material antecedents of his existing attributes were no more present in this ancestral ereature than specific determiners for the oceans, continents and topographieal features of the world to-day were present in the original nebula which preceded our solar system. The great central problem of evolution is just this very one of how the determinative accumulations which exist in germ-cells to-day have been incorporated step by step into this erstwhile primitive protoplasm. Certain pos- sibilities have. become realities and concomitantly as a basis of this reality the old mechanism has in part been altered, or a new mechanism has come into being which persists as a part of the established constitution of the germ-cell. Before entering upon a diseussion of whether or not any of the remarkable serological activities which have come to light in recent years may be possible or probable sources of germinal modifications, we must recall briefly the general nature of immunologic reactions. As you know, foreign proteins of either plant or animal origin when injected directly or indirectly into the circulation ‘of an animal will engender antagonistic or neutralizing substances to which the general name of antibodies is applied. Thus the toxins of bacteria incite the production of antitoxins; the bacteria themselves lead to the pro- 88 THE AMERICAN NATURALIST [Vor. LVI duction of bacterial immobilizers or solvents termed bacteriolysins, or sometimes to agglutinating substances termed agglutinins which clump baeteria of the species used in their production, if the two are brought together in the blood-serum of the animal into which the bacteria were originally introduced. Likewise a tissue of one kind of animal injected into the circulation of another induces the formation of antibodies of various kinds such as precipitins which form a precipitate when the blood-serum of the treated animal and an extract of the special tissue used are brought together in vitro; or other antibodies termed cytotoxins or cytolysins which possess a specific toxic or solvent action for the kind of protein used in their production. The alien substance employed to pro- duce antibodies is commonly called the antigen. | In this connection, the phenomenon of anaphylaxis should perhaps also be mentioned. Anaphylaxis is a name given by Richet to designate a highly supersensi- tive state which, after a period of incubation, an animal develops toward certain protein substances that were practically harmless on first injection. Sometimes, par- ticularly in guinea pigs, death results. The sensitizing dose for the production of anaphylaxis may be very small; one millionth eubie centimeter of horse-serum, for example, has been known to render guinea pigs sensitive. The reaction is specific; for instance, an animal sensitized to sheep-serum, though reacting violently to this antigen, displays little or no hypersusceptibility to other sera. In the main all of the immunological reactions show a considerable degree of specificity; the antibody will react fully only with the particular kind of protein used as antigen. The specificity is not absolute, however; a .milder reaction may be obtained with homologous proteins of related species, the extent of the reaction being determined by the nearness of relationship of the species to that from which the original antigen was ob- tained. Similarly with bacteria, the reaction is in the main specific, although so-called group-reactions may ap- pear. The serum of an animal immunized against ty- phoid, for example, may not only agglutinate Bacillus typhosus but may also show this reaction in a less degree with such related forms as the colon bacillus. Thus, ir- No. 642] SEROLOGICAL REACTIONS 89 respective of whether the antigen consists of bacteria or of other protein materials, there is a gradational spe- cificity of reactions which apparently corresponds to tax- onomie relationships. An even more delicate biochemical measure of kinship than the known immunological reactions has apparently been established through the extensive researches of Leo Loeb‘ and his associates on transplanted tissues. In numerous of his papers Loeb has called attention to the remarkable power of transplanted tissues to indicate dif- ferent degrees of even close individual relationship, sueh as the individual to itself, to a brother or sister, to a parent, to a more distantly related individual of the same species, or to an individual of a different species. Par- ticularly the lymphocytes of the host serve as a delicate indicator of such relationships. Loeb assumes that a specific chemical group which he designates as individuality-differential is common to all the tissues of an individual and that in virtue of this characteristic each creature differs from the others of the same species. The individuality-differential of a trans- plant (except in autotransplantation), since it is not adapted to its new environment, assumes injurious prop- erties, probably by engendering toxins. The relative - strengths of these are determined by the degree of rela- tionship that exists between the source of the transplant and the host. In the circulating fluids of a given indi- vidual, he believes that there are ‘‘ autosubstances "' which exercise important regulative functions, such, for example, as keeping the vascular supply of the various tissues at an optimum, or holding in check lymphocytes and invasive fibroblasts which when inadequately re- strained, as in old age, become destructive agents. In one place he speaks of their stimulating effects and he regards them as responsible directly or indirectly for the marked vascular reaction called forth by autotransplants. In sexual reproduction, obviously two different ** indi- viduality-differentials ’’ must combine to form the new individuality-differentials of the offspring. These, Loeb finds, are of varying degrees of intermediacy. This in- termediacy is continued into the next generation. What 4 Amer. Nar., LIV, 1920. 90 THE AMERICAN NATURALIST [Vor. LVI is of much interest from the standpoint of our quest of a possible connection between reaction-products of the body and alterations of the germ is the fact that he feels con- strained to link up his serumal phenomena with the ehro- mosomes. Thus, he says, ‘‘ The chemical individuality- character of the chromosomes should lead to analogous chemical differences consisting perhaps in the formation of chemical side-chains attached to proteins; they should be present primarily in cell-proteins and secondarily in the proteins of the body-fluids. . . . These side-chains must be identical in all the proteins of the same indi- vidual and differ in the case of different individuals.’’ Another great group of influences which extend to the furthermost reaches of the body and profoundly affect the entire organism in development and in maturity— those emanating from the various endocrine struetures— I have barely time to mention. They must be kept in mind, however, when we attempt to pieture the ebb and flow of chemical influence which is indispensable to the maintenance of general physiological equilibrium, inclu- ding that of the gonads no less than of the other body structures. ; You may feel that in reviewing the nature of the protein molecule, the behavior of the proteins of the cells in morphogenesis, the gradational specificities of the im- munological reactions, the relationships which exist be- tween host and transplant, and in reminding you of the intricate functions of the endocrines, I have wandered far afield into irrelevant byways, but I hasten to assert that these phenomena are not as unrelated as might ap- pear at first sight; they are but different aspects of the great salient fact of organismie unity, whether it be a matter of chemical constitution, taxonomic relationship or physiologic response. And now I wish to raise the question of whether or not in the light of the foregoing facts it is irrational to believe that in all probability a thread of chemical iden- tity persists between the chemical constituents of the germ and the chemical substratum of the tissue-cells. The nuclei of the various tissue-cells differ little in ap- pearance from the nuclei of the germ-cells, and inasmuch as the new germinal and somatic cells descend alike No. 642] SEROLOGICAL REACTIONS 91 directly from a common source, presumably bearing in their chromosomes samples of all the chromosomal com- ponents of the original zygote, is it unreasonable to sup- pose that if changes come to pass which can affect certain constituents of tissue-cells, this influence, if borne in the circulating fluids of the body, could also affect the ho- mologous constituents of the germ-cells? Personally, I think that such a hypothesis is not unr ble. But is there even the least bit of evidence on this point? I believe that there is. I feel that in the transmission of eye-defects secured by Dr. E. A. Smith and myself in fetal rabbits by means of serum immunized against rabbit crystalline lens, we have a bona fide case of such parallel influences. Since I have already presented the facts before this Society and inasmuch as the details are avail- able in printed form,’ I need not repeat them now. It is sufficient to recall to you that we secured a fowl-serum immunized against rabbit crystalline lens which when injeeted into pregnant rabbits penetrated the placenta and occasionally attacked the lens of the fetal young, the outeome being marked eye-anomalies in such young. Since, once produced, the defects were transmitted to successive generations through both male and female lines, we interpreted our results to mean that the immune serum was not only specifically cytolytic for the newly forming lens-tissue of the fetus, but that it also attacked the representatives of such tissue—its genes, if you please — in at least some of the germ-cells of the fetus. If true, this must mean that there is some degree of constitutional identity, probably protein homology, between the mature substance of a tissue and its correlative in the germ. And in view of the fact that, basically, inheritance is mainly a question of the perpetuation of specific protein-com- plexes, and development, the result of differential reac- tions of these same fundamental constituents under dif- fering conditions of environment, is this an unreasonable inference? : But does anything comparable to this occur in the ordinarv course of animal existence? Do cytolysins or kindred substances which can modify or destroy both 5 Jour. Exp. Zool., 31, 2, 1920, AMER. Nar., LV, 1921. Proc. Nat. Acad. Sci., 6, 3, 1920. 92 THE AMERICAN NATURALIST [Vor.LVI tissue-elements and their germinal correlatives ever occur in animals without being introduced by man? Do animals ever form such antibodies or other equally active sub- stances against their own tissues? It is obvious, since tissues persist intact under conditions of normal physio- logical equilibrium, that they are not being subjected to: such influences, or if they are, that they resist them. As a matter of fact, Römer,’ using the complement-fixation technique, found that the serum of adult human beings may possess antibodies for their own lens proteins. It seems reasonable to suppose that if the tissues of an animal became injured or displaced in some way, or met- abolically unbalanced, immunity reactions might be es- tablished against them. We have some evidence that such is the ease. During the late war, for example, it was found that toxic reactions resembling anaphylactic shock often followed extensive injuries of the soft tissues. The matter can be tested experimentally. Because of their distinctive nature and the ease with which they may be isolated, I chose spermatozoa for such an experiment.’ I found that a rabbit will build antibodies against its own spermatozoa when these are injected into its blood- stream; also, that rabbits injected with rabbit sperma- tozoa not only develop antibodies in their blood, but also have their own spermatozoa greatly weakened, a condi- tion shown in vitro by their lessened resistance to anti- sera. This clearly shows that an animal can on occasion build antibodies against its own tissues; and since anti- bodies can apparently directly or indirectly affect germ- cells, it seems reasonable to suppose that such influences, especially if continued over a long period of time, might be one source of germinal variations. It is known from the experiments of Kuntz? and others that the blocking off of the ductus deferens of one testis may induce degeneration of the germinal epithelium, not only of that testis, but of the other as well. Inasmuch as the spermatozoa in the testis on the operated side must die and be resorbed, is it not probable that in this process spermatotoxins have been formed which have then attacked the living germ-cells of the other testis? Again, we are familiar with the fact that oculists fre- 6 Zinsser: ‘* Infection and Resistance.’’ 7 Paper in press, Jour, Exp. Zool. 8 Anat. Rec., 17, 4, 1919. No. 642] SEROLOGICAL REACTIONS 93 quently find it necessary to remove a severely injured eye to prevent the ‘‘ sympathetic” degeneration of the other eye. I am told by competent oculists that the ex- tension of the degenerative influence involves more than the atrophie effects which might result through direct nerve connections. Does it not seem probable that here, too, the disintegrative influence which comes to operate on the uninjured eye is cytotoxic or cytolytic in nature? And if it can operate on the tissues of the normal eye, "why not on the corresponding protein constituents in the germ, the prototypes of those which were originally in- eited to form the ocular tissues? It may be, it probably is true that there is sufficient dif- ference between these factors of the germinal protoplasm and those of the finished organ to render the former less susceptible to such agents. It is not improbable that even if some of the numerous germ-cells were affected many others might not be. But any new organism which sprang from such an affected germ would have its own germ-cells similarly modified, since these would all be derived from the same zygote. Even so, the defects might not be manifested in offspring because of the prob- ability of dominance by the corresponding factors from their partner in fertilization. | The only way to settle the matter, of course, is through experiment. I know of no existing experimental evidence - on this point. In my own laboratory, however, an in- vestigator has an experiment in progress which I hope will ultimately throw some light on the matter. There are many bits of evidence to show that an or- ganism may react against the tissues of other individuals . of its own species. Thus Bradley and Sansum,’ employ- ing anaphylactic reactions, found that guinea pigs in- jected with various guinea-pig tissues such as heart, liver, muscle, testicle, and kidney developed immunity reactions. Moreover, certain changes in the blood of the mother during pregnancy, apparently induced by cells or cell-products set free from the newly-forming placenta, seem to be of the nature of antibody formation. Then again Turck'? has shown that products of the lung-tissue of the eat, autolysed under sterile conditions in vitro, pro- 9 Jour, Biol. Chem., Vol. 18, 1914. 10 Med. Rec., 1919, 95, pp. 719-21. 94 THE AMERICAN NATURALIST [Vor.LVI duced characteristic pulmonary lesions when injected into other eats. Similarly autolysed lung-tissue of other mammals had no effect on cats. But, the question arises, in order to get parallel influ- ences, in soma and germ, would there not have to be ab- solute identity between the two sets of proteins con- cerned? Before answering this question let us glance for a minute at two types of specificity which are recog- nized in serological reactions: namely, ‘‘ species-specifie- ity" and ''organ-specifieity." What the serologist means by species-specificity is the fact, shown through precipitin reactions, that blood immunized against one tissue of an alien species will react, although in a less degree, with extracts of the other tissues of that species. And that there may be a specificity of certain organ com- plexes which is independent of species is shown by the fact that an immune serum produced by using the crys- talline lens of one species of animal yields a precipitin which reacts more or less with the lens proteins of even unrelated species. Similar results have been obtained with proteins derived from the testis, and confirmatory evidence of such organ-specificity has also been estab- lished by means of anaphylactic reactions. Such facts as these, together with those cited in the discussion of the gradational reactions of various immune sera according to the systematic relationships of animals, it seems to me, answer our question affirmatively; there need not be ab- solute identity between the proteins of the somatic cells and their correlatives in the germ-cells for immune sera engendered against the one to react also against the other. I raise the issue because it might be urged that such tissues of an organism as become so abnormal as to excite the production of antibodies are no longer suffi- ciently similar to the normal tissue-elements, and there- fore to their germinal representatives, to make the anti- bodies effective against either normal somatic or ger- minal constituents. It seems to me that all available facts indicate that the constitution of an organism, whether germ or soma, is not to be regarded as a congeries of cooperating, equi- potent units, but rather as the outcome of interacting systems which differ in their orders of organization; sys- tems which in themselves possess more fundamental and No. 642] SEROLOGICAL REACTIONS 95 more supplementary or fluctuating components ; chemical groups which represent the more constant features of organization coupled with subsidiary groups of more re- stricted significance. That is, there seem to be series of substances of like chemical constitution common to all the cells of an organism, possibly to even various groups of organisms, and superimposed upon these central or foundational constituents, probably as parts of the same molecules, are secondary systems, or possibly systems within systems, which modify the main configurations in various ways. This conception certainly squares with the fact that degrees of specificity paralleling the kinships of animals may be shown by immune sera. It harmonizes with what we know of the architecture of the native proteins as well as with our whole scheme of natural biological taxonomy in which we find certain fundamental stable features representing a broad series of organisms, and less and less inclusive characteristics which grade down to the minor differences that separate species, varieties and in- dividuals. Nor is it incompatible with what we know of chromosomes and genes. The very fact that heritable grades of a single gene in a given chromosome may occur (e.g., in Drosophila) and that one of these variants may in turn be modified gradationally by a series of secondary factors located in other chromosomes suggests the type of organization just discussed. With the remarkable and abundant evidence of hand- and-glove relationships between unit-characters and chromosomes that has been accumulated in recent years through the painstaking studies of workers on Droso- phila, not to mention other. corroborative work, it seems to me that there is no longer a reasonable doubt that the differentials, whatever they may be, responsible for the distinctiveness of the so-called unit-characters, reside in the chromosomes. And while I have always believed” and still believe that for the final outcome the cytoplasm 1s just as necessary in its way, and must be just as char- acteristic of the species as the chromosomes are, its dis- tinetiveness must be of a fundamental organismie (prob- ably chemical) type common to the species as a whole, 11 Bull. No. 2, Univ. of Cincinnati, Vol. III, Ser. II, 1902. Science, June 28, 1907. fab Cincinnati Studies, Sept—Oct., 1909. AMER. NAT., XLV, 1911. 96 THE AMERICAN NATURALIST [Vor. LVI since it ean be, in fact is, contributed in reproduction al- most wholly by one parent, the female. It is apparently a medium which responds specifically to the action of the respective chromosomal incitants, whether these be of maternal or paternal origin. All geneticists agree to-day; I think, that any character of an adult can not be merely the outeome of a unitary germinal antecedent; it is the produet of many faetors. And ordinarily what we see as a character-difference is. probably merely the outcome of a factor-difference in one of the chromosomal cooperants. In conclusion, let me say first of all that no one more than myself realizes the inadequaey of my present argu- ment as a complete or satisfying theory. The knowledge in the fields on which it is based is as yet far too frag- mentary to warrant anything but tentative conclusions. But since various facts seem to me to point toward the view that certain types of immunological reactions, no- tably the cytolytic, engendered against various somatic constituents may occasionally also affect chemically re- lated substances in the germ, and inasmuch as many other facts lend themselves to such an interpretation without undue violence to scientific credulity, I have felt justified in presenting the whole matter in the form of a working hypothesis. In the short time remaining I can not enter into the important question of whether or not changes induced in the blood-serum might be instrumental in leading to pro- gressive rather than regressive evolution, and even had I time for such a discussion, there are not sufficient data available to support such a discussion affirmatively. I should like merely to point out in closing that through exercise we can initiate and promote growth in various parts of the soma, we can induce hypertrophy, and in so doing we are in some way leading the protein and other constituents of the cells in question to make more of their own kind of substance, in other words, to reproduce their kind. We do not know what stimulates them to do so, but, in part, it may well be something that is or can be transported in the circulating fluids of the body; and if so, then there exists the possibility that the correspond- ing germinal representative of such a part, however tenuous the thread of chemieal connection, might also be modified in the direction of progressive germinal change. THE AMERICAN NATURALIST Vor. LVI. March-April No. 643 ORTHOGENESIS SYMPOSIUM ON ORTHOGENESIS BEFORE THE AMERICAN SOCIETY OF ZOOLOGISTS, TORONTO, DECEMBER, 1921 ORTHOGENESIS FROM THE STANDPOINT OF THE BIOCHEMIST PROFESSOR L. J. HENDERSON HARVARD UNIVERSITY Ir does not seem likely that physical science should have much to say about the theory of orthogenesis. In the first place, it is hard to see what the term means if one adopts a physico-chemical standpoint. In the second place, organic evolution is more remarkable in its morpho- logical aspects than in its chemical and physico-chemical aspects. I The first point,may be dismissed with a few remarks. Orthogenesis presumably means that evolution has taken place in a straight line or in a very restricted path, and that the straightness of the line depends, at least partly, upon something which is internal to the organism, though, of course, the process may be released by a stim- ulus from the environment. The straightness of the proc- ess must be largely a matter of definition. Physico- ehemieally, it eould hardly mean more than that quanti- tative ehanges have steadily the same sign over a con- siderable period of time. One might, perhaps, adopt such a view, if one could believe, as has been often suggested, that variation is the expression of a process which is approaching a condition 97 98 THE AMERICAN NATURALIST [Vorn LVI of equilibrium, because then, so far as there is no unto- ` ward interference from without, it would be natural to think that the eourse of the process must be in a certain sense a straight one, with a negative acceleration. Taken literally, such a consideration is, however, purely specu- lative and for the present, I think, a sterile speculation. Somewhat more clearly intelligible is a hypothesis whieh arises from the study of hormones and their róle in development. It appears to be quite possible that the effect of increase or decrease in the amount of a single chemical substance in a species might be a complex change in its strueture, including modifieations of size, . of the proportions of the different parts, of pigmenta- tion, or of the other peculiarities which ordinarily arrest the attention of students of evolution. This would be. especially true if, instead of a change in the amount of a hormone or other substance, it were a case of the forma- tion of a new compound. Such changes, while directly due to a single substance, might be greatly modified by readjustments following the disturbances of the physio- logical equilibria between the different parts of the body. Compensatory readjustments of similar nature are, of course, among the most familiar and interesting phenom- ena in pathology. We are, accordingly, fully justified in taking their possibility for granted. It is, therefore, conceivable that evolutionary changes should be occasionally progressive and apparently ortho- genetic, although due to a simple physico-chemical modi- fication. No doubt, if it were desirable, such considera- tions might be developed into a clear and possibly useful theory of orthogenesis, but I am not qualified to do so. My object is only to insist that changes which from a morphological standpoint are complex, continuous, and progressive, may conceivably be due to a single, simple, physico-chemical change. Such reflections, vague though they may be, clearly point to a conclusion which is, I feel sure, inevitable for the physical scientist; morphological phenomena in them- No. 643] ORTHOGENESIS 99 selves are not suffieient to establish the validity of any theory of the mechanism of evolutionary variation. II More important than speeulating about such questions is the fact that the underlying physico-chemical processes in living organisms seem to have remained about the same throughout the whole process of evolution. So far as it is possible to form any opinion on the matter, this conclusion is inevitable. In considering the question of organic evolution it should always be remembered that, with very trivial ex- ceptions, the economy of life on the earth is now and prob- ably always has been founded upon the synthesis of car- bohydrates from water and carbonic acid with the ac- companying fixation of energy, followed by the conver- sion of carbohydrates into fats, proteins, and a great variety of other related substances. Later there is an oxidation of these substances back to water and carbon dioxide, accompanied by the utilization of the energy in various forms of organic activity. Correlated with this is the fact that cells are made up of water, carbonic acid, carbohydrates, fats, proteins, and certain other sub- stances. They are enough alike in chemical composition and in physico-chemical structure fully to justify the concept of protoplasm as a fairly constant physico-chem- ical apparatus throughout the organie world. These familiar facts of chemical physiology and chem- ieal morphology undoubtedly depend upon the properties of the substances involved. Water and carbonic acid, with which the process begins and ends, and which seem to be everywhere the foundation of protoplasm, possess in themselves such a large number of remarkable char- acteristics and lead directly through the formation of sugars to such a great variety of chemical substances and chemical reactions, that it is hard to believe in the pos- sibility of the existence on a large scale of any very dif- ferent kinds of living organisms. 100 THE AMERICAN NATURALIST [Vor. LVI This is a subject that I have elsewhere discussed at length. Hence it will perhaps suffice briefly to recapitu- late a few of the more striking facts. Because of its peculiarities as a solvent, as an ionizing medium, etc., water makes possible the formation of an almost indef- initely greater variety of physico-chemical systems than does any other substance. On account of its high latent heat of vaporization, its high specific heat, its high sur- face tension, and the peculiarities of carbonic acid, such systems often possess a very remarkable stability. The elements hydrogen, carbon, and oxygen, of which water and carbon dioxide are composed, seem to be unique in ' the number and variety of the substances which they can form. In particular, the production of sugar from water and earbon dioxide fixes a very great amount of energy and leads directly to the greatest variety of chemical sub- stances and reactions which are known to occur as the result of one chemical process. Finally, water and ear- bon dioxide are the two substances which are everywhere available. Anything so complex, so stable and yet so variable, so widespread and so active as life can only occur when a great variety of conditions are fulfilled. In other words, the physico-chemical systems of the organism, in order that life shall be capable of its evolution, must possess altogether exceptional characteristics, which appear to be quite impossible unless water and carbonic acid, and compounds of the three elements, hydrogen, carbon, and oxygen, and no doubt also of nitrogen, are at the basis of them. These substances possess a set of properties each one of which by itself and also in cooperation with the others is necessary for the highest physico-chemical com- plexity and variability. So far as we know, no other ele- ments or compounds possess another set of properties which permit similar physico-chemical complexity or variability. It is, I believe, for this reason that life has always op- erated on the same basis. No. 643] ORTHOGENESIS 101 Thus while the evolutionary process has certainly pro- duced a large number of well-defined series of changes when it is looked at from the morphologieal point of view, it still remains very probable that such physico-chemical changes as have occurred are not only of a secondary nature, but that they are much less of the character of serial modifieations. Indeed, one is tempted to say that in a physico-chemical sense, the variations are distributed in rather a random manner, without any particular indi- cation of a general progressive tendency, such as we seem obliged to think of in studying morphologieal variation. No doubt the evolutionary proeess has, from time to time, invented new chemieal substanees and greatly modi- fied colloidal systems. In the total these changes are very numerous and of the utmost importanee to the stu- dent of evolution. But progressive change is more par- tieularly a morphologieal phenomenon and it seems to be almost self-evident that progressive morphological evolu- tion should not be accompanied by the same degree of continuous variation in straight lines in physico-chemical properties. Such a parallelism would, I think, be well nigh unaccountable. However that may be, there is no evidence for it. III Another consideration which makes the theory of or- . thogenesis seem very different to a physieal chemist from what it must seem to a biologist, is the fact that chemistry tends to deal with individual substances which either exist or do not exist. The case of hemoglobin will illus- trate this point. Hemoglobin is an individual substance of very marked peculiarities. So far as known there are no essential differences between the hemoglobins con- tained in the bloods of different species. It is possible that the known differences in crystal form depend upon something more than trivial differences in the amino acids which make up the molecule, but this seems unlikely. In any case, it will do no harm to speak of hemoglobin as 102 THE AMERICAN NATURALIST [Vor.. LVI a single chemieal individual in order to illustrate a par- tieular point. This substanee is the sole means of transporting more than a small amount of dissolved oxygen in the blood of those species which contain it. It is, therefore, apparent that it may be thought about from the evolutionary point of view, much as one thinks about an organ. I believe that the success of Aristotle's system of classification justifies this view. But while it is easy to think of the gradual evolution of an organ as something which ean not be regarded as appearing at any point in the evolu- tionary process, being related by a process of continuous differentiation to something which was certainly not the same organ in an ancestral species, there is not the slightest evidence for anything of the sort in the case of hemoglobin, and it must seem to most chemists nothing less than fantastie to assume such a continuous evolution of a substance more and more closely approaching hemo- globin. Moreover, it is almost as difficult to imagine such a thing from the standpoint of a biologist, and it is cer- tainly true that any given organism either does or does not eontain a substance which is capable of de. a loose chemical combination with oxygen. But the difficulty in the case of hemoglobin is more serious than this, for it has been found that hemoglobin, like other organs, has more than one function. It has, in fact, at least three; for it is the sole means of transporting oxygen, almost the sole means of liberating carbonic acid in the lung and absorbing it in the tissues, and the instru- ment of the final delicate adjustment of the alkalinity of the blood. The last two functions depend upon the same property in the hemoglobin molecule, but this property is a different one from that which enables hemoglobin to combine with oxygen. We are, therefore, here confronted with the task of imagining the origin of a chemical sub- stance, quite different in its nature from any other known substance, which possesses two chemical peculiarities, and which, as a result of these two peculiarities, performs three highly important functions. No. 643] ORTHOGENESIS 103 Now it may be that originally hemoglobin possessed only one of these peculiarities, so that its sole original function was to carry oxygen. And accordingly one of the most interesting questions of comparative physio- logical chemistry concerns the respiratory function of the blood. It would be a very important discovery to find a kind of hemoglobin in which there is no specialized action upon the transport of carbonic acid and upon the alkalin- ity of the blood. But even if the earliest hemoglobin were of such a nature, the first production of hemoglobin would still seem to have been relatively an extremely dis- continuous variation involving an unmistakable physio- logical unit of great importance. It is true, and should be noted in order to avoid con- fusion, that there has been a later evolution of the proc- ess of oxygen transport. This has been commented on by Bareroft as a result of his own important researches. It appears that variation in the electrolytes of the red cells is aecompanied by remarkable variation in the af- finity of their hemoglobin for oxygen, and that this is the explanation of the differences in the so-called oxygen dis- sociation curves of the bloods of different species of mam- mals. Here, as Bareroft points out, there is no difficulty in imagining a process of adaptation, for the fact of chemical discontinuity is not involved. It is a question of changing proportions of the different substances. But in spite of the possibility of sueh phenomena, it seems probable that there are, even in the human species alone, a considerable number of important individual sub- stances whose appearance in the course of organic evolu- tion it is very difficult to imagine, except as a radical in- novation. Accordingly, it must be apparent that in the present state of our knowledge, any theory which postulates con- tinuity in evolution is very unsatisfactory to the chemist. Moreover, in this case one seems to be confronted with an appearnace of discontinuity which does not depend, as is too often the case, upon a judgment of the magni- tude of a difference. 104 THE AMERICAN NATURALIST [Vor. LVI Of course, it is not difficult to imagine a sufficiently close approach to continuity of evolution, and therefore, to orthogenesis, in the case of simple proteins. But here, very likely on account of our ignorance, there is no indication of anything more than indefinite variation and variability, accompanying variation in a definite direction in the morphological characteristics of species. On the whole, variation in the ultimate physico-chem- ical nature of organisms seems to have been rather dis- crete than continuous, not orthogenetie, but distributed at random. Such a conclusion may possibly be illusory, for our ignorance is greater than our knowledge. But whatever the nature of the changes which it has under- gone, the most striking thing about the physico-chemical nature of protoplasm seems to be its uniformity through- out nature. Therefore, with due reservations because of the incom- pleteness of bio-chemical knowledge, it seems reasonable to suppose that apparent instances of orthogenesis may sometimes depend upon a single important chemical change in an organism, followed by slow and progressive modifications leading up to a definitive morphological re- sult. Such a process would be somewhat analogous to the establishment of a condition of equilibrium. ORTHOGENESIS IN BACTERIA PROFESSOR CHAS. B. LIPMAN UNIVERSITY OF CALIFORNIA Ir is well to understand at the outset that bacteria, un- like plants and animals, can not be studied from the orthogenetie standpoint in the strict sense, owing to the lack of a proper vantage point, or, perhaps more correctly speaking, a basis from which one may start such a study. In my opinion, all attemps at the establishment of sys- tems of bacteria, and there have been many, have ended in creating greater confusion than there was at the start. Such frustration of well-intentioned programs was inevi- table for at least three reasons, to wit: (1) Every bac- teriologist used criteria of his own for the establishment of new species. This is bound to lead to chaos. (2) Nearly all bacteriologists used a morphological basis for classification. Owing to the relative simplicity of form of bacteria, this must inevitably result in erroneous dis- criminations. (3) Nearly all bacteriologists in the past and even to-day are laboring under the misconception that bacteria are simpler forms of living organisms than they really are. That this is incorrect has been shown by the studies of Lohnis and Smith in 1916* and of Lóhnis alone very recently.’ But assuming the foregoing to be true, it must follow that it is impossible to trace the evolution of new species . of bacteria in a definitely directed course. Not being cer- tain what constitutes a forward or what a backward step 1 (a) ‘* Life Cycles of the Bacteria,’’ preliminary communication, Jour. Agric. Res., Vol. 6, No. 18. (b) ** Life Cycles of the Bacteria,’’ paper read at New Haven meeting of Soe. Amer. Baeteriologists, Dec. 27, 1916. ? ** Studies upon the Life Cycles of the Bacteria,’’ Part I, Review of the Literature, 1838-1918, Nat. Acad. Sciences, Vol. XVI, Second Memoir. 105 106 THE AMERICAN NATURALIST [Vor. LVI in bacterial development, it is obvious that we can not apply as justifiably as we do in the case of the higher organisms the criteria of definitely directed evolution. To be sure, we have a number of instances of the estab- lishment of permanent new characters in bacteria, yeasts, and other fungi through the influence of a change in en- vironment. Most of these are, however, induced through changes in the medium under artificial conditions and they do not necessarily indicate a change in the direction of improvement of the organism or of greater complexity in its organization which may in turn point to the evolu- tion of a higher form from a lower one. Despite the foregoing, it is probably well to examine into certain facts with which we are familiar with regard to microorganisms, and which may, perhaps, have a close bearing upon what might be regarded as orthogenesis in bacteria. The first fact to which I wish to refer is that . of parasitism. There can probably be very little doubt that parasitism on the part of bacterial cells is not an original, but an acquired character, using the term ‘‘ac- quired’’ in its literal, and not technical, sense. If that is granted, it would also follow that the aequirement of such a characteristic by a microorganism would mean the gradual adaptation of a bacterial cell from one kind of a medium to another. It would mean the gradual acquire- ment of partiality on the part of a microorganism towards certain chemical substances, certain tempera- tures, or certain other conditions which obtain only in a living host and not in an inanimate medium. The steps, gradual or rapid, by which the aequirement of such pecul- iar characteristics on the part of the microorganism would occur in its change from a saprophyte to a para- site would almost seem to imply evolution in a definite direction. In a sense, therefore, we may regard para- sitism in bacteria as an evidence of orthogenetie develop- ment in such organism. It is, moreover, a case of evolu- tion in a definite direction through the influences of en- vironmental factors of the natural order and not those No. 643] ORTHOGENESIS IN BACTERIA 107 which are produced artificially. In respect, therefore, of causal factors in the evolution of bacteria, we have parasitism exemplifying the antithesis, so to speak, of changes which we induce in bacteria in our artificial media, or by changes in the environment. These observa- tions would seem to possess cogency, not only in the ease of obligate parasitism, such as that characterizing the organism of human tuberculosis, or of anthrax, or of the fungus of wheat rust, but also that of what we may call facultative parasitism in which the organism may have adapted itself to life, both as a saprophyte and as a parasite through the influence of certain chemical or physi- cal-chemical agencies in its environment which have ren- dered its protoplasm more highly adaptable than that of the obligate parasite. We may, therefore, regard facul- tative parasitism as an instance of orthogenetie evolu- tion, just as we may so regard obligate parasitism. The puzzling question whieh may, however, arise, from these considerations is, which is the more advanced step in orthogenesis in parasitic bacteria. Is the obligate para- site the more advanced form, or is the facultative parasite the more advanced form? While many would probably, on first impulse, regard the former as the correct answer, it does not necessarily follow that such is the case. Cer- tainly in this regard, a great many more facts are needed before any definite statements can be made. Examples of other cases of orthogenetie evolution in bacteria other than the case of parasitism, which I have just discussed, may be multiplied ad libitum. But, owing to limitations of time and space, it will suffice to mention a few only. The adaptation of bacteria to the physiological char- acteristic of nitrogen fixation, such as is possessed by all the Azotobacter species, and the Clostridium species and, to a slight extent, by many other species, can scarcely have been the result of anything else than a case of def- initely directed evolution. This was probably accom- plished through the influence of an environment in which 108 THE AMERICAN NATURALIST [Vorn LVI it was impossible for the organisms existing therein to live without aequiring a power of employing energy exist- ent in carbonaceous material to fix atmospheric nitrogen and make it available for their own life processes. The next ease which may be cited is that of the lactic-acid bacteria, which possess the power of transforming lactose (or milk sugar) into lactic acid. These cells are not in form or otherwise in function appreciably different from any other bacteria with which we are acquainted. They have, nevertheless, this specific and peculiar power to which I have alluded. Is it likely that they have acquired this power through any other influence than the influence of environment which operated in a definite direction and hence orthogenetically? The sulphur bacteria, or par- ticularly those species of sulphur bacteria which have the power of oxidizing sulphur to sulphuric acid, are another ease in point. The nitrifying bacteria are still another case in point. The iron oxidizing bacteria are still an- other case, and so we might go on and mention very many classes of bacteria, in each case of which there is a def- inite, distinct, and strikingly peculiar functional power which could not well have been developed without the influence of some environmental factor or factors opera- ting in a definite direction. It is not so easy on the mor- phological side to give examples like those which I have just cited from the point of view of function of the bac- teria. The reason for that has already been touched on above, namely, the simplicity of form and particularly the slight variety in form which characterizes the bac- teria. In fact, it is my conviction that it is best to ignore, largely, morphological factors in bacteria when we study the problem of bacterial evolution. My conviction arises from a study of many and varied experiments which I ean not discuss here. Viewing our subject, then, from the standpoint that orthogenesis in bacteria would be concerned with pro- gressive changes in the organism, prineipally physio- logical, due to its response to changes in environment, it No. 643] ORTHOGENESIS IN BACTERIA 109 seems that we must admit that orthogenesis does exist there. But if, on the other hand, progressive changes like those in question must also be in the direction of pro- ducing a more advanced form of organism, we are con- fronted by a quandary resulting from a lack of an ae- cepted definition for the term **advanced.?' The argument that bacteria do not at all lend them- ` selves to appraisal as regards evolution by the standards applying to the higher organisms is, perhaps, not sound as shown again by the researches of Lóhnis, which I have just cited. The objection to viewing bacteria in a man- ner similar to the higher organisms because no sexual reproduction is known among them is removed by Lohnis’ observations, which indicate that something akin to a real conjugation of cells does occur in the bacteria. His stri- king monograph in the Memoirs of the National Academy should be read and studied by all those who seek new light on the origin and nature of bacteria. Another point of view which I believe may be intro- duced into this discussion with some justification results from a broad comparison of natural phenomena gener- ally. In the inanimate world, we are confronted by the evolution of substances in series in which the first mem- ber of the series is simple and by small accretions be- comes progressively more complex in the succeeding members of the series until very complex materials are finally built up. We are all well acquainted with the seriation showing progressive complexity in the hydro- carbons beginning with methane and going up; in the car- bohydrates beginning with formaldehyde and going up; in the proteins beginning perhaps with amino acids and going up. Such examples of progressive seriation may be multiplied ad libitum. Why, therefore, should it not be possible that similar series should arise in the progres- . sive evolution of bacteria through certain forces as yet largely unknown which cause accretions of characters, so to speak, to occur in bacteria through their being ren- dered more complex and complicated by the influence of 110 THE AMERICAN NATURALIST [Vor. LVI certain factors of the environment. It seems inconceiv- able to me that the great diversity and complexity of funetional nature in the baeteria eould have arisen other- wise. Nevertheless, analogies between phenomena in animate and inanimate nature must not be pushed too far in the absence of the necessary facts for their sup- port. While I believe them to be of great significance, I do not desire to be dogmatie on the subject in the slightest degree. While all the foregoing as regards the evidence for orthogenesis may be accepted as true, it does not follow that the doetrine of orthogenesis is anything new or sig- nificant or was so when it was first enunciated. It seems to me to constitute merely one way of describing the aetual eondition of progressive series in evolution, but it seems to me that it explains nothing. In so far, how- ever, as its advocates espouse the cause of those who be- lieve in and give evidence for the inheritance of acquired characteristics, the potency of environment in inducing fundamental and permanent changes in the organism, and the theory of mutation, they do contribute something sig- nificant to the discussions and experiments which consti- tute the amorphous symplasm, metaphorically speaking, from which our knowledge of the well-defined and real nature and origin of life may some day be expected to emerge. It is, perhaps, of particular importance now to con- sider the bacteria as a class and their probable origin as bearing on the question for which we are trying to find an answer. There is a general disposition, and particu- larly is it true that there has been in the past, on the part of biologists and natural philosophers, so called, to place the bacteria in point of origin among the most primitive of living organisms. There is much inclination, indeed, to regard them as the most primitive organisms. While, superficially, this view seems attractive and cor- rect, it loses much of its eogeney when one takes into consideration the following situation: In all but a few No. 643] ORTHOGENESIS IN BACTERIA 111 exceptional forms of bacteria, some of which I have named above, the physiological characteristic is either that of a saprophyte or of a parasite. It seems obvious to me that neither a saprophytic nor a parasitic organism can well be expected to originate in an environment which is devoid of elaborated organic matter. Subject to considerations which I shall diseuss later, we must, therefore, accept one or two conclusions with regard to the origin of bacteria in the seale of evolution of organ- isms generally. Either they are the most primitive forms of organisms which have lost their primitive powers of living in purely inorganic media, or they are a much more advanced form of life which came to be after other organie forms had for some time been developing on the earth's surface. The first possibility is merely tantamount to saying that some cells of the most primi- tive forms have gradually adapted themselves to either a saprophytie or a parasitie existence and, therefore, is of little assistance to us. The correctness of the second con- clusion, however, would seem to depend on many little- known factors. Still, it is the belief of many scholars. Putting the matter in another way for greater clarity and emphasis, I may state it as the opinion of several plant physiologists who have speeulated upon this sub- ject, that the primitive forms of living cells were prob- ably those which could live in a purely inorganic medium. Obviously, such cells must have been limited to the group which we now eall the autotrophie organisms, and of the autotrophic organisms, since the higher plants are cer- tainly a very advanced form, we must have had some- thing very much simpler, and the natural conclusion is that such a simpler form of organism must have been the single-celled green alga, or forms closely similar to it. If we assume that such was the case, then it is not diffieult to propose a scheme of evolution of the baeteria which involves the gradual change of the unicellular green alge into a variety of bacterial forms through the influences of environmental faetors as I have already in- 112 THE AMERICAN NATURALIST [Vor. LVI dicated. It is not at all inconceivable that a green algal . cell may have adapted itself gradually to life within a higher plant cell or within an animal cell, or to a sapro- phytic existence in soil or other media devoid of light. It may first have come there accidentally and then, through the power to respond to such an environment and to tolerate it, has gradually evolved new powers and has lost some of its old powers. It is conceivable, there- fore, that whether we regard parasites and saprophytes among the bacteria as degraded forms or not, they may be examples of evolution in a definite direction, pre- sumably in this case in the direction of greater complex- ity of function resulting from the urge of a constantly and markedly changing and potent environment. Since the foregoing observations on orthogenesis in bacteria have led me to enunciate in another form a theory accounting for the origin of bacterial forms which has been discussed before, I feel constrained to go one step farther into that subject in order that my own views may not be misunderstood. While the idea of accounting for the origin of the bacterial cell from the single-cell alga seems attractive and appears to be in consonance with cer- tain well-known facts, there are several troublesome features about it. In the first place, it assumes the de- velopment of so complicated and intricate a substance as chlorophyll before any form of living substance was evolved. While this may, of course, have been the case, it seems doubtful, in view of what we must consider to be the highly specialized nature of the green pigment of plants. In the second place, we have seen that the strong argument in favor of the theory of the single-celled alga as the primordial cell, or rather against the theory that bacteria may have been such primordial cells, lies in the well-known fact that most bacteria require organic com- pounds as energy for their life processes and that no organic matter could have been available without the activity of chlorophyllous organisms. This argument, however, overlooks two points, viz., first, the existence of No. 643] ORTHOGENESIS IN BACTERIA 113 autotrophic bacteria and, second, the possibility and even probability that suffieient amounts of organie matter for baeterial purposes may have been elaborated at the dawn of life by chemical means, using the term ** chemical "' in the broadest sense. It is, of course, well known that the autotrophie bacteria, for example, the nitrifying bac- teria, can live and build organie matter out of purely inorganie substances, carbon being obtained from earbon dioxide of the air, and in the absence of light and chloro- phyll. But if this is so, why may it not be that of the known forms of living cells, the autotrophie bacteria were the first, since they are capable of living in a purely inorganic medium, the ammonia which is necessary to them being supplied from the small amounts resulting from chemical reactions induced by electrical phenomena in the atmosphere. As we have seen thus far, it may be argued, with equal justice, that the activity of the nitri- fying bacteria is a highly specialized one on the one hand, and a very primitive one on the other. But if, as just indicated, it should be argued that, after all, the autotrophic bacteria are exceptions in the bac- terial world and that most bacteria need elaborated or- ganic matter and hence they could not have been the primordial living cells, the second objection which I have stated may be urged, namely, that organic matter may have existed on the earth before living cells came into being. Mature reflection will render it highly plausible with the high temperatures, great electrical activity, and probable intense radioactivity which existed on the planet prior to the appearance of living cells, that un- usual chemical activity inducing rapid and general com- binations among the elements should have prevailed. This, moreover, involves the assumption of the existence of a degree of all these conditions which is requisite for the synthesis, but not for the rapid destruction of the or- ganic matter, which must also be conceded as probable. Under such conditions, it is reasonable to suppose that bacteria, on being evolved as the primordial cells, may 114 THE AMERICAN NATURALIST [Vor. LVI have found the conditions requisite to their growth and further development. It seems, on careful deliberation, that strong arguments may be brought forward for both the theory that single- celled green alge and the theory that bacteria were the primordial organisms, if we consider merely the argu- ments which enter into the usual diseussions of the sub- ject. But it appears to me that we must penetrate beyond what is ordinarily called careful deliberation, if we would see other possibilities for explaining the origin of living matter. There is no logical reason for confining our attention in these discussions to the single-celled alge and the bacteria which we know. There are, in addition, bacteria so small as to challenge and defy our ingenuity for devising means for rendering them visible. What may not further discoveries about their nature and re- quirements for life unearth for us which may be of the most vital significance to the solution of our problem? I have tried in imagination to go beyond, far beyond, the ultramieroseopie bacteria and have pictured to myself the following condition for the origin of living matter: A single molecule of organic matter, let us say, a polypeptid or a proteid molecule produced by the force which I have discussed, exerted as chemical energy, may, in floating about in its aqueous medium on the earth’s surface, sud- denly find itself in a field of radioactive ‘force or some similar force which causes its atoms to orient themselves in such fashion and to vibrate in such a manner as to endow it with certain activities which we now regard as attributes of life. Crude though this conception may be, it constitutes a step, though perhaps a very bold one, into the realm of possibilities for explaining the origin of the first living cell, a subject which we must consider together with all our theories of evolution if we do not wish to remove the inspiration to progress by arriving at an impasse in our theories and our hypotheses. In conclusion, it is well to review briefly the discussion which I have just presented in a very brief form. Out No. 643] ORTHOGENESIS IN BACTERIA 115 of regard for your time and patienee, I have merely pre- sented in outline each of the important considerations which I deem of direct significance to the question at issue. I have presented the diffieulties whieh lie in the path of treating bacteria from the point of view of orthogenesis, and yet have shown that they may be so treated with certain justifiable assumptions as a basis. Having thus treated them, however, I have shown that whether the theory of orthogenesis holds for baeteria or not, it can not be considered as explaining anything, but merely as a mode of deseribing our observations. I have gone into the more fascinating and what seems to me to be the more useful diseussion of the origins of living cells and the position of the bacteria with regard to such primordial cells. I have mentioned the various hypoth- eses which, in my opinion, may be considered to be the most plausible in that connection, and have shown the weaknesses and the strength of each. It has been my purpose to give an unbiased presentation of my own hypotheses and those of others without prejudiee to any so that you might be enabled to diseuss them all and arrive at your own conclusions. Without a thorough review of the literature of bacterial physiology and mor- phology, it is not easy to obtain a broad enough view of the subject to do it justice and I would urge particularly that those who are interested in it acquaint themselves with the absorbing and inspiring literature of the subject of mutation in microorganisms. I believe that it is full of significance for biological progress and I wish that cir- cumstances made it possible for me to present a brief review of it for your consideration. As it is, I must content myself with directing your attention to it and with expressing the hope that my humble efforts in preparing and presenting this paper will constitute a step forward in our progress of thought and experimentation on prob- lems in the evolution of living matter. ORTHOGENESIS AND SEROLOGICAL PHENOMENA PROFESSOR M. F. GUYER UurivERSITY OF WISCONSIN As my diseussion progresses I fear that some of my hearers may be reminded of the old joke about the mon- goose. A stranger carrying an odd-looking box was asked by a man whose curiosity got the better of his good man- ners, what was in the box. The stranger replied that it was a mongoose and went on to explain that his brother was subject to delirium tremens, during the attacks of which he believed that he was being strangled by snakes; this mongoose was to catch the snakes. To the reminder by the inquisitive man that these were imaginary snakes he retorted, ** Yes, I know, but this is an imaginary mon- goose." Since some of our most competent investigators in the fields of geneties and evolution are skeptieal apparently about the whole question of orthogenesis, to them, at least, I shall be making an imaginary attack upon a mythical phenomenon. President Kofoid, however, seemed to think that some of the recent work with immune sera done in my laboratory, which strongly indieates the induction of permanent germinal modifieations, might have possible theoretieal implieations bearing on the question of ortho- genesis, and I agreed to diseuss the subject, although realizing at the outset that the net result would not be a scientifie proof, but merely a suggestion which might possibly be of some value as one of various working hypotheses. First as to orthogenesis itself; is there such a process? Our answer must depend largely upon how we define orthogenesis. It takes but a glance at the literature of the subjeet to see that it has meant many different things to many different people, ranging from a mystical inner 116 No. 643] SEROLOGICAL PHENOMENA 117 perfecting principle, to merely a general trend in devel- opment due to the natural eonstitutional restrietions of the germinal materials, or to the physical limitations imposed by a narrow environment. In most modern statements of the theory, the idea of continuous and pro- gressive change in one or more characters, due according to some to internal factors, according to others to external eauses—evolution in a ‘‘ straight line "—seems to be the central idea. To many, faced by the seeming im- possibility of explaining by natural selection the origins of new characters, it has been apparently merely a wel- come general utility concept by which one may account for the beginnings of new organs, or the development of parts along definite lines, irrespective of utility. For present purposes nothing is to be gained by a review of the different theories of orthogenesis, all so well summarized in Kellogg's ‘‘ Darwinism To-day,’’ and I shall proceed merely on the assumption that, judging from the statistical law of errors, certain variations are apparently not fortuitous, since they tend to accumulate in certain directions. It is customary to add that the lines of development which result are independent of, and in extreme cases may be opposed to, the operation of natural selection. I see no reason, however, for believing that if variations occur in definite directions of no use to the organism, why they may not also occur just as definitely in directions which lend themselves to the per- fecting influences of natural selection. The difficulty in determining this point lies in the fact that an evolution based on the selection of even fortuitous variations must in one sense be orthogenetic, that is, along definite lines, so that there is no way in retrospect of telling whether the underlying germinal variations were purely fortu- itous, or whether they were biased toward an adaptive outcome. Of the various lines of evidence brought forward in support of theories of orthogenesis, the ones which ap- peal most convincingly to me are: (1) those based on 118 THE AMERICAN NATURALIST . [Vor. LVI parallelisms in variation which appear in different branches of the same large group of organisms; (2) those argued from the premise that the very nature of the chemical complexes which constitute the body of a living organism necessarily limit changes to relatively few di- rections; and (3) some of those instanced in the field of paleontology. As to examples of parallelisms in related forms, I am most familiar with conditions to be seen in the color patterns of the pheasants (Phasianine) and the guineas (Numidine). Ina paper’ written some years ago I sum- marized my observation on various features of the colora- tion in a number of groups of genera and species in these two subfamilies as follows: . . there are certain basic tendencies for particular elements of the coloration, such as the formation of eye-spots, barring, and the like, to follow along definite paths of development. When arranged with reference to one of these elements, such, for example, as bar- ring, which is one of the most universal, instead of possessing dis- tinct and unrelated markings, the different species in a given group are seen to be standing merely at different levels in the develop- ment of one, or at most a few, continuous progressions of the special pattern in question. Since when so grouped the gradation in pattern is as much in evidence between collateral kinsmen as be- tween those of direct lineage, one can only conclude that the bias toward a particular line of patterns is the product of fundamental protoplasmic peculiarities implanted in the group as a whole. Further on in the same paper it is shown that where the pattern has become obscured it may be brought to ex- pression again through hybridization. In the interesting group known as the peacock pheasants (Polyplectron) which by systematists is regarded as intermediate be- tween the peafowls and the pheasants in the narrower sense, the varying stages and types of ocellation to be seen afford a good illustration of the point at issue. Quo- ting again from this earlier study: Again as regards ocelli or *eye-spots" in P. chalcurus, which appears to be the most generalized species, one finds no ocellation. The only hint of what is to be realized in the more specialized 1 Jour. Exp, Zool., VII, 4, 1909. No. 643] SEROLOGICAL PHENOMENA 119 members of the group is found in a pronounced purplish and green- ish “metallic” coloration present on certain feathers of the tail. In the male of P. emphanes, while there are numerous green metallie iridescent areas on the feathers of the upper wings and back, they have not yet progressed to the condition of being definite ocelli, although on the tail of this same individual there are two trans- verse bands (the one on the retrices, the other on the upper tail coverts) of ocelli. Still a step in advance, in the male of P. thibet- anum, Gm. (P. alboocellatum Cuv.; type, Mus. d'hist. nat, Paris) ' the small feathers of the wings and the feathers of the inter-scapular region bear distinct small purple ocelli ringed successively with black, light brown, and white. The tail is also banded with ocelli. In the male of P. germaini the wing-coverts and back bear numer- ous green ocelli. The female of this species, as usual less advanced phylogenetically than the male, has the ocelli of the body much less distinctly marked An idea frequently attached to the theory of ortho- genesis, yet which I believe is in no wise a necessary part of it, is that the various grades of a feature supposed to show orthogenesis have arisen as a connected succes- sion, the supposedly most advanced stage having emerged from the stage next in order below it. That the individ- uals of a large group which show some particular char- acteristic expressed in different degree can be arranged in a gradational series with reference to that characteristic is obvious, but, as pointed out by various critics of the theory, this does not prove that the various expressions of the character in question arose thus sequentially.’ In a recent study yet unpublished, made by Miss Sarah V. Jones in the department of genetics at the University of Wisconsin, on the genetical behavior of checks and bars in inheritance in pigeons, for example, Miss Jones corrob- orates the results obtained by Staples-Browne and by Bonhote and Smalley which show that the two patterns are independently inherited in Mendelian fashion with checked-wing dominant to barred. She also found the relation of uniform black to check to be one of simple dominance, and furthermore, that certain grades of check- ing are inherited independently. But she found no evi- dence in support of the well-known view of Whitman that 120 THE AMERICAN NATURALIST [Vor. LVI the various grades of checks form a series moving in one direction, the ultimate outcome of which is the two-barred and finally barless types. And Miss Jones points out that it does not ‘‘ necessarily follow that because the interaction of these several factors produces an apparent epistatie series, the mutations producing the various grades of checking should have occurred in any particular order.’’ And here, to my mind, is the erux of the matter. In order to have what may legitimately be termed orthogen- esis, do the underlying mutations have to occur in any particular order? Is not the very fact that, instead of existing as a medley of wholly unrelated elements, certain characteristics of organisms, such as color markings, can frequently be arranged as parts of a definite pattern or as stages in a general process—does not this in itself indicate directional variation? When one sees in its incipiency, as it were, in one species a character which has attained to an advanced expression in a kindred group, especially where there are intermediate expres- sions of the same characteristic in other related species, is not this indicative of a general trend in variation? For instance, is not the tendency toward the formation of ‘‘ eye-spots "' in the plumage of the pheasants hinted at even in the greenish-black iridescence so often visible in the tail feathers of the common rooster—is not this tendency the expression of what in last analysis must be a germinal bias? To be sure, this bias finds different ranges of expression in different species: as eye-spots on the wing-feathers in some species, on the body-feather: in others. In one species they may occur as a single row of ocelli, in another as two rows, across the tail. And it is obvious that certain of these patterns have not been derived directly from others, since they appear in what are clearly collateral lines. Nevertheless, the tendency to form ocellations is present in many if not all species of this great group. We know nothing of the order in which the mutations occurred which brought about any partic- No. 643] SEROLOGICAL PHENOMENA 121 ular condition that at present exists in the group. Some of them may have been small and sequentially related, and it is not impossible that the extremes were thrown in one line, while grades of less advanced type came into expression in collateral lines. It is also clear that even should certain grades have arisen as a progressive series there is no reason, from the viewpoint of the mutation theory, why any two partieular grades should not show the characteristies of Mendelian unit-characters, irrespec- tive of the order of their origin. The important point is that in this group when mutations occur in certain col- or pattern-controlling factors, whether great or small, they tend toward the formation of eye-spots in some degree. While we know little of the chemistry of animal pig- ments, the reactions involved in color-production in cer- tain plants are better understood, since many of the pig- ments have been extracted and analyzed. Along with the understanding of strueture that has been gained in the chemical studies of synthetic dye-stuffs, has come con- siderable knowledge of the relation between color and molecular structure. In many cases, for instance, as Nietzki? has shown, where the pigments of most simple construction are yellow, by increase of molecular weight they change to red, next to violet, then to blue. A good review of the theories of color in organic compounds, given as an introduction to her own painstaking and valuable chemical researches? on ‘‘ Pigments of Flower- ing Plants,’’ will be found in a recent paper by Dr. Nellie A. Wakeman. Most of the information about organic pigments used in the present discussion has been obtained from this source. Upon reading such a piece of investigation together with the accompanying discussion of related studies one is impressed by the fact that a comparatively few proc- esses underlie most pigmentary changes in plants. En- zymes—hydrolases, reductases and oxidases—frequently 2 Trans. Wis. Acad., XIX, Part II, pp. 767-906, Madison, Wis. 122 THE AMERICAN NATURALIST [Vor. LVI play an important part in the formation of pigments, or in changes in pigments already formed. Shade of color, for example, is evidently often merely a function of oxidase content. With a graded increase of oxidase, therefore, a plant might be put through a regular gamut of color effects. In general, the mere addition of hy- drogen to dye-stuffs reduces them to the corresponding leuco base. Armstrong, in his quinone theory of color, makes much of the quinones as the colored compounds in dye-stuffs and maintains that the corresponding color- less compounds are hydro-quinones. The structure and size of the pigment molecule itself seems to be an impor- . tant factor in color. Hydrocarbon radicals, for instance, deepen the tint; the addition of hydrogen raises the tint. In analyzing any particular case one has to take into ac- count the original molecule, the position of any group introduced and the number of groups introduced. It is possible, for example, by the introduction of a tint-deep- ening group to deepen the color, but by introducing two or three more such groups to throw the absorption wholly outside the visible spectrum and thus do away with the color. From this it is clear that two compounds may be closely related in constitution and yet one be colored, the other colorless. As to why, physically, in both the aro- matic and the aliphatic series, color is produced in certain compounds and not in others, various investigators have arrived at the conclusion that the cause of color is due to ‘‘ the making and breaking of contact between atoms, thus giving them marked activity,’’ a process known as isorropesis, and Miss Wakeman goes on to explain: This change of linkage which must accompany the transforma- tion of one modification of the compound to the other is the source of the oscillations producing the absorption bands. If these oscilla- tions are synchronous with light waves of a high frequency they give rise to absorption bands in the ultra violet and the compound is colorless. If, however, they are less frequent, the absorption band appears in the visible portion of the spectrum and this absorption of colored rays results in the compound taking on complementary color. No. 643] SEROLOGICAL PHENOMENA 123 Among interesting facts that come to light in Miss Wakeman's summarization may be mentioned the fol- lowing: all organie pigment molecules are unsaturated ; the quinone grouping is one of the best known of the chromophorous groups; by far the largest number of plant pigments are referable to hydrocarbons of satura. tion C.H, .,, and C,Ha.,,; and finally, that it is the relation of the ehromophorous groups to each other and to the rest of the molecule, and not their mere presence in the molecule, that postulates color in a substance. I may seem to have dwelt unduly upon pigmentation in plants but I have tried to go into the matter only suffi- ciently to give a glimpse of some of the real conditions which underlie some of the ‘‘ unit-characters ’’ we are juggling about in genetics, and around which we are attempting to frame hypotheses of evolution. Color, per- haps more than any other one thing, has in recent years been utilized for genetical observations. And when it is known that in many cases color is merely a function of the degree of oxidation of some fundamental com- pound, or the introduction or subtraction of some hydro- carbon radical, it does not tax the imagination to conceive of how it is possible to have series of color ** characters ”’ that in the parlance of organic evolution represent ortho- genetic series. As simple a matter as the relative degree of oxygen supply, probably determined by the amount of an oxidase present, may account for various stages of color. It is manifest, moreover, that in a given group there might easily be a tendency toward increase in oxi- dase-production, which if present unequally in different collateral lines might give us just the uneven condition, with different species standing at different levels of ex- pression of the trait in question, which exists in various alleged cases of orthogenesis. It is obvious, furthermore, that any higher state of development of the character in a particular species need not have sprung from the next lower stage, but may have had its immediate origin from any level of the scale. 124 THE AMERICAN NATURALIST [Vor. LVI Most of the constitutional changes which go on in the living organism seem to center in the proteins of the protoplasm. Metabolism is largely a question of the dis- ruption and reconstruction of the various cell-proteins. In the cell, moreover, the characteristic protein-complexes themselves determine the nature of the synthesis that shall go on. When synthetic activity is more than sufficient to make good metabolic waste, growth results, and when such increase becomes overgrowth and takes the form of a de- tached individual, we pronounce it reproduction. Weare then in position to talk about inheritance—the fact that a new individual possesses the properties and, under similar conditions, therefore, will express the activities and take on the appearances of the earlier or parent form. Thus the germ-cell is a reduplication of the zygote from which it sprang, a detached bit of living matter made up largely of certain characteristic protein-complexes. Even the simplest protein is a huge molecule built up of a series of different kinds of amino-acid ‘‘ nuclei ’’ which in differ- ent proteins differ in numbers, kinds and arrangements. Certain of them seem necessary to all proteins, others are present in only some proteins. Each amino-acid, be- sides being linked to its fellow, has a replaceable hydrogen atom which may be exchanged for any one of several radicals or ‘‘ side-chains." Furthermore, the ordinary native proteins are secondarily compounded of a number of the simpler blocks formed of linked amino-acids. The unitary amino-acids which enter the blood as protein digestion-produets are used as building units again, each cell selecting the kind of units it requires for replace- ments in its own proteins. We have no reason to believe that the proteins of the germ-cells have any mysterious powers associated with them that are not shared by any or all of the somatic cells. The modern view of embryogenesis and histo- genesis no longer finds it necessary to picture troops of pangenes departing from their home in the nucleus at just the proper time to take possession of the cytoplasm No. 643] SEROLOGICAL PHENOMENA 125 and turn the cell into a specifie tissue-cell. Each tissue- cell probably retains all the essential properties of the original fertilized ovum from which it has directly de- scended. In many cases, among lower organisms, at least, we know that somatie cells detached as buds or experimen- tally ean through regulation and new growth reconstruct themselves into complete organisms. That a particular cell takes on the characteristics of a specific tissue seems to be determined by the special environment in which the cells happen to be placed in the organism.? Exactly how much chemical difference there is between - two unlike tissues or between the cells of a particular tissue and the germ-cells is not known. As far as we have any cytological evidence to the contrary the nuclei, at least, of the tissue-cells are not essentially different from the nuclei of the germ-cells. While the various tis- sues differ very much in appearance, this is mainly the result of the accumulations of intercellular products or of cytoplasmic modifications. And many of the latter may be largely changes in colloidal state rather than fundamental changes in chemical composition. In any event, the new condition is one which has sprung from a cellular chemical constitution similar to that of the orig- inal zygote. And if this is true, would not any internal or external agent which could affect particular proteins of the somatic cells be able also to influence the homol- ogous elements in the germ-cells? Inasmuch as I have already twice reported to this Soci- ety on the work‘ of Dr. Smith and myself with fowl-serum immunized to rabbit-lens. by means of which, through injections into pregnant rabbits, we succeeded in obtain- ing defective-eyed young, I shall not again relate the details. The most interesting thing about the experiment was the fact that the eye-defects were transmissible to subsequent generations, and inasmuch as the condition 3 Cf. Child, ‘‘ Individuality in Organisms ’’; also ‘‘ Origin and Develop- ment of the Nervous System.’’ 4 Jour. Exp. Zool., 31, 2, 1920. Am. Nart., LV, 1921. 126 THE AMERICAN NATURALIST [Vor. LVI could be passed down through the male line alone it is evident that it is based on changes in the germ-cells. In later experiments we obtained similarly defective young by injeeting rabbit-lens into pregnant rabbits, al- though we secured this result only after repeated trials and in the young of but one female. Our belief is that the eytolytie serum not only attacked the newly forming fetal lens, but also its representatives in some of the germ- cells of the fetus. "This implies, of course, that there is a sufficient thread of chemical identity between the two to render them both susceptible to the same specific in- fluence. ; For such serumal. effects to be of significance in evolu- tion, however, the antibody or other factor operative on a given tissue-protein would have to be one that could arise directly in the organism itself. But since it is known that animals will develop anaphylaxis against tis- sues of their own species, and that a rabbit can be made to build spermatotoxins against its own spermatozoa, it is reasonable to suppose that if an animal’s own tissues became displaced, injured or otherwise modified, they might cause the production of antibodies. And these, carried by the circulating fluids of the body into the gonads, would have opportunity to influence suċh protein- complexes there as were similar to those in the tissues which served as antigens. Nor need the germinal and somatic elements in question be identical in constitution, for it is known that while an antibody against a particular tissue shows its highest degree of specificity only against that tissue, nevertheless, it will also react in some degree with other tissues of the same individual. This phenom- enon, termed species-specificity, clearly indicates that there is a broad common basis of chemical identity under- lying all the tissues of an organism. It is not unreason- able, then, to believe that there is sufficient chemical iden- tity between the proteins of tissue cells and the related proteins of the germ for both to be influenced by the same agents. To construct a working hypothesis upon the possibilities No. 643] SEROLOGICAL PHENOMENA 1A before us we might suppose that as long as all tissues are in normal physiological balance no antibody or similar modifying agents are developed. The germ-cells, there- fore, maintain the exaet constitution they derived from the Zygote from which they descended. But with the occur- rence of injury, undue stimulation or pronounced change in any part of the body, serological changes would prob- ably be produced in the blood-stream and the germ-cells would then be exposed to possible modifying influences. This would be more likely to happen, of course, if the ex- posure continued through a long period of time. If the in- fluence were disintegrative or poisonous as the eytolysins or cytotoxins evidently are, then probably degenerative changes would ensue. Such a hypothesis affords, perhaps, a plausible explanation of such deteriorative evolutionary processes as those seen in the formation of vestigial or- gans. As a concrete illustration, purely hypothetical of course, we may suppose that such a species as the mole in gradually changing to a subterranean existence would meet with frequent injuries to the eyes, and that, as a result of the ensuing inflammatory and suppurative con- ditions, resorptive influences would be set to work which not only affected the proteins of the eye, but also the related proteins of the germ. Once the degenerative proc- ess got to going, it might for a time be based in each new generation upon both the direct chronic irritation to the eye and the parallel changes induced in the germ. Fi- nally, we may suppose that the somatic influence would cease when the eyes became of small size and the eyelids remained permanently sealed, but that the conditions induced in the germ would persist. If such atrophied eyes continued to be resorbed more or less in each gen- eration, however, variation toward still further reduction might continue in the germ. Such a progressive degen- eration might possibly be ranked as an instance of re- gressive orthogenesis. But what of the progressive aspects of evolution? Can serological reactions be invoked here with any show of reason? One great difficulty in dealing with progressive 128 THE AMERICAN NATURALIST . [Vor LVI |. variation is that we know almost nothing about the chem- ical and physical factors which underly growth, hyper- trophy, hyperplasia, metaplasia, or other changes in somatie tissues due to changed internal relations or to unusual environmental stimuli. If. we only knew, for instance, what happens in even as simple a case as when epidermal cells develop into a callous in response to undue pressure or friction, we might have a clue as to how, also, constructive changes might occur in germ-cells; but we have no such knowledge. Certain types of tissue-overgrowth' in which there is increase in the size of the tissue-elements (hypertrophy) or in the number of such elements (hyperplasia) are interesting in this connection. For example, increased strain in bone leads to increased growth of bony tissue, or excessive exercise leads to overdevelopment of certain muscles. In such cases an increased demand on the nutri- tive stream caused by unusual katabolism results in a physiological hypertrophy. That is, an excessive syn- thesis of certain types of proteins is set up. Does the impetus to such extra synthesis extend also to the related, though unstimulated, tissues? I know of no evi- dence bearing directly on this point, although such phe- nomena as compensatory overgrowth show that there are influences at work outside the immediate tissue itself which are instrumental in inducing the hypertrophy. For example, if one of a pair of organs (lung, kidney, testicle, thyroid) is lacking or is destroyed, the other en- - larges in a short time to the size and functional capacity of the pair combined. There is considerable evidence, particularly in the field of pathology, to show that under ordinary conditions the tissue-elements exert a sort of balanced reciprocal restraint, but disturb this and the whole system is more or less deranged until a new equilib- rium is established. Since in compensatory adjust- ments the compensating organ is generally not in direct connection with the one which is missing or disturbed, it seems probable that the agent which incites the hyper- 5 Cf. any General Pathology. No. 643] SEROLOGICAL PHENOMENA 129 trophy is carried by the blood, although the possible influ- ence of the nervous system must also be reckoned with in higher animals. And if there is such a serum-borne agent in the case of compensatory hypertrophy, may there not also be one in that of the ordinary physiolog- ical hypertrophy of the exercised muscle or stressed bone? If so, we must keep our minds open to the pos- sibility that it may also stimulate the germinal proto- types of such proteins to additive functioning. For if we had but a single side-chain in common between a pro- tein of somatie tissue and a protein of the germ, any- thing that could affeet one might well be expected to affect the other. Again, mechanieal stimuli, if not too severe, and vari- ous irritative substanees in amounts sufficiently small not to be destruetive or poisonous to the tissues, may stimu- late cells to overgrowth. Very small amounts of arsenic or phosphorus, for example, may thus affect the kidneys and liver, and minute doses of phosphorus may cause in- creased growth of bone. In such cases also it is probable that serological changes are involved. This is almost certainly true in eases of viearious overgrowth, where an organ supposedly of related funetion takes over wholly or in part the work of another tissue. An example of this is seen in the enlargement of the pituitary gland when the thyroid is atrophied or removed, or the com- pensatory enlargement of the hemolymph glands and bone-marrow following removal of the spleen. Still further may be cited the phenomena of metaplasia, in which, through modification of function and nutrition, specialized tissues develop from cells which normally pro- duce tissue of another order. It sometimes happens, for example, that the choroid coat of a severely injured eye, after the lapse of considerable period of time, will de- velop a layer of true bone. In fact, metaplastic forma- tion of bone is common in many tissues. Such facts show that many, if not all, tissue-cells have the capacity to form very different kinds of tissue in different en- vironments, and suggest that they retain all the inheren- 130 THE AMERICAN NATURALIST [Vor. LVI cies of the germ. They are what they are somatically : because of the special restrictions or excitations of their particular situation in the organism. But if these highly specialized tissue-cells can be so stimulated as to form an entirely different type of tissue, may not such stimulative influences invade even the germ-cell with modifying ef- fects? Lastly, there are the endocrinal secretions to be reck- oned with. Since they are at present popular subjects of research and are constantly being alluded to and dis- cussed in the biological literature of the day, I need not review the field. It is evident that in them we have cir- culating through the body a series of powerful substances eapable of producing profound effects in any or all parts of the body. Through them, apparently, various organs effect reciprocal stimulations and the tissue-complexes of the entire body are maintained in a state of general physiological equilibrium. Both clinical and experi- mental evidence reveals that hypertrophy or atrophy of an endocrine gland may be followed by marked altera- tions of structure or function in one or more regions of the body. Thus the eretinism or the myxcedematous condition which results from removal of the thyroid or arrest in its function, or the symptoms resembling exophthalmie goiter following hyperthyroidism are famil- iar examples. The remarkable overgrowth of the bones of the extremities and head known as acromegaly—mor- bid giantism— associated with enlargement of the pitui- tary body is another. Also the relations of the gonads to the secondary sexual characters are well known, as is that of the fetus to the normal hypertrophy of the mammary glands in pregnancy. Since hypertrophy or atrophy of an endocrine may pro- duce deep-seated permanent changes in various tissues of an organism, I would again point out the possibility that the germinal homologues of the proteins of such tissues, if such there be, might likewise be permanently modified, and that if for some reason there came a con- stant inherited increase or diminution of an endocrine No. 643] SEROLOGICAL PHENOMENA 131 gland, or an environmental modification of it generation after generation, we might have in its waxing or waning output the exeitant necessary for germinal changes which become outwardly expressed as a series of orthogenetie changes. That this suggestion of endocrinal influence on the germ is not so far fetehed as would appear at first sight, is evident, I think, when we recall that certain of the eonditions which ean be indueed in individuals by experimental or pathological endocrinal upsets are known to occur also as congenital defects which are inheritable. For example, short-fingeredness (brachydactyly) may be induced after birth by too much pituitary secretion, but such a condition is also well known as a congenital defect which is hereditary. In the latter case, is it more reason- able to suppose that the short-fingeredness, could we trace it back into the germ, is really represented by a factor that has to do primarily with the finger or with a factor directly concerned in some way with the pituitary body? And did hypertrophy of the pituitary body origi- nally induce the heritable type of brachydactyly? We do not know, but the parallelism of the two conditions, it seems to me is highly suggestive. And let us glance for a minute at one of the well- known studies on orthogenesis; that of Ruthven? on the variations of seutellation in the garter snakes. In his own words: . it seems to me that the most tenable hypothesis of the evolu- tion of the genus T'hamnophis is that it originated and became dif- ferentiated into four main groups in northern Mexico. From this region the groups radiated in all directions, but principally to the northward, and wherever they entered different regions the changed environmental conditions acted as an unfavorable stimulus which retarded growth, and differentiated the groups into dwarfed forms. And in another place he generalizes as follows: (1) That the maximum scutellation and size in the genus Tham- nophis occurs in the center of dispersal, and the forms that have been produced in the history of its migration have been formed principally by dwarfing and by a reduction in scutellation; (2) that the variation in the number of scales in the different series is 6 Bull. 61, U. S, Nat. Mus., 1908. 132 THE AMERICAN NATURALIST [Vor. LVI definite and not promiscuous, and is correlated in a remarkable degree with changes in the environment. Have we not here a condition strikingly like what we should expeet to find if some factor or factors, external or internal, were operating in such a way as to lessen the output of some endoerinal secretion concerned in growth or the determination of size? This is at least a possi- bility worthy of consideration. In elosing may I say that in what I have put before you I do not pretend to have supplied the established facts necessary for founding a scientific theory. The dis- eussion is largely a series of suggestions, a mere work- ing scheme which takes into aecount various phenomena that appear to be related, and whieh in their present states of disclosure seem to lend themselves to some such interpretation as I have tried to give. It is presented because, in my estimation, it suggests a line of thought we may well entertain when we are wrestling with our several problems of geneties, variation and evolution. For if it can be clearly established that any one of the serologieal influenees ean reach specifically from soma to germ then it becomes a plausible hypothesis that many of them do. If a changed or changing external or in- ternal environment causes a long continued physiological stress of certain parts, then as long as this stress is ac- companied by changed conditions of the circulating fluids of the body, so long also will the germ-cells be exposed to these influences. If they are such as to induce varia- tions in definite directions, orthogenesis must be the out- come. And if serological influences play an important part in adaptive somatic changes, such as adaptive hyper- trophies—or for that matter, adaptive atrophies—then we have the way open to conceive of how adaptive ger- minal changes may likewise be the outcome of these same influences. It is a noteworthy fact that in the geological past whenever conditions suitable for new types of existence occurred, new forms of life well adapted to the condi- tions appeared. This has happened not only once, but No. 643] SEROLOGICAL PHENOMENA 133 repeatedly. And since satisfactory adjustment to the new conditions must mean not only one, but many favor- able and interrelated variations, it seems almost incredible that the adaptedness characteristic of the organisms in question was attained merely through the operation of natural selection generation after generation on as- semblages of purely accidental mutations. Paleontolo- gists tell us that times of marked evolutionary change have coincided with periods of great geological change— extremes of temperature, moisture or drought, or fairly rapid fluctuation between such extremes. And while such conditions would undoubtedly favor a maximal opera- tion of natural selection, it is well to remember also that the severe strains of somatic adjustment forced upon or- ganisms existing at the time would doubtless result in a maximal sweep of serological influences through the sorely pressed body. Although I have emphasized one side of the problem of variation, I am not unmindful of the remarkable stability of the germinal protoplasm as we see it expressed i in or- ganisms to-day. It is obvious that not every minor or temporary alteration in somatic mechanism is reflected in the germ to any measurable extent. Since probably no two living things of any kind are equally susceptible to external influence, individual germ-cells doubtless vary in susceptibility, possibly even the same germ-cell would respond differently at different stages of maturity. It is not unreasonable to believe, moreover, that only a few out of many germ-cells might be sufficiently affected to make a perceptible difference. I have already expressed the opinion’ elsewhere that ‘‘no one to-day, qualified by his knowledge of embryology and genetics to the right of an opinion, would, I think, deny that the new organism is in the main the expression of what was in the germ- line, rather than of what it got directly from the body ." But we know that the germ does change from irk "s time and it seems to my mind not illogical to sup- pose that at least some of the changes are specifically re- lated to changes in the soma. 7 Ax, Nart., LV, Mar.-Apr., 1921. ORTHOGENESIS AS OBSERVED FROM PALE- ONTOLOGICAL EVIDENCE BEGINNING IN THE YEAR 1889! DR. HENRY FAIRFIELD OSBORN AMERICAN Museum or NATURAL History 1. THE OBIGIN oF SPECIES Tur Origin of Species is now clearly understood in the hard parts of invertebrates and of vertebrates, and there is little to be added as to the modes of mechanical evolution. No chances or experiments are tried by Na- ture. The process is continuous, adaptive, mechanically perfect in every Mutation of Waagen. As shown in actual observations by all close students of vertebrate and invertebrate morphology during the last fifty-two years, and as summed up in the remarkable contribution of D'Arey Wentworth Thompson (1917) on ‘‘Growth and Form," animal mechanisms compete with each other in close analogy to humanly made machines—automo- biles, typewriters, aeroplanes. Consequently, while Na- ture is constantly standardizing her machines through individual competition and producing flocks of birds and shoals of fishes which are so precisely alike that animals of the same age, sex, environment and heredity show no perceptible variation, she is also frequently substitut- ing more perfect and more adaptable machines and dis- carding older and less adaptable ones, exactly as man is doing in the case of his automobiles, his typewriters, and his aeroplanes. Thus the naturalist and the paleontolog- 1 Illustrated by twelve lantern slides exhibiting mutations of Ammonites, of Spirifer, of Paludina; rectigradations of the grinding teeth of lemuroid Primates; evolution of proportion from the rhynehocephalian type of Hat- teria to the dinosaurs and birds of the genera Deinodon, Struthiomimus, and Diatryma. 134 No. 643] ORTHOGENESIS 135 ist are alike impressed with the incessant action of Nat- ural Seleetion on animal mechanisms and with the new testimonials to this aspeet of Darwin's great principle. When it comes to the origins either of new characters or of new proportions quite different is the attitude of observers of mechanical evolution; no evidence whatever has been fortheoming from the same fifty-two years of elose observation and research as to the causes of origin, at the same time the modes of origin of all mechanical characters are indubitably orthogenetie. To further clarify the bearing of paleontology on orthogenesis, I desire to point out that all visible me- chanical evolution goes hand in hand with invisible physi- cochemical evolution; and that there are steps in evo- lution which are primarily physical, others which are primarily chemical, others which are primarily mechani- eal. Therefore the experimental botanist, zoologist, bio- chemist, biophysicist, or geneticist, has the opportunity to win immortal fame by discovering the causes of me- chanical evolution. Meanwhile the paleontologist enjoys the entirely unique position of being the only competent observer of the Origin of Species so far as specific characters are recorded in the hard parts of animals and the relatively few soft parts which are preserved in a fossil condition. 9. ORTHOGENETIC ORIGIN or NEw CHARACTERS All agree that sound induction either as to the origin of new characters or their transformation is an exceed- ingly difficult matter. It has taken me thirty-three years of uninterrupted observation in many groups of mam- mals and reptiles to reach the conclusion that the origin of new characters is invariably orthogenetic. ; In 1889 I first observed (Osborn, 1889.46) that new eusps originate on the grinding teeth of Eocene Pri- mates, now recognized as lemuroids, in a definite and adaptive manner from minute shadowy beginnings which are mechanieally adjusted to similar minute shadowy 136 THE AMERICAN NATURALIST [Vor. LVI beginnings of opposing eusps in the other jaw; whereby there evolves a continuous reciprocal mechanism not dis- similar to the reeiproeal serviees of the Yale key and the Yale lock. The evolution of the key below proceeds with the evolution of the lock above. ‘The process does not go very far in the Primates, but in the purely herbivo- rous ungulates, like the horse and the elephant, the re- eiproeal grinding mechanism reaches a degree of com- plexity to which the most intrieate lock and key devised by man present but a feeble parallel. Every mechanical deviee in the upper grinding teeth, adapted to the fine comminution of grasses, is reversed in the lower grinding teeth, on the principle of mechanical action and reaction; nowhere in nature is reciproeal mechanieal co-adapta- tion more perfectly evolved than in the upper and lower grinding teeth of mammals. Between 1889 (Osborn, 1890.47) and 1891 (Osborn, 1891.53) I made what I now believe to be an wnsound in- duction from this evidence that this continuous mechani- eal origin tended to support the Lamarekian theory of the inheritance of adaptive reactions. I first termed the orthogenetie process ‘‘definite variation’’; later I termed it **progressively adaptive variation’’; by the year 1908 I realized that these new adaptively arising tooth ele- ments were not variations in Darwin’s sense at all, and I applied to them the distinctive term rectigradations (Osborn, 1908.314). In the meantime I abandoned the Lamarckian explanation and in 1895 (Osborn, 1895.97) I started out upon a search for the unknown factors of evolution, a search in which I am still busily occupied. To return to the difficulty of making sound inductions as to the origin of new characters in hard parts, in 1889 I opened a long correspondence with the leading expo- nent of Darwinism in Great Britain, Edward B. Poulton, who admitted the evidence but interpreted the facts in the Darwinian way, namely, as the selection of mechani- eal successes from non-observed mechanical failures. It is a good thing to have a number of skeptical friends No. 643] ORTHOGENESIS 137 about; it sharpens your powers of observation and makes you much more eautious about your inductions. My original observations on the Primates required cor- roboration, and this I have sought through the observa- tion of the origin of new charaeters in many other kinds of mammals traced in their evolution over very long periods of time, especially the horses, the rhinoceroses, and recently the proboscideans, but most profoundly and exhaustively the titanotheres, an extinet family remotely related to the horses, which I have studied monographi- cally for twenty-one years. Even by trying to keep an absolutely open eye and mind, entirely uninfluenced by any theory, or preconcep- tion, or opinion, I have been unable to find a single ex- ception among these many different kinds of mammals to the observations made on the Primates in 1888 and 1889; not a single new organ is observed to arise for- tuitously or indefinitely; it always arises gradually, con- tinuously, and adaptively from its minute shadowy beginnings. This continuous reciprocal, mechanical eo- adaptation seems to be an established fact in evolution, and is established most strongly where explanation or search for causes seems to be most difficult. I am not enthusiastic about the adoption of the term orthogenesis, admirably significant as it is in its Greek derivation, first, because Eimer connected it with La- marck’s and Buffon’s principles of inheritance of ac- quired modifications, and, second, because it does injus- tice to the first great observer of direct adaptive origins in nature, namely, the German paleontologist Wilhelm Heinrich Waagen, whose observations in 1869 laid the foundation of all subsequent work both among the in- 1 Osborn, H. F., ‘‘ The Titanotheres of Ancient Abbe cs Dakota, and Nebraska. Life and Geography of the Central Rocky Mo ain Region in Eocene and Oligocene Times. Evolution of the ri ig Th Causes of Development and Extinction of Mammals,’’ U. S. Geol. Survey Monograph No. 55. [Unpublished.] Completed for the Survey June 30, 1920. This monograph is the most complete and exhaustive analysis that has thus far been made of the evolution of any family of organisms 138 THE AMERICAN NATURALIST [Vor. LVI vertebrates and the vertebrates. To the best of my knowledge he was the first naturalist to observe how new species actually arise in nature. Compare Waagen’s description (1869) of the genesis of new characters in the shells of cephalopods (Ammonites subradiatus) with those which Osborn (1889-1921) has observed in the teeth: ‘Thus the species if considered as such may be con- ceived and considered as a species, but in contrast with earlier or later forms [1.e., ancestors or descendants] as a mutation. Now as regards the value of these above- defined conceptions, variety and mutation, on closer con- sideration a quite decided difference in value becomes apparent. The former conception [variety], in the high- est degree variable, appears to be of small systematic value; while the latter [mutation], although in minute characters, 1s highly constant, always surely recogniz- able; on which account far greater weight must be put upon Mutations, they ought to be very precisely denoted and held fast to with great persistence." Twenty years later the German paleontologist Mel- chior Neumayr observed this process of continuous de- velopment, generation after generation, in a certain defi- nite direction for which he proposed the term ''Muta- tionsrichtung.’’ Thus the ‘‘mutation of Waagen"' arises continuously through the inner working or tendency, the ‘‘ Mutationsrichtung’’ of Neumayr. It was not until 1894 that William B. Seott brought Waagen's term ** Mutation" to the notice of vertebrate paleontologists in this country, in antithesis to Dar- win’s term Variation. Waagen's ‘‘Mutation’’ means one thing, Darwin’s ‘‘Variation’’ means quite another, as pointed out by Scott above. The term Mutation in Waagen’s sense is now widely but not universally used by paleontologists to designate intermediate gradations of minor taxonomic rank which are observed in ascend- ing or descending series of animals to connect the larger stages of evolution which we call Species. As an ele- No. 643] ORTHOGENESIS 139 mentary species a ‘‘Mutation’’ of Waagen is compa- rable to a ‘‘Mutant’’ of De Vries in external appearance, but not in mode of origin, because one arises through a continuous '' Mutationsrichtung," while the other arises through aecidental germinal saltation. To my mind the continuous or discontinuous mode of origin either of a ‘‘mutation’’ or of a ‘‘mutant’’ is of small account as compared with the fortuitous or orthogenetie nature of the impulse in the germplasm which gives rise to it.” So far as I know all observers of the hard parts of extinct animals, whether vertebrate or invertebrate, con- firm this classic observation of Waagen, and many in this special field of observation also confirm the ‘‘ Muta- tionsrichtung’’ of Neumayr. So far as I personally have observed, this principle of ‘‘Mutationsrichtung”’ is especially dominant in the origins of charaeters; here at least other interpretations are not applicable; there is no question of Seleetion between two alternatives, adaptive and inadaptive, because the inadaptive does not occur, the whole process is adaptive and the differ- ‘ence between two organisms is the rapidity and direc- tion with which the ‘‘Mutationsrichtung’’ is acting. This is the same in the hard parts of the molluses Am- monites, Paludina, and Planorbis, as it is in the mam- mals Equus, Rhinoceros, and Elephas. 3. Tue Ortarn of New Proportions In the evolution of proportions, that is, proportions in the different parts of skeleton and skull as in Spheno- don, Deinodon, Struthiomimus, Diatryma, it appears probable that Selection may be constantly working on all adaptive fluctuations of proportion in connection with ontogenetic modifications in proportion which are also adaptive, as in the classic case cited by both Dar- win and Lamarck of the length of the neck of the giraffe. 2 TF, A, Bather in 1905 (Proc. Geol. Soc., Vol. 61, pp. Ixxii-Ixxiii) most clearly elucidated Waagen's conception of the Formenreihe and of the Muta- tion in Ammonites. 140 THE AMERICAN NATURALIST [Vor. LVI It has been demonstrated experimentally that the limb proportions in the brief life of a dog may be modified from the cursorial to the saltatorial type by amputating the fore limb. "This is a process of reciprocal Modifica- tion and Selection which Osborn, Baldwin, and Morgan term Organic or Coincident Selection. I have devoted an immense amount of study to the causes of the evolu- tion of proportion and have come to the conclusion that orthogenesis in the evolution of proportion may be ap- parent rather than real. In other words, whenever a character assumes a survival or elimination value, it may develop very rapidly through the selection of fluctua- tions in the right direction and may result in apparent but not real orthogenesis. 4, Summary AS TO ÜRTHOGENESIS The visible ‘‘mutation of Waagen,’’ or ‘‘definite vari- ation" or ‘‘rectigradation’’ of Osborn appears to de- pend on the ‘‘Mutationsrichtung’’ in the germ-plasm. The final question in my mind, as in yours, must be, if such a ‘‘Mutationsrichtung’’ exists, is it the ‘‘internal perfecting tendency,’’ is it the ''vitalism," is it the ‘creative evolution” which the majority of biologists are so skeptical about? I observe that it is not. I observe that while the ‘‘Mu- tationsrichtung’’ is a real process, it differs from any kind of internal perfecting M in the faet that it which it ‘migrates. For example, the internal Sertost- ing tendency to arboreal life does not manifest itself when the animal seeks an aquatic life. Conversely, aquatic adaptations are not constantly springing up among arboreal mammals. Observations on fossil forms have led to Dollo’s remarkable generalization regarding *alternate adaptation," which renders any form of in- ternal perfecting tendency in any predetermined direc- tion inadmissible. No. 643] ORTHOGENESIS 141 Summary of Observations.—In the hard parts of ani- mals orthogenesis is observed both in the origin of new adaptive characters and in the evolution of proportions. (1) The induetion as to eause may be different in the two eases. (2) In the origin of new adaptive characters orthogenesis is attributable to definite germinal tend- encies. (3) The origin of changes of proportion which are subject to modification may be partly attributable to Organie Seleetion. (4) There is positive disproof of an internal perfecting tendency (Vitalism) in either the origin of new characters or the origin of proportions. (5) There are certain changes of length and breadth proportion both in the shells of invertebrates and the skulls of vertebrates which can not be explained by Or- ganic. Selection. (6) There is very strong support in fossil series for Selection incessantly acting on all char- acters of survival or elimination value. The above six principles are those which I have de- rived from forty years of continuous observation; they are actual modes of the mechanical evolution of new species for which we have no theoretic explanation, un- less it be that of Organic Selection in the single case above noted. Summary of Opinions.—I may add as a matter of personal opinion and hypothesis three points: first, that we are as remote from adequate explanation of the na- ture and causes of mechanical evolution of the hard parts of animals as we were when Aristotle first speculated on this subject three hundred years B.c.; second, that the chief outlook for experiment is in the domain of physics; third, that the explanation, if ever it is to be found, is to be along the lines of four systems of energy (— Tetra- plasy, Tetrakinesis, Osborn) which surround the origin and development of every charaeter in every organism; fourth, I think it is possible that we may never fathom all the causes of mechanical evolution or of the origin of new mechanical characters, but shall have to remain con- tent with observing the modes of mechanical evolution, 142 THE AMERICAN NATURALIST [Vor. LVI just as embryologists and geneticists are observing the modes of development, from the fertilized ovum to the mature individual, without in the least understanding either the eause or the nature of the process of develop- ment which goes on under their eyes every day. In conclusion, it is the great biological achievement of the last half century that paleontologists have dis- covered how new characters and new species originate. It may be the achievement of the experimental biologists during the next half century to explain why new char- acters and new species originate. HENRY FAIRFIELD OSBORN. BIBLIOGRAPHY ON SINGLE CHARACTERS, MODES OF ORIGIN AND TRANSFORMATION, ORGANIC SELECTION 1889.46 The Paleontological Evidence for the Transmission of Acquired Characters. Amer. Naturalist, Vol. XXIII, No. 271, pp. 561- 1890.47 The * Paleontological Evidence for the Transmission of Acquired cters. Nature, Vol. XLI, pp. 227, 228. 1891.53 Pare imei riations Inherited? (Opening a Diseussion upon the Lamarekian' 21 in Evolution. American Society of Naturalists, Boston, Dee. 31, 1890.) Amer. Naturalist, Vol. 1894.92 Certain Principles of Progressively Adaptive Menten ye served in Fossil Series. Nature, Vol. 50, No. 1296, p. 1895.97 The Hereditary Mechanism and the Search for the ebd Factors of Evolution. Biol. dip Marine Biol. Lab. of Woods Holl, 1894, Ginn & Co., Boston, 1 1896.108 [Abstr.] [A Mode of Evolution requiring tieiiber irgend hod d acte N. Y. Acad. Sci., Vol. XV, Mar. 9 and Apr. 13, rae pp. 14i, 142, 148. 1898.134 The Biological Problems of To-day: Palzontological Problems. [Diseussion before the annual meeting of the Ameriean So- ipd of Naturalists.] Science, N.S., Vol. VII, No. 162, pp. 145-147. 1902.212 Homoplasy as a Law of Latent or Potential Homology. Amer. a Vol. XXXVI, No. 424, e 259-271. 1907.303 Evolution It Appears to the Paleontologist. Science, N.S., ol. muito No. 674, pp. pie 1908.314 Coincident Evolution Through Rectigradations. Science, N.S., Vol. XXVII, No. 697, pp. 749-752, 1909.331 To the Philosophic Zoólogist. Science, N.S., Vol. XXIX, No. 753, pp. 895, 896. No. 643] 1911.353 1912.362 1912.372 1914.412 1915.416 1915.421 ORTHOGENESIS 143 Ti om une ge of Waagen and ‘‘Mutations’’ of De Vries or [and] ctigradations'' of Osborn. Science, N.S., Vol. Darwin's Theory o f Evolution by the Selection of Minor Salta- tions. Amer. Naturalist, Vol Sind No. 542, pp. 76-82. ird Continuous Origin of Certain Unit Characters as Observed ya EISA Harvey Soc. Volume, Tth ser., Nov., pp. af [Abstr.] OUR NUM and Allometrons in Relation to the Conception of the ‘‘Mutations of Waagen,’’ of Species, D m Phyla. Bull. Geol. Soc. of Amer., Vol. 25, No 3, pp. 16. Origin es i ngle Characters as Observed in Fossil and Living Animals and Plants. Amer. Naturalist, Vol. XLIX, No. 580, pp. 193-239. The Origin of New Adaptive Characters. Nature, Vol. 96, No. 2402, pp. 284, 285. THE EFFECTS OF ENVIRONMENT ON ANIMALS' PROFESSOR A. S. PEARSE UNIVERSITY OF WISCONSIN As Henderson? has pointed out, the environment on the surface of the earth is suited to, and largely responsible for, the existence of living organisms. After an organism comes into existenee, it strives to live in harmony with its immediate environment. An organism is a ‘‘system of aetivities'? which devotes its energies primarily to three funetions: (1) eapturing energy for and releasing energy from its own system, (2) protecting its system from injury, and (3) producing other systems of activi- ties similar to itself. If possible an organism reacts with its environment in such a way that its system continues to exist and carry on its three primary functions. It is limited in its responses to a particular behavior pattern, inherited from the system from which it came, but in general it reaets in such a way toward its environment that it selects by trial the optimum conditions for its own existence. In other words, an organism generally re- sponds in an adaptive way and selects the best environ- ment that it ean. If the behavior patterns of certain sys- tems, similar or dissimilar, are well suited to a particular environment, such systems often are ‘‘successful.’’ They may take possession of the environment, perhaps exter- minating other systems, and, thus demonstrating their **fitness,"' constitute what ecologists call a climax forma- tion. Every organism in such a group must remain a sys- tem of activities and must make continual physiologieal adjustments to keep in harmony with the environment, or it ean not continue to exist. Each organism assumes a 1 An address before the Geographical Society, University of Wisconsin, January 11, 1922 Fr" The Order of Nature,’’ Cambridge, 1917. 3 This definition is not intended to exclude the possibility that an organism may be more than matter and energy. It may contain an entelechy or some- thing similar, but as yet there is no scientific proof that it does. The limita- tion, and value, of science is that it must always deal with facts. 144 No. 643] EFFECTS OF ENVIRONMENT 145 partieular internal pattern that consists of a graded series of metabolic activities which (de Child) is a direct response to stimuli received from the environment. In responding to environment plants and animals show fundamental similarity. Many plants adjust themselves to their surroundings by assuming the form that best suits them to the particular space in which they happen to take up a sessile life, and many animals secure a place which is suited to their system of activities by moving about until they find it. This difference between plants and animals is largely due to the faet that the former usually are able to subsist on inorganic foods, wuile the latter require organic substances as a basis for their metabolic activities. However, animals often respond to the environment by assuming a particular growth form, and plants have many motile systems of activities that find favorable environments through active or passive migrations. Being trained as a zoologist and knowing little of the activities of plants, I gladly take the task as- signed to me—‘‘to discuss the effects of environment on animals’’—but I can not refrain from expressing my opinion, that there is no essential distinction in this con- nection between the two great kingdoms of life. Animals are continually active and must continually re- act with the environment. Alcock‘ said, ‘‘the three great exigencies: to find something to eat, to avoid being one’s self eaten, and to disseminate one’s species, give rise to a perpetual struggle in which the fittest are successful." The environment furnishes matter and energy to main- tain the activities of each system and a considerable quan- tity of both is neccessary. A silkworm during its short life eats food amounting to 86,000 times its own weight at the time of hatching. Animals take the most diverse mate- ials from the environment and use them to build substance or furnish energy. The clothes moth flourishes on a diet of wool, which consists entirely of keratin. From this almost pure, and to most animals wholly indigestible, pro- *** A Naturalist in Indian Seas.’’ London, 1902. 146 THE AMERICAN NATURALIST [Vor. LVI tein substance the moth makes carbohydrate, fat, and water to supply the needs of its system. The bee moth subsists on bee comb, which contains less than one per cent. of protein and a large amount of rather insoluble wax. Ants not only acquire food from the environment, but give up what they have already swallowed to their fellows, even when they are hungry themselves. In this ease the ‘‘system’’ of the colony is more important than that of the individual. In order to keep their systems of activities intact, ani- mals have adopted many means to escape dangers. There are lurking enemies, physical changes, accidents, insidi- ous parasites to be met or avoided continually. A walk- ing-stick spends nine-tenths of its life in a ‘‘perfectly quiescent’’ state, depending on being overlooked by hungry enemies. A house fly escapes through endless agility. A rotifer avoids drying up by secreting a cyst about itself, and may remain dormant for years. Many animals are able to change the usual rate of their meta- bolic activities in response to changes in temperature and pass cold periods in a hibernating state. Animals, before all things, use the means they possess in order to perpetuate their particular systems. New indi- viduals must continually be started on new life cycles and such reéreations involve reorganizations of systems, changes in metabolism, and various responses by organ- isms to the environment. Such qualities as odors, colors, and songs may be very important for the survival of a race. A male moth will migrate a mile or more to find a mate—attracted by her odor. The daily routine of seek- ing and escaping dangers is often neglected by animals when the survival of their race is concerned. Greedy penguins allow any youngster that comes to feed from their crops. An adult bull seal takes no food from May to August, but devotes all his energies to the defense of his rookery. The male gaff-tops'l catfish takes the eggs from his mate and carries them in his mouth for ninety days—denying himself food in order that his offspring No. 643] EFFECTS OF ENVIRONMENT 147 may survive. The little spider that spins a cocoon under stones guards her treasure with watchful care and, if she is compelled to leave her cocoon, spins a ground line as she runs in order that she may return without delay. A male spider dances, postures, and uses all his arts to secure a mate. As soon as he has mated, Nature usually sacrifices his life to his offspring—for his hard-earned mate devours him if she can. Thus it is wherever one considers animals. There is adjustment, frequently of a very specialized type, to en- vironment. The wonder of it all is the degree of adapta- tion that animals show. In speaking of food relations Semper® said, ‘‘there is scarcely a constituent of the earth’s crust, whether on land or in water—not an animal nor a plant, whether living, dead, or even in decomposi- tion—which does not afford nourishment to some living animal." The first more or less self-evident generaliza- tion justified by this discussion may be stated as follows: Animals are adapted to the environment. That animals are adapted, probably noone disputes, but there has been much controversy as to the means by which they have become adapted. There appear to be three effects that it is possible for the environment to produce in animals: (1) a direct transformation or modi- fication of the living system of activities, (2) the de- struction of systems-unsuited to the environment and the ‘‘survival of the fittest," and (3) the migration of sys- tems from unfavorable to favorable environments. Animals are modified by external changes and may even take on different forms to suit different environments. Sponges and corals growing in deep water usually have a branching form; the same species in shallow water form flat, encrusting growths. The brine-shrimp, Artemia salina, is a classical instance of an animal that has many forms, and these are rather closely correlated with the salinity of the water in which it lives. Sumner® and Shel- 5° Animal Life as Affected by the Natural Conditions of Existence,’’ N. Y., 1881. 6 Bull. U. S. Bureau of Fisheries, 1910, 148 THE AMERICAN NATURALIST [Vor. LVI ford,’ working independently, have shown that very slight structural differences that distinguish closely related species of amphipods and tiger-beetles are correlated with distinct habitat preferences. The structure and physiology of animals are modified by environment—the structures and activities of the systems are changed. Different species may possess almost identical structures, but show specificities of behavior in relation to environ- ment. Darwin made much of the struggle for existence among animals, pointing out that many species hold their places on the earth through wide dissemination and selective survival. One who has seen the strangler trees gaining a foothold in the tropical forest, the fiddler crabs fighting to hold a favorable place on an ocean beach, or the oysters in an overplanted area striving to survive, can not doubt that there is such a struggle. More animals are produced than can find a place to exist, and in general those survive that are best suited to the environment that is available. Animals are not always obliged to adjust themselves to the environment or struggle for a favorable place to live init. They migrate from situations where their sys- tems can not well carry on activities to some spot where conditions are more propitious. In such migrations ani- mals have very definite relations to the environment. They are limited by their reaction pattern to certain habitats; they must disperse from their ‘‘centers of origin" through ‘‘highways,’’ and are prohibited from migration into certain regions called ‘‘barriers.’’ Bar- riers are areas where certain environmental factors vary beyond the limit of toleration for a species. A ‘‘center of origin” as usually understood by geographers, may be the real place of origin of a species or it may merely rep- resent the locality where the most environmental factors are favorable. In general a uniform environment cover- ing a wide range of territory permits the species suited . to such an environment to have a wide geographic range. 7 Biol, Bull., 1911. No. 643] EFFECTS OF ENVIRONMENT 149 Variable species usually have wider ranges than unvary- ing, because they ean adapt themselves to more environ- mental variations. A second generalization is appropriate here: Animals become adapted to environment by (1) transformation, (2) selective survival from an overpopulated condition, (3) migration from unfavorable to favorable situations. It will be profitable now to examine two or three typical associations in order to study animals in action with the environment. If, in this connection, one thinks over the great responses that animals have made to environment in the past, he will probably conclude that the greatest habitat change has been that from water to land. It is generally supposed that life first appeared in water. As a habitat, water has certain inherent advan- tages—the chief of which is perhaps the slowness with which temperatures change. It also has certain disad- vantages, the most important of which are probably the variability of its dissolved gases (the higher the tempera- ture, the less gas can be held in solution) and its general solving power, which makes it a transporting medium for all sorts of substances, some of which are poisonous. All animals require a more or less constant supply of water and of oxygen for metabolic processes. When ani- mals forsook the water for land habitats, they gave up surety of water supply and conditions of reasonable thermal stability. What did they get in return? Ap- parently nothing but a stable gaseous condition for re- spiratory needs. The danger of desiccation and the wide variations of temperatures incident to land life were ap- parently compensated for by this gaseous stability. Yet the attractions of the water have at times led many ani- mals, like the aquatic insects, that had become adjusted to life on land to revert to aquatic habitats. In the past races have doubtless many times become adapted by transformation, selection or migration on account of the advantages or disadvantages of one of two habits. If one walks along a rocky shore, where the ocean 150 THE AMERICAN NATURALIST [Vor. LVI waves and tides sweep, he may be surprised to find an abundant fauna in the ** ’tween-tide " zone. The moving water, teeming with mieroseopie organisms, brings an abundance of food to those animals that are able to stand the beating of the waves and the alternate submergence and exposure due to the ebb and flow of the tides. A roeky wall along the sea shore is no plaee for weaklings. One minute the blistering sun bakes the exposed animals; ` the next, the rising tide has covered them with cold water. The waves beat ceaselessly. The changing seasons bring ice and torrid heat. How are the animals on these rocky shores responding to the environment? Here one finds a variety of hardy species which, though not closely re- lated genetically, have many characteristics in common. There are sponges, anemones, hydroid colonies, barnacles, mussels, snails, small crustaceans, and a few scavenger crabs. These animals for the most part obtain their food by net fishing or by straining water through their bodies. They are mostly attached firmly to the rocks, and thus withstand the violent movements of the food-laden water. The barnacles, sponges, and hydroids are grown fast; the anemones and snails have sucking dises that enable them to adhere firmly; the crustaceans have claws for attach- ment and hard armor covering their bodies. Some of these animals are small and can easily hide in crevices; some of those of larger size, like crabs, are able to migrate to other habitats during violent storms. If an animal is attacked, it is advantageous for it to be able to receive stimuli with facility from all directions of the compass, and, as would be expected, many of the animals on rock beaches are radially symmetrical. Radial symmetry has marked advantages for sessile animals, but puts a weighty limitation on psychic development. An animal that is able to perceive stimuli equally well, through equally efficient sets of sense organs that are symmetrical- ly disposed about a central axis, is never able to develop its power of paying attention to any considerable degree. Its simple mind, if such an animal may be said to have No. 643] EFFECTS OF ENVIRONMENT 151 a mind, must attend now to a stimulus received from one side, now to that on another. Such vacillation is not con- ducive to the development of higher types of mental life through the delegation of psychic authority to one nerv- ous center. The rocky ocean shores, then, put a pre- mium on radial symmetry and thus as an environment tend to foster psychically unprogressive animals. The barnacles, that appear to have come from progressive, bilaterally symmetrieal ancestors, have become degraded with the taking on of the sessile life and radial symmetry that suits them so well to wave-beaten shores. | : The ebb and flow of ocean tides have a pronounced effect on shore animals. "Those speeies that are not able to survive alternating exposure to the desiccating effects of air of varying temperatures and the activity of vio- lently moving water of rather constant temperature ean not exist on rocky shores. This fauna must be resistant, and is so. An anemone can be kept out of water for a week—until it looks like a dried raisin; or kept in a tightly eorked bottle for ten days, and when replaced in the ocean appear to be perfectly normal in an hour or two. Such an animal will not readily succumb to the exposure between tides or even to the stagnation likely to oeeur in a beaeh pool that is eut off from the ocean during a prolonged period of low water. The barnacles and molluses on rocky shores are protected by heavy cal- eareous shells. Flattely? has suggested that land animals perhaps arose in the past on ocean beaches as a result of _ the resistance developed during exposure between tides. As a whole the environment occurring on rock beaches offers abundant food, but hard conditions for life. The fauna is highly adapted to resist the two important en- vironmental faetors —moving water and exposure to varı- able conditions— and in this adaptation the fauna has in- cidentally but of necessity become unprogressive and de- votes most of its activities (1) to feeding rapidly when the opportunity comes, (2) to resisting, (3) to resting. * Science Progress, 1921. 152 THE AMERICAN NATURALIST [Yon LVI One does not imagine such a fauna as developing, even through countless ages, great appreciation of beauty, or of any of the esthetic qualities of ‘‘higher’’ animals. The adaptations here are to resist the unfavorable in the evironment, and still live. If a man walks in a tropieal forest, he is amazed at the abundanee and variety of the life about him. He may see a certain species of tree in one spot and not encounter another like it for a mile. Meanwhile he has seen a hun- dred other species of trees. What is the striking en- vironmental faetor in this forest? It is life itself! "The environment is favorable for so many systems of meta-: bolie aetivity that hundreds of kinds of animals are ready to live in it—if they can find a place. Is there a struggle; is there adaptation? Nowhere on the earth are these responses to environment more striking. Most of the struggles to live in the forest are competitions with other living systems that are trying to continue to exist. And the adaptations are not often for resisting, for eat- ing, for resting. Think of all the animals in the tropical forest—is there one that is radially symmetrieal? Here keen senses are at a premium. Life has always depended upon seeing, hearing, feeling better than something else. Lately it has come to depend upon thinking better than something else. And the climax of adaptation in this tropieal forest has been the greatest thinker of the ages. A third generalization appears to be justified: Each habitat, representing environment, limits the patterns of the systems of activities that may persist from reactions within it. The type of adaptation is set by the environ- ment. ` Environment has a quality that any system of activi- ties that attempts? to live in it must respond to. This is its changefulness. The paleontologists say that environ- ment punishes too much adaptation by ehanging. I think it is proper to say that the chief cause assigned for the 9 The writer realizes that ** attempts ’’ may be pus A ger as teleologieal —and rejoices in the sinfulness of it. If an organism does anything, it strives to keep on existing. As far asit possesses means, it responds to what- ever interferes with living No. 643] EFFECTS OF ENVIRONMENT 153 dying out of extinct types of animals is ‘‘over-adapta- tion,’’ or better ‘‘too much specialization." A system of activities, as represented by an organism, can not depend absolutely on another system of activities, as represented by environment. The organism changes and the environ- ment changes too. If the environment continues for a time in a fairly stable condition, an animal may become adapted to it to such a degree that, if the environment then does change, the animal ean not respond enough to continue to live. The wood frogs in the United States breed when the water is at freezing temperatures ; frogs, belonging to the same genus as the wood frog, that live in Cuba die when the temperature falls below seven degrees. These frogs are adapted to different environments and those in Cuba will be in greater danger of extinetion if there is a prolonged cold period. There is a general tendeney among animals to find sue- cess during conditions of stability. Certain arthropods left the water and attained stable respiratory conditions and freedom from water-soluble poisons by going on land. Later, certain of these arthropods again gained a thermally stable environment in the water and continued to enjoy a stable gaseous environment by carrying air into the water with them. When any raee of animals attains a stable environment, it may become specialized to it. We see a manifestation of the same type in the psychology of man. It is ‘‘human nature” to desire stability—to be free from care and worry; to know where one stands. On the other hand, continual change is a stimulus to progressive response—in fact, one is tempted to say that laek of change is injurious to living organisms and that changes often stimulate living systems to renewed activ- ity. Payne kept fruit flies continuously in the dark; Calkins and Woodruff maintained protozoans on unvary- ing culture media. All these investigators agreed that lack of variation in the environment was injurious. This raises a dilemma—on one hand animals tend to become highly adapted (or specialized) when the environment is 154 THE AMERICAN NATURALIST [Vor. LVI stable, and on the other hand a changing environment is a stimulus to progressive changes in organisms. A few animals have lived for ages in a stable environment. Thompson cites the brachiopod, Ligula, as a ‘‘ supreme instance of static racial inertia." However, most animals must live in environments that change. How do these respond? It is a matter of common knowledge that animal sys- tems of activities can become adapted to changes in the environment, even when such changes constitute new racial experiences. By taking increasing doses of certain poisons at regular intervals animals develop enough im- munity to be able to take daily a dose which in the begin- ning would have been fatal. If a pigeon is fed nothing but meat the lining of its stomach changes its character and the bird’s metabolic activities become adapted to an unusual diet. Many other instances of acclimatization . to new conditions might be cited. Every physiographer knows that earth environments change by succession. Land forms erode and water forms fill up with sediment. Physiographie succession brings about a suecession of environments, or habitats. "These are successively occupied by different groups of plants and animals and there is thus an ecological succession, which is a succession of species or groups of species. Shelford has worked out excellent examples of ecological succession in the streams and ponds along the shore of Lake Michigan. Pioneer species of animals invade hab- itats soon after they are formed, and as the habitats change the pioneer species are succeeded by others that are adapted to later stages in physiographie succession. Ecologie succession is a succession of species; animals do not change as the environment changes, but die or migrate to more favorable localities. Animals do not appear to have special means for adapting themselves to such changes. There are other types of succession, however, to which animals show striking adaptation. The types are all rhythmic (seasonal, monthly and daily) and depend pri- No. 643] EFFECTS OF ENVIRONMENT 155 marily upon the motion of the earth and moon. As the earth makes its annual journey around the sun, the ani- mals of temperate and polar regions, and to a less extent those in the tropics, are subjected to seasonal changes in environment. These changes are related chiefly to tem- perature, available moisture, and food. Animals gen- erally respond to such environmental variations by adjusting appropriate activities to favorable times. In general winter is a season for resting; spring, for mating and propagation; summer, for feeding and growth; and autumn for fructification. Seasonal succession is a succes- sion of stages in life cycles. The seasonal rhythm has a short enough period to permit animals to become adapted to it. Their systems of activities vary to fit the seasons. Every one is familiar with the seasonal migrations ot animals. The arctic tern travels from pole to pole, and thus always lives in sunshine. Many animals do not migrate, but pass the winter in a dormant condition. In the tropics animals frequently estivate during the annual dry season. Now many of these seasonal responses are certainly due to stimuli received from the environment. The little Daphnias, that live in fresh-water habitats the world over, usually have long helmets in summer and short helmets in winter, but long-helmeted forms can be made to produce short-helmeted offspring in summer by keeping them at low temperatures. In this instance the effective stimulus appears to be thermal in nature. But animals are adapted to seasonal succession beyond merely responding as far as they are able to stimuli that come with rhythmic changes without their bodies. The living system apparently has arhythm of its own that is adapted to the seasons. Smallwood” kept a female dogfish (Amia calva Linneus) in an aquarium, practically without food, for twenty months at rather constant temperature. Dur- ing this time the fish twice took on its bright nuptial coloration. Another instance of similar nature has come to the notice of the writer. A tame spermophile, Citellus tridecimlineatus (Mitehill), was kept in a steam-heated 10 Biol. Bull., 1916. | 156 THE AMERICAN NATURALIST [Vor. LVI house for four years. In the autumn of the first year it became very fat and stored a large quantity of food in its burrow. About December 1, it went into its burrow, closed the opening, and remained underground for 119 days. The following autumn the spermophile behaved in a similar way but remained underground for only 28 days. It did not hibernate during the two years following. This animal had an established seasonal metabolic rhythm that was correlated with seasonal environmental changes, but the rhythm had a physiological basis for it persisted when appropriate environmental stimuli were not present. The rotation of the moon about the earth introduces certain rhythmic variations into the earth environment to which animals respond in adaptive ways. Such responses are of course not due directly to the moon as such, but to effects of the moon’s motion on matter belonging to the earth. The famous Palolo worm and various other ma- rine annelids come from their hiding places to spawn only during certain phases of the moon. In these worms the eggs do not ripen except when the moon is new or full; the internal activities respond to outside changes, chiefly referable to tidal variations, and a physiological rhythm is established. The earth rotates on its axis and thus the animals on its surface are subjected to alternating light and dark. Animals readily respond to this short-period rhythmical change. Every one is familiar with nocturnal and diurnal animals. They are adapted to rhythmical environmental changes to such a degree that they may keep on respond- ing periodically when the environment does not change. Keeble and Gamble" have described an interesting shrimp (Hippolyte varians) that has day and night color phases. During the day this shrimp matches the background on which it rests with a high degree of accuracy, assuming quite a variety of colors and patterns. At night it turns green, regardless of its background. When kept contin- uously in light it undergoes rhythmic color changes at about the time periods that correspond to day and night 11 Phil, Trans., London, 1904. No. 643] EFFECTS OF ENVIRONMENT 157 for two days; and makes similar changes in the absence of light for about a week. There is a physiological rhythm that corresponds to periodic environmental changes. A fourth generalization must again relate chiefly to adaptation—Though animals possess considerable power of adjustment to new or changed factors in their environ- ment, they apparently do not usually become adapted as species to physiographic changes, but are eliminated by the variation of factors beyond their limit of toleration. One species or group of species succeeds another during physiographic succession. However, animals do respond in an adaptive way to rhythmical daily, monthly and seasonal successions. Some animals show adaptive re- sponses to rhythmical environmental changes only once during their life cycle. Salmon, for example, do not migrate up rivers to spawn until they have reached a certain age. Animals apparently become most special- ized, or adapted to particular environments, when con- ditions are most stable. Even the striking instances of adaptations to rhythms show this tendency of adaptation to attain stability—in this case a regularly changing sta- bility. Environmental changes have been important in their effects on the evolution of animals. In this paper it has been shown that living systems of activities are adapted to the environment ; that they respond to the environment by transformation, selective survival, or migration; that each habitat limits the patterns of the systems that exist within it; and, that, though adaptation to environment may permit precise adjustment to rhythmical changes extending over considerable periods, and though animals generally become most specialized when conditions are most stable, there is no evidence that living systems are caused to change from one species to another by the transformations of habitats due to physiographic suc- cession. The pattern of evolution is set by environment, but there is little or no evidence that changing envi- ronment causes adaptive variations of such a degree that new species are produced. Animals adapt them- 158 THE AMERICAN NATURALIST [Vor. LVI selves to environment by changing their systems of ac- tivities, but such responses are apparently limited in extent to the inherent possibilities of variation already within the system. Animals have great powers of adapta- tion to environment, but are not fundamentlly changed by it. Environment permits evolution and controls its course, but does not appear to cause it. If variations fit environment, they are adaptive; if they do not, systems cease to exist. Environment does not appear to cause variation. The living mechanism still holds the mystery of variation within itself. Until there is conclusive evi- dence, this one great remaining problem of evolution ean not be solved. Yet, notwithstanding this lack of evi- dence, there are still those who belive the environment does cause evolution—though their only foundation for such belief is what Bergson calls ‘‘ intuition.’’ Until there is proof, science, if it would be scientific, must keep in mind that these ‘‘ faithful " believers may be right, and be content to wait, perhaps a hundred thousand years —for evidence. SUMMARY l. Animals are systems of activities that are adapted to environment. 2. Animals become adapted to the environment by transformation, selective survival, migration. 3. Each habitat limits the patterns of systems of ac- tivities that may result from reactions within it. The type of adaptation is set by the environment. 4. Though animals possess considerable power of ad- justment to changes in environment, there is no evidence that they became adapted as species to slow changes due to physiographic succession. They do respond to rhyth- | mical daily, monthly, and seasonal changes in an adaptive way. Animals appear to become most specialized, or adapted to partieular environments, when eonditions are most stable. 9. Environment permits and directs evolution, but does not appear to cause it by forcing the acquirement of new characters. A SUMMARY OF THE FOOD HABITS OF NORTH AMERICAN COLEOPTERA HARRY B. WEISS New JERSEY DEPARTMENT OF AGRICULTURE Tug Coleoptera or beetles contain a very large number of species and show a great diversity of habits. Most of them are terrestrial and they live under almost all con- ditions where insect life is possible. The economic status of this group of insects is important. To the Coleoptera belong some of our most pernicious agricultural pests as, for example, the cotton boll weevil, which has caused such ruin in the cotton belt, the Colorado potato beetle with its familiar destructive activities and various other species which attack forests and field crops with varying degrees of intensity. However, many species of Coleop- tera are engaged in useful activities and it is the purpose of this paper to summarize briefly, and in a very general way, the food habits of the families in this order. _ For the purpose of convenience in handling and for the sake of simplicity, the families have been grouped into a few important classes and the placing of each family was based mainly on the predominating larval activities of its members. In some families considerable variation occurs in the food habits of the different spe- cies. For instance, in the Scarabeide, some are destruc- tive to green vegetation and others thrive on vegetable decay. On the whole, however, their activities are saprophytic and for this reason the entire family was placed in the group Saprophaga. The Staphylinide were placed in this group also, although this family contains members which live in fungi, in animal and vegetable decay, in the nests of ants and some which are predatory. In quite a few of the families, the activities of the species are practically identical. 159 TOO es THE AMERICAN NATURALIST [Vor. LVI The classes into which the families are grouped are as follows: Phytophaga, Saprophaga and Harpactophaga. In addition to these three important ones, the species attacking mammals and those whose family habits are N 5 © © a 26 per cent E Y a =} © < 44 per cent B 100 PER CENT 2 a ca PLEET d ` B G - L^ 27 per cent E o : rA N n aS $ per cent Diagram illustrating the comparative abundance ke the various types of food-habits in the Coleopte obscure have been grouped separately. In the Phytoph- aga have been placed those species which feed upon the higher plants. In the Saprophaga will be found those forms which feed for the most part upon disor- ganized tissue, vegetable and animal decay and such No. 643] FOOD HABITS OF COLEOPTERA 161 species which remove or change the form of animal and vegetable remains and aid in reducing such substances into shape for assimilation by plants. While not strictly belonging to this group, species feeding on low forms of plants such as fungi and those living on dry vegetable and animal matter have been included for the sake of convenience and in order to avoid numerous subdivisions. In other words, the term Saprophaga is used in a very broad sense. poe The Harpactophaga contains the predacious and car- nivorous species, of which there are a great number, and whose aetivities help to preserve a natural balance be- tween certain groups. Many of them are general feeders, appearing to be not partieular whether their prey is a plant feeder or another predatory form. However, in some families, such as the Coccinellide, there is a decided specialization as to the prey, and such a group is very often an important specific check to unusual increases in the numbers of plant lice. The Coleoptera attacking living mammals are few in number. The species in the family Platypsyllide consists of a wingless beetle found on beavers. In the Leptinide, the species have been found in the nests of field mice and bumble-bees, but their exact habits are somewhat obscure. It has been suggested that the bumble-bee nest is the natural home and that the field mice afford transportation from one nest to another. The last group is made up of those families of which little or nothing appears to be known concerning their food habits. While this same lack of information is true for a large number of individual species placed in the other groups, yet enough is known of their general family habits so that little risk is run in placing them as family units. This, however, could not be done with any cer- tainty in the case of the last class and they are presented simply as a group difficult to classify from a food stand- point. The following tables show the name of each family, the number of species in that family described up to and in- 162 THE AMERICAN NATURALIST [Vor. LVI cluding 1918 and a brief statement indicating the more important food habits. The information in the first two columns was compiled from the recently issued ‘‘ Cata- logue of the Coleoptera of America North of Mexico ”’ by C. W. Leng. Family No. Species Habits Lymexylide .......... Bore in hard wood. Buprestide .......... 3 Wood borers in healthy and unhealthy trees. erambycid® ......... 1,1123 Borers in dead, dying and healthy trees and plants somelide ,........ 974 Feeders on vegetable tissue. MISES a eras 93 In men TOMEI saech esan 6 ood. MN Ie 1 Like Drait Cureulionide ......... 1,839 Feeders e vegetable tissue, Platypodide .......... Boring in epi? bic u.s. 379 Borers in od healthy and sick trees. SAPROPHAGA ily No. Species Habits Büphide aissi eunis? 137 Scavengers in dead animal and vegetable . matter, in fungi. Ones bidet ceci sarisi, 6 Same as above, Orthoperide ......... 57 In decaying €—9À under bark, ete. Staphylinide ........ 2,748 Varied, in ants’ nests, in fungi, in decaying animal and PSEk matter, ete., preda- tory. Pselaphidsm .......... 355 Varied, in ants’ nests, under vegetable de- cay, in wet moss, in rotten stumps, ete. Clavigeride ......... 7 Same as above Pude iser ET 83 In T vegetable matter, excrement, fun ^ HIS olives 3 In uus decay. Seaphidiide ......... 90 In rotten wood, fungi Spheritide .......... 1 Same as Silp zs cere 4 Under bark, in dry wood, iod be pm Cidemeride ......... 49 In timber cast up by se rdelide ...... v 142 Vows adults on flow mu larve in dead ood, fungi, stems of live plants. IAO soei 17 te mber. Pyroehroldg ......... 11 Mts bark of tree stumps. upende ons soe cd 39 In dead wood, Cerophytide ......... 2 Probably like those of Elateride. Cebrionide .......... 9 Probably like those of E aterida. MEN iioii. 976 In decaying wood, in soil on roots of grasses, PHYTOPHAGA No. 643] FOOD HABITS OF COLEOPTERA 163 Melde oes. 57 In dead tree Throscdade ciwon: 25 Like those of Elaterida. Hide i oss 29 On roots of aquatics, in fungi. Dermestidm .......... 129 In dried animal matter. Ini ays ch 64 Varied, under bark, in granaries, in fungi, predaceous, Nitidulidae oe eek 132 Sap beetles, on flowers, predaeeous. Rhizophagide ........ 14 Probably li M Monotomide ........ 36 In ants’ nests, probably have no relations with ants Jrotyhdb os aka Gis 71 Mainly in füngl: Cryptophagide ...... 135 In fungi and aaen zie vegetable matter. Mycetophagide ...... 32 Under bark, in fungi. Colydüdi- 605. cis: 84 In fungus covered wood, LathrdHde |... 104 In fungi. Mycetwide .......... 4 In fungi. Endomychide ........ 34 In fungi. Phalaeénde seleeni 117 Under bark, on flow Aleéenhid ...ss 5.058 124 Larve in rotten ice adults on leaves, flower Tenebrionide ........ 1,139 In dry vegetable sae fungi. Lagrüdb .. ics.. 17 Probably like Melandryide ........ 81 In dry wood, nn. Fin ilo 37 In dry onam and vegetable matter, wood, drugs, hobide ss 233 In dry ord matter. aae spi eis 01 In d. yetidi 215r 16 In dry wood. sphindite s Verus 6 In fungi. SB .. eee eta us 85 In fungi. éatabaiijie E 996 Varied, in decaying vegetation, on roots of plants, on green vegetation bte sane REA ERA 30 In deeaying wood, sald no ee eens 2 In decaying wood. Pinkie umm 62 On dead wood, in fungi. HARPACTOPHAGA Family No. Species Habits Cincindelide ........ 114 Predaceous. RD eu Nes 2,165 Predaeeous. Omophronide ........ 15 Predaceous. Halplide ...... e 41 Aquatic, predaceous, Dytissidte aee nnn 333 Aquatic, predaceous, Mire oU cis ce esse 41 quatie, predaceous, drophilide ........ 190 Predaceous. Secydmenide ........ 174 Feeding on aeari, in ants’ nests. Histor esee oe 384 Found in same situations as scavengers but probably predaceous. 164 THE AMERICAN NATURALIST [Vor. LVI byecide :.. cece ees 50 Carnivorous as larve. Lampyride .......... 52 lLarve carnivorous, adults on flowers. Phengodidm 2... vcs oes 23 Larve carnivorous, adults on flowers. Cantharide ......... 155 Larve vun adults on vegetation. BB i Se eae a 321 Predaceou ts flowers Omni ie ids 181 Se Acces adults on flowers, oe Can SUE T 38 Predaceous. ephaloide .......... 8 Probably similar to those of — mow sos eves 26 Larve in asitie in ants’ nests, on coc roac POM A es 227 Larve PPR adults on green vegeta- tion. Cuna so soari 85 Varied, under bark, predacious, in stored roduets. Coeeinellide ... í .... 862 Predaceous. ANIMAL PARASITES No. Species Habits Piatypsyllide ea as 1 Animal para: KOPRAD Sock ck ne es s,s 3 Probably uen on mammals. Foop HABITS OBSCURE Family No. Species Habits Amphizoidme ......... 2 Aquatic. Brathinidw .......;.. 3 Telegeuside ........ 1 Micromalthide ....... 1 Eurystethide ET se 3 Mi sais pes 54 Probably like those of Anthicide. Anthieide ....... ^... 191 On surface of earth like ground beetles. Plastoeeride ........ 19 hipieéeride isss 6 In wood, Psephenide ......... 4 Semi-aquatic. Dryopidibb viis P. 17 Aquatic. Helle ci o 36 Aquatic. Heteroceridm ..i..... 11 Semi- T e Ory ser : 2 In wet plae lodids$ os oe ss ves 32 Probably ae treno ES I Byrrsthide ess 97 Habits obseure, on ground beneath cover, about grass roots. Riysodidë oe ea 4 Derodontide ........ 5 Hs vee c cus 5 ORIGIN oo esses 1 Monommide SU EL es 6 No. 643] FOOD HABITS OF COLEOPTERA 165 SUMMARY No. Species Per Cent. of Tota Phytophaga o doses CER Yel 4,801 26 Baprophapa orse n a Nn eee 8,252 44 Harpactophaga cser osa S£ EXER 4,985 27 Animal parasites ...... 4 eek Res 4 Food habits obscure ;.......,. 7 e 501 3 18,543 100 About 26 per cent. of the species of Coleoptera are phy- tophagous, most of this pereentage being made up of the families Curculionide, Cerambycide and Chrysomelide. Almost one half of the species of beetles, or 44 per cent., appears to be saprophagous for the most part and im this group the families Staphylinide, Tenebrionide and Scarabeide supply over half of the species. In the pre- daceous group, consisting of 27 per eent. of the total, the Carabide with its 2,165 species is the largest single con- tributor. Thus almost three fourths of the species of beetles in North America are apparently engaged in what we call useful activities. INDIRECT EVIDENCE FROM DUPLEX HYBRIDS BEARING UPON THE NUMBER AND DIS- TRIBUTION OF GROWTH FACTORS IN THE CHROMOSOMES DR. D. F. JONES Connecticut AGRICULTURAL EXPERIMENT STATION, New Haven Surricrent evidence has accumulated to indicate that the main features of the chromosome theory of hereditary transmission, as worked out for Drosophila, are appli- cable.to plants. Peas, primula and maize have been the best materials so far to demonstrate linkage of factors in plants. Owing to the ease of culture, large number of seeds produced and the great genetic variability the maize plant is becoming very useful in this line of in- vestigation. The agricultural importance of the plant and the large number of people working with it have al- ready made the list of Mendelian factors definitely de- termined large and increasing rapidly. Due mainly to the industry of Professor Emerson and his co-workers at Cornell University, six linked groups are already visible in rough outline, some of which have a goodly number of factors fairly well located. It therefore seems pertinent to consider some indirect evidence funished by this plant having a bearing upon the chromosome mechanism. In working out the best means of utilizing inbred strains of corn for the purpose of increasing production it has been found to be advantageous to cross again two different first generation hybrids each of which were the result of combining two different self-fertilized families. Altogether four homozygous types, each differing from the other in many visible characters, are brought together in this way in a progeny which has an.extremely complex composition. Assuming that the inbred strains have been reduced to complete homozygosity, the first generation hybrid is uniform. Statistical measurements show this 166 No. 643] GROWTH IN CHROMOSOMES ` 167 to be so. Theoretically, all the plants are hereditarily exactly alike. When such a hybrid with segregating gametes is again crossed with a similar first generation hybrid but having a different genetic construction, the result is a mixed lot of plants in which practically every individual differs in some degree germinally from every other. This statement holds for any numbers that it would be possible to grow. Every inbred strain of maize, that has so far been obtained by continued self-fertilization with one progenitor in each generation, has differed in many ways from every other inbred line, whether they came originally from the same or different varieties. All the inbred strains coming from different individuals at the start show a noticeable increase in vigor when crossed and a rapid reduction of growth and great increase in variability in the immediately following generations when again self-fertilized. It is therefore not at all improbable that most of the self-fertilized strains differ from each other by a large number of genes in every chromosome. If such is the case, then the duplex combination will have an extraordinary amount of genetic diversity. This may be made clearer in the following illustration. If, instead of being crossed, a hybrid was self-fertilized and there was only one factor difference in each pair of ehromosomes, over one million plants would have to be grown in order to have an even chance of securing all the possible combinations (assuming maize to have 10 chromosomes). But with more than one factor in each chromosome the situation is far different. Two factors in each chromosome having a linkage ratio of 10 per cent. would necessitate 20?' individuals in the segregating generations to obtain the same result. This is calculated - from the formula [2(r 4- 1)"?]?* where r+ 1 is the link- age ratio, in this ease 10 per cent., or 9+1, n is the number of factors in each chromosome, and c is the num- ber of chromosome pairs. This number of plants to be grown would require an area roughly 57,346 million times the total surface of the earth. But instead of being 168 THE AMERICAN NATURALIST [Vor. LVI self-fertilized, the hybrids with their segregating gametes are again erossed and certainly there are more than two factor differences in most of the chromosomes having varying degrees of linkage with each other. At present almost nothing is known about the heredity which the two first generation hybrids may have in common. But all the four homozygous types when erossed singly in the six possible combinations show about an equal amount of heterosis. The double-erossed combination shows no re- duction in vigor of growth, but on the other hand this appreciably increased. This is due, however, in part to a better start as the plants come from large, well-nour- ished seeds grown on vigorous plants, whereas the first cross is handicapped in this respect. The doubly hybrid plants are theoretically more di- verse than self-fertilized second generation progenies coming from the same parents, but compared with the first and second generations the double cross has features of both. In respect to growth characters the plants are a group of many different first generation hybrids. Very little recombination can take place to allow recessive weaknesses to appear. In fact any recombination that does take place is probably out-balanced by an increase in heterozygosity in other factors. A critical comparison of such double hybrids with their parental first genera- tion hybrids and with their second generation self-fertil- ized sibs in respect to variability of different characters ought to give some indication of the distribution in the chromosomes of the hereditary factors affecting growth. In those factors which are independent of the growth of the plant the variability of the double cross should approach or exceed that of self-fertilized second genera- tion. In those characters which are directly dependent upon the vigor of the plant the double cross should re- semble more closely the first hybrid generation. Five characters have been taken and measured in three different but similar lots of plants. These are: number of rows of grain on the ear (pistillate inflorescence) ; nodes of plant, height of plant, length of ear; and pro- No. 643] GROWTH IN CHROMOSOMES 169 duction of grain (weight of entire pistillate inflorescence with mature seeds). A previous study of a large number of first generation hybrids between inbred strains of maize has shown that the average number of rows of grain of the hybrids was inereased 5.29 per cent. above the mean position of their parents; similarly nodes per plant 6.45 per cent.; height of plant 27.44 per cent.; length of ear 28.57 ; and total produetion of grain 180.00 per eent. The variability of these F, plants was slightly decreased below the parental average in nearly every ease in respect to these characters. Rows of grain and nodes are therefore much less in- flueneed by the vigor of the plant than are the other characters, notably production of seeds, which is very largely determined by the amount and rapidity of growth. Assuming that the complementary aetion of dominant favorable growth factors is responsible for the vigorous growth of the hybrids, it would be expected that F, x F, combination would not be as variable as the second gen- eration resulting from self-fertilization in respect to produetion of grain per plant, provided a large number of essential growth factors were acting and that these were distributed rather uniformly throughout the chro- mosomes. On the other hand such characters as rows of grain on the ear and nodes per plant being largely independent of growth vigor, would not be expected to show a reduction in variability when-compared with the second self-fertilized generation. The distribution and statistical constants for the sec- ond generations grown from self-fertilized seed of the parental hybrids have been compared to the reciprocal erosses of the same parental hybrids in three different sets of plants. In each ease the cross-fertilized seed, which produced the F, X F, plants, and the self-fertilized seed, from which the F, plants were grown, came from the same ears. The two kinds of pollen were applied in a mixture at one time and the seeds separated by their color at maturity. Without giving the extensive data upon which the 170 THE AMERICAN NATURALIST [Vor. LVI figures are based, the averages of the coefficients of vari- ability of the F, X F, and the F, families are brought together in table 1. With these are given some figures averaged from the F, parents. These are not from the exact first generation parents of the progenies used to TABLE I A COMPARISON IN VARIABILITY OF SINGLE First GENERATION, DOUBLE First GENERATION AND SECOND GENERATION HYBRIDS Characters | Rows of grain...... | 8.90 40- .98 | 12.76 .D2— .65 | 12.381 45- .82 Nodes of plant..... | 5.54 25- .73 5.88 .25- .27 6.20 26- .33 Height of roti d 708 .25-1.04 6.20 .25- .33 6.92 26- ,40 Length of ear...... |.- 13.83 ./2-1.66 | 13.23 .41- .75 | 16.90 .69-1.03 Weight of ndn SM | 24.13 |1.15-1.42 | 26.99 |1.19-1.44 | 32.68 /1.11-2.31 give the other results in Table I. They are similar but were not grown in the same years. They can not be compared as closely to the F, X F, and F, lots as these can be compared with each other. The coefficients for variability of the F, x F, and the F, plants, averaged from three different combinations with a fairly large number of plants in each grown from seed of which the two contrasted kinds came from the same ears, are strictly comparable. The greater growth of the double hybrids as shown by the increase of the means makes comparison of the coefficients of variability The appearance of the plants in the field apports the statis- tieal data, as it is the uniform produetion whieh makes the hybrid plants so valuable for agricultural purposes. There is a noticeable difference between the double cross and the self-fertilized second generation in even size, similar appearance and general excellence as the plants are harvested in the field. The figures show that the variability of the F, X F, families is about the same as the F, families in rows of grain and nodes per plant. In height of plant, length of ear and weight of grain per plant, all characters which No. 643] GROWTH IN CHROMOSOMES 171 are markedly influenced by the vigor of the plants, there is a reduction in variability. Partieularly is this true of length of ear and weight of grain, which are fairly reliable measures of the plant's reproduetive ability, whieh in annual plants sums up the organism's entire energy. In other words the plants are uniformly vigor- ous and are not dependent upon exceptional individuals for their high average position. This is indirect evidence that those hereditary factors which are concerned with the growth of the plants are numerous and widely dis- tributed throughout all or many of the chromosomes. As a means of corn improvement it would be highly desirable to bring together into a pure breeding homo- zygous condition all those factors which cause the hybrid plants to excel their parents. Such individuals should be even more efficient in their growth processes than the heterozygous combinations of the same factors because the determiners responsible for hybrid vigor seldom show complete dominance. The recombination of linked fac- tors is a problem that demands the most careful attention of the plant and animal breeder. It is the closely linked factors which are the main concern. When the distance between any two loci is fifty units or more, then all the factors situated outside of these points are independent of each other in transmission and it makes no difference from the standpoint of recombination whether the factors are in the same or different chromosomes. Therefore the number and arrangement of the individual genes themselves seem to be more important than the number of chromosomes. Although as yet it is impossible to com- pare the numbers of factors in different species, it does not seem likely that the genus Rosa with 8 chromo- some pairs is genetically less complex than Nicotiana species with 24 pairs. Some crustacean species with 84 pairs of chromosomes are contrasted with various mammals with 8 to 12. Even in the Arthropods alone the haploid number ranges from 2 to 100. It seems profitless to look for any significance in chromosome, numbers. Leaving aside the matter of doubling of chromosomes any 172 THE AMERICAN NATURALIST [Vor. LVI differences that there may be are probably qualitative rather than quantitative. It is possible that there may be very little difference in the amount of essential hereditary material. But the word ‘‘ amount " must be considered as equivalent portions. The visible size of the chro- matin mass fluctuates greatly even at different stages of growth in the same individual. Although the cytological proof of the chromosome theory is still so meager as to make speculation somewhat useless, nevertheless, looking at the matter from the stand- point of diffieulty of recombination the important consid- eration is the number of fifty-unit lengths of chromo- somes. However, the Morgan school unit of measure- ment, the one per cent. of erossing over, is not a stable unit, as they have shown that crossing over fluctuates rather disconcertingly, due both to environmental and ger- minal modifying factors. Detlefsen' finds the rate of cross- ing over between certain loci to be very profoundly altered by continued selection for high and low cross over stock. So that for the present the terms proposed by Haldane* of morgan and centimorgan as measures of chromosome length do not have any precise applieation. At the same time the rate of erossing over is the only measure avail- able and ean not be given up until a better one is found. The term morgan, referring to a one-hundred-unit length of chromosome is convenient but does not have the bio- logical significance that a fifty-unit length of chromosome would have. Since every gene is independent in trans- mission from all other loci in the same chromosome more than fifty units distance from it, and has the usual Men- delian relation with them as well as with all the factors in the other chromosomes, the term mendel would perhaps be useful, if the employment of such terms can be justified at all. Applied in this way a mendel is a measure of chromosome length equivalent to fifty per cent. of eross- ing over. It should be noted that a mendel is not comparable to a 1 Proceedings of the National Academy of Science, 6: 663—670, 1920. 2 Journal of Genetics, 8: 299-309, 1919. No. 643] GROWTH IN CHROMOSOMES 173 single short chromosome fifty units long. It is not to be thought of as a fixed portion of any chromosome. The chromosomes are, or course, not to be considered as marked off into fifty-unit lengths. But the result of re- combination with a large number of factors is approxi- mately the same as if such were the case. Because it brings out the fact which has not always been fully appreciated, that recombination within a chromosome takes place as easily as between different chromosomes, when the distance between the loci is sufficiently great, the term mendel as a measure of chromosome length may have some value. In the primitive unicellular organisms it is conceivable that the hereditary substances were not located in a mechanisn as well regulated as in the higher organisms. As specialization increased, the grouping of factors in chromosomes has undoubtedly been of very great evolu- tionary significance. The chromosome mechanism has been subjected to natural selection as severely as any external morphological feature and has developed co- ordinately with sexual reproduction—the one to make recombination possible, the other to make that process orderly. Although it is largely speculation it seems necessary to believe that there is some functional relation between the factors associated together in a chromosome or por- tion of a chromosome. There is some evidence for this in the quick and exact return of certain species hybrids to one or the other parental type. Evidently only those individuals resulting from gametes in which crossing over has not occurred are able to live. So far the factors which have been located seem to be placed at random in the chromosomes, and it is impossible to make out any significant relation among them. This in itself may be an indication of an immense number of hereditary de- terminers which play a part in the organism. For as yet the function of only the relatively superficial factors can be seen. The vitally important ones can not be dis- pensed with and therefore can not be studied except as the lethal factors show some effect in hybrid combination. EXPERIMENTAL STUDIES ON THE DURA- TION OF LIFE II. HEREDITARY DIFFERENCES IN DURATION or LIFE IN LINE-BRED STRAINS or DROSOPHILA + PROFESSOR RAYMOND PEARL AND SYLVIA L. PARKER INTRODUCTION Ir was shown in the first paper in this series (27)? that there was a marked difference in mean duration of life, and in the form of the ls curve, between wild-type stocks of Drosophila on the one hand and the synthetie quin- tuple mutation stock on the other hand. It was further made clear that, because of the technique used in the ex- perimental work, there could be no doubt that the basis of this difference must be hereditary and not environmental. Furthermore, Hyde (11) and Pearl (6) have presented evidence for the Mendelian inheritance of this character duration of life. Given it to be the fact, as the just cited work demon- strates to be the case, that there are hereditary differ- enees within the same species of Drosophila in respeet of duration of life, the problem which next presents itself is to determine whether within a particular strain of Droso- phila hereditary differences exist, and if so what their magnitude may be, their degree of permanence, ete. In 1 Papers from the Department of Biometry and Vital Statistics, School of Hygiene and Public Health, The Johns Hopkins Uni , No. 48. 2 A word of explanation is necessary as to the ed of handling biblio- graphie referenees in this series of papers. In the first paper a list of 26 references numbered eonseeutively from 1 was appended. It is proposed not to duplieate referenees in any subsequent paper in the same series. nse- quently the first new bibliographie eitation in the present paper is numbered 27. When any reference is made to titles already cited in the first paper in the series, the numbers whieh they bear in the list appended to that paper will be used. This practice will be adhered to in all subsequent papers in this series of Studies. 174 No. 643] STUDIES ON THE DURATION OF LIFE 175 short one wishes immediately to get a kind of knowledge for this organism and character similar to that which Johannsen (28, 29) got for the size character of beans from his pure-line work. The first, and in a sense pre- liminary, investigations on this problem will be presented in this paper. Later in the series we expect to publish much more extended and penetrating evidence on the same problem. Some, however, must be presented early in the series in order to make the account of subsequent experiments intelligible. It is obvious that in the ease of an organism like Droso- phila it is impossible to have a pure-line in the strict sense of Johannsen. The most that one can do is to have inbred lines, and the most intense degree of inbreeding possible in the premises is by brother X sister mating. The gen- eral plan of the experiments reported in this paper ean be outlined as follows: 1. Mate a virgin brother and sister, chosen at random each from the same one of the original 5 foundation . stocks (cf. 27). 2. Repeat this for as many pairs as the facilities of the laboratory make possible. 3. Test the progeny of each mated pair separately for - duration of life, and form for each group of such progeny a life table. | 4. Each such mated pair constitutes the beginning of a line, in which at any time the processes noted under para- graphs 1, 2, and 3 above could be repeated. In this paper will be reported the results of one such repetition. The general technique of the experimental work has been fully described in the first paper of this series and need not be repeated. It should merely be emphasized again that the environmental conditions in respect of food, housing, temperature (25° C.) and atmospheric con- ditions were identical for all the flies in the experiments here reported. 176 THE AMERICAN NATURALIST [Vor. LVI Duration or LIFE IN DIFFERENT PROGENY Groups OUT OF BROTHER X Sister MATINGS The survivorship data (ls frequencies) for 7 progeny groups each out of a mating of brother X sister are ex- hibited in Table II. All distributions are put on the same basis of 1,000 flies at emergence from the pupal stage. The absolute numbers of flies involved in each experiment, are given at the foot of each column. These numbers are TABLE I BROTHER X SISTER Matines. First TEST Lines | Original Stock Date of Date | (Described in (27)) Mating Emergence 108... | Old Falmouth April 8, 1920 April 19-May 3 IUE, s | ais " Apiil 7, 1920 April 17- Eemi : s. | New Falmouth April 7, 1920 April 18- May 200 .—— | 3: i April 10, 1920 April tire E 300... | Sepia April 7, 1920 Apri! 17-May 301..... | cs April 6, 1920 Apul 17-May : 308... ss | e April 8, 1920 April 18- May 2 TABLE II SURVIVORSHIP DISTRIBUTIONS OF PROGENY OF BROTHER X SISTER MATINGS. BOTH SEXES TOGETHER Numbers of Survivors up to Indicated Age in Lines No. Age in Days 100 101 201 202 300 301 | 303 1... 5 ud Lu Lt 1,000 | 1,000 | 1,000 | 1,000 | 1,000 | 1,000 | 1,000 ru eir eee 983 | 2993]|1,000| 689; 926; 870; 882 120 a eee 937 | 987 | 1,000 8| 727 64 IB. Oe a qa IK a ey 891 934 952, 492 809 02 702 Thee EE ie T ee 811 1 943 | 426 623 441 621 40... oM EU 743 875| 85 549 522 BO iro NUR ee 589 855 790 197 383 255 | 429 Loc D e eee tees 514]. 770] -6 1 2| 205; 311 481 is iaa oe et ae ae 406 | 599 5 148 148 130 | 261 o WD AN C NR ee AN 240 | 493 381 105 75 1 60. GE E E E eee TT 91 219 33 12 31 112 80. a R ton a A 29 99 133 | 16 6 12 56 TA 12 UTILI UNA RUE CU 6 29 10 | 0 2 0 TES hh dio as kia eee 0. IOl e i ES Crus DM re ER Ww Wed os c ur SU. VoU ce eee Kod sa | ^ie | 0 = = pur A rt Abs. No. of flies..........-- | 175| 152| 105 | 61| 162| 161| 161 No. 643] STUDIES ON THE DURATION OF LIFE 177 smaller than is desirable, but these experiments represent a relatively early stage of the work before the technique of getting maximum progenies for life table work had been perfected. Further it must be remembered that the individuals in any eolumn are the progeny of only one single pair of parents. The source of the lines together with other pertinent data are shown in Table I. 1000 a OE o, 100 o Big S z c CN iO ET ES has Eve tow E EE Y EE IY au d B iu BAAS Aah Lob cb ee ee p bob | LB les 16 4 30 36 dB 4B SR cO 060 72 8 4 9 AGE IN DAYS Fie. 1. Survivorship (lx) graphs for lines 100, 101, 201, 202 and 301. Five of these distributions are shown graphieally in Fig. 1, and their biometrie constants are given in Table TABLE III FREQUENCY CONSTANTS FOR d; DISTRIBUTIONS. First TEST i j rati i Deviati Coefficient of e Mean — of Life "— Seton ion meant 3005.55 40.45 = .84 16.38 = .59 40.49 + 1.68 101... 50.02 + .85 15.51 = .60 $1.01 = 1.31 AZUL tae 47.40 + .99 .03 + $1.71 = 1.51 208 css | 22.04 = 1.57 18.18 = 1.11 82.49 = 7.74 300....| 3119 = .83 15.76 = .59 53 = 2. S01... | 25.28 = .92 17.25 = .65 68.24 + 3.56 808.,... | 32.02 + 1.07 04 62.59 = 3.14 178 THE AMERICAN NATURALIST [Vor. LVI HI. In calculating these constants, the absolute d; fre- quencies, and not the per mille frequencies, were of course used. From these data it is at omce apparent that these progeny groups show distinct, and in some cases decidedly large, differences both in mean duration of life and in the form of the mortality distributions. Lines 101 (Old Fal- mouth stock) and 201 (New Falmouth stock) show the longest mean duration of life, and they are sensibly iden- tical in the form of the life curve, having regard to the errors of random sampling. The difference in the means for these two lines is 2.62 + 1.31 days, an obviously insig- nificant difference, only 2 times its probable error. Simi- larly these two lines do not significantly differ in absolute or relative variability, the difference between the stand- ard deviation being .48 + .92. Line 100 (Old Falmouth stock) has a distinetly and sig- nificantly lower mean duration of life than 101 or 201. Comparing it with line 101 the difference in the means is 9.57 + 1.20 days or approximately 8 times its probable error. The l- curve lies throughout its course below the lines for 101 and 201. Line 100 is also relatively more variable in duration of life than 101 and 201, but largely because of the difference in the means. The individuals in line 202 (New Falmouth stock) are the shortest lived of any here dealt with, and the shortest- lived wild-type strain we have as yet isolated. Its mean duration of life is less than half that shown by lines 101 and 201 and only a little more than half that of line 100. Line 202 shows the highest relative variability in duration of life of any of the lines here discussed. It also has the highest absolute variability with one exception (line 303). Lines 300, 301 and 303 (Sepia stock) are all relatively short-lived lines. 300 and 303 are substantially identi- eal, while 301 has a lower mean approaching that of line 202. These sepia lines are also characterized by high relative variability. No. 643] STUDIES ON THE DURATION OF LIFE 179 RESULTS or [INBRED RE-TESTS ror Constancy During the progress of the experiments described in the preceding section the offspring flies (from original brother X sister matings) in each of the lines, whose dura- tion of life was being tested, were allowed to mate at random in their bottles, and their progeny removed to form stocks of the several lines. These stocks were al- lowed to reproduce in stock bottles, all matings being therefore random within the line, for a period of about 7 months (ef. Table IV). At the end of that time it was de- cided to make a re-test of each line to see how it was then behaving relative to duration of life. There was then made, at dates indieated in Table IV, a random selection from each line stock bottle from which a brother and sister pair was bred, and these two individuals were mated to get a set of progeny on which to carry out a sec- ond set of life duration experiments. The necessary facts as to line numbers and dates on this re-test are given in Table IV. TABLE IV BROTHER X SISTER MATINGS. SECOND TEST Line from which Number of Line Date of Date of Second Selection of Po" of Vie uh Second Brother of Brother and Second Brother Brother X X Sister Sister Was Made X Sister Mating Sister Mating Mating TOO ieee T 104 April 8, 1920 November 6, 1920 LOL Gea vv ORAN 107 April 7, 1920 October 14, 1920 201.1. va 3x 207 April 7, 1920 October 18, 1920 ZUR as 208 April 10, 1920 October 14, 1920 MO ee I 304 April 7, 1920 November 6, 1920 SUP CR aus eee 307 April 6, 1920 o , 1920 803,1: 4 bees oan ` 809 : April 6, 1920 October 14, 1920 The survivorship distributions of the progeny groups of this second brother X sister mating are given in Table V, and the biometrie eonstants ealeulated from the ob- served d. distributions in Table VI. These tables are for eomparison with Tables II and III above. 180 THE AMERICAN. NATURALIST [Vor. LVI TABLE V SURVIVORSHIP DISTRIBUTIONS OF PROGENY OF SECOND BROTHER X SISTER Matines. BOTH SEXES TOGETHER | Numbers of Survivors up to Indicated Age in Lines No. Age in Days | 104 107 207 208 304 307 309 Eli... Oh oar s UR » WEE eS | 1,000 | 1,000 | 1,000 | 1,000 | 1,000 | 1,000 | 1,000 B ry S See eG dE EA | 997 | 1,000 973 833 | 1,000 862 | 1,000 PES Crise M vsus nacre gs 923 950 926 738 0 700 978 EN als Vereen Ce RI Es 871 926 819 643 870 623 911 SE Lu sU Usati iU alis 713 917 792 595 674 469 00 De OUT Au ea uu e FES | 629 901 785 478 392 489 AB. c OW VAN STI LEA ee 552 860 711 286 435 285 456 Bee Win TS. E DEN P S Vis 469 wae 644 167 261 177 267 SÉ ees eee et 395 686 530 0 152 92 89 CRM E veo seattle eee V E ae 595 3 — 109 4 67 | pete Set ee Gd er alee VOV S dr 178 488 255 0 23 44 SC ee eee Cea ip E 66 264 141 — — 8 0 WE. ERY aA) oer NIA UNE e 0 83 20 — — 8 — COS EU pue E MCN E — 8 20 — — 8 — BA RE FRHRR RE eA a dE — 0 T — — 0 — DE ae M as — — 0 — — — — Abs. No: of flies ......... 286 121 149 42 46 130 90 TABLE VI i FREQUENCY CONSTANTS FOR d; DISTRIBUTIONS. SECOND INBRED TEST Line Mean Duration of Life Standard Deviation Coefficient of No. (Days) (Days) Variation IOE .... 39.59 += .74 18.63 = .53 47.06 + 1.62 100155. 53.74 = 14.40 & .75 32.38 + 1.54 25230 45.34 = 1.10 19.97 + .78 4.04 2.03 208. 1. 25.65 = 1.53 14.68 = 1.08 57.23 = 5.42 304..... 32.09 + 1.43 14.43 = 1.01 44.97 = 3.75 807. iv. 25.22 = 16.70 = .70 66.22 = 3.79 S08. ei. 1 2.84 38.91 + 2.23 The purpose of this second test was, of course, to see to what extent duration of life was holding constant in the line. During the period between the first and second test the stocks of the several lines had been subjected to vary- ing environmental influences, in particular in relation to temperature, the stock bottles having been kept at room temperature, which varied rather extensively. Did the lines after 7 months have the same characteristic life curves that they exhibited on the first test? Allowing 12 No. 643] STUDIES ON THE DURATION OF LIFE 181 days from generation to generation in the case of flies re- producing freely at random in stock bottles, the interval elapsing between the first and second tests would cover roughly almost 18 generations. "This is a long period and affords abundant opportunity for change in the aver- age genetic constitution of the population. 1000 W N e LS Pm ms. ` va, Bu a LC ~, - - -> me ee `, oT TTT ~ o o LT TTT SURVIVORS S TTT Lp ee rs i E 4 Ir TE M a o 4 XP 4B 54 66 72 AGE IN DAYS Fic. 2. Comparing the lx lines of the first and second inbred tests. of lines 101, 100 and 301. An examination of Tables V and VI and Fig. 2 shows at once, in a general way, that the characteristic features of the several lines in respect of duration of life did in fact hold remarkably constant during this period. A more precise comparison of the means is made in Table VII. There can be no question of the substantial constancy of these lines, over the period covered in the tests in re- spect of duration of life. The le curves run well together till the upper end of life is reached, where, because of the small numbers involved, there is some irregularity. In no ease is the difference between two comparable means, as shown in Table VII, as much even as three times its prob- 182 THE AMERICAN NATURALIST [Vor. LVI able error, nor is there any certainly significant change in variability having regard to the probable errors of the differences involved. TABLE VII DIFFERENCES IN MEAN DURATION OF LIFE BETWEEN THE FIRST AND SECOND INBRED TESTS OF THE SEVERAL LINES Corresponding Lines (Mean of Second Test Difference of Means inus Mean of First) (Day: P.E. Diff £04-100 15244427542 Gg oe ca a CaN — .86 +1.12 IE FE IE A VE Re ARE T E Cer E es + 3.72 + 1.37 2.72 BUT EUIS. uA cd E ola Oe — 2.06 + 1.48 1.39 AUN MUSS Qu cceli RET eA Cn T ME + 3.61 + 2.19 1.65 BOO QU NS ons VENT CCP ERA LUE RO os + .90 + 1.65 54 E i S Sui a ci creek A CaS + 062+ 1.35 04 i a a IAS. E RC G HEC TEE + .98 + 1.40 70 Resuuts or Mass CULTURE Re-rests FOR Constancy The point may well be made that in the re-tests of the lines described in the preceding section an additional ele- ment is introduced in the faet that the flies for the re-test were the progeny of a second brother X sister mating. What one wishes to know is: what degree of constancy in duration of life is exhibited by the general stocks in each line, mating purely at random, after the initial se- lection and inbreeding? "We wish now to present some data on this point. Table VIII gives the biometric con- stants for this material. Mass culture re-tests have been made on two of the original lines, 100 and 101. "These mass culture re-tests were made in two ways as follows: (a) From the stock bottles of the line to be tested a large sample of progeny was taken at random each day as the flies emerged from the pupal stage, and these progeny flies were put in small bottles for a duration of life experiment in the usual way deseribed in (27). (b) From the stock bottles of a partieular line to be tested a number of virgin flies (usually 8 to 10 of each sex) were taken at random immediately upon emergence, and mated as a group in a mating bottle. The progeny from this sample was then removed, upon emergence, to small bottles and a regular duration of life test carried as described in (27). 183 No. 643] STUDIES ON THE DURATION OF LIFE IZ ES eoe -+ 6’ = Qc + EET E O6TV'T die. isi s&up 6c ‘our OL «9*45»4€92a42*9209299292952829*289*229292929225929^25^5 eee eee * Hao 9ouod9gH(] OL'T E g£rILd- 22* E ZO'L d OL'T E 208g 4 FER. s&up €I *OUI q Tiere «tee eee ee hi ee ee he eer ee ae Se E VO 99uo319gT I£'I + IO'I£ 09' + IG'GI 98' = 0'09 est Z enw- “Ady ; ; üt — ie 19]8IS X JeYJOI BUBUO 99'I = FSFE GL’ F OL'9OT COI = ES°SP Pol y ‘“Ady-ST UN ia | pen m n 808 gr o1nj[nz) ssvyy 60°L = prot 6P £9'66 OL = 60'E9 Ly IG 390-96 '1dog OtBI — [oeMÍ y ommo sep [coos TOT soc E 92'0g4- 08 un OL + ery EM OFZ due RES s&up eI ‘oul 9 wer i| $i] | | ss |] |t n | || ccs rccte tatc tt 9ouo19gt(T 89'I + 6T'O0v 69' + 8€'9I v8' F SV'OF GLI € Avyw-61 ‘dy DI — 01 19j818 X 10q]01q [BUISLIO Tes = SO'TA vo F SV'$6 94' * G0'e8 EE 13 390-96 '1dog ar — o Ira emymnosesW | 9099969060606 ****001 UOT)ULIU A (s&&(q) uon (s&vqq) Song Jo 9ouoz.1ouip ‘ON jo qn yo -UIAO([ prepuvjg uBoyy S1oquinNr jo 898A 3891, eury SANI TVNISIUO IIIA WTA Vib ‘SLSAL-AY AUALIAH SSVI O4 SLNVGSNOD AONANOAU,T 184 THE AMERICAN NATURALIST [Vor. LVI It is at once apparent that the mass re-tests on line 101 gave extremely satisfactory results as to constancy of duration of life in the line, after intervals of approxi- mately 5 and 11 months. The mean value for either the A or the B test does not signifieantly differ, having regard to its probable error, from the mean shown on the original test at the start of the line. The mean of the A mass re- test almost exactly agrees with that of the second inbred test of the same line, as given in Table VI. In the ease of line 100, the mass re-test after 54 months approximately does not give such close agreement. The mean is signifieantly lower, the difference being 6.6 times its probable error. No explanation of this result is, as yet, forthcoming, but it probably means no more than lack of genetic purity in the line. It is, however, interesting to note that the sense of the change is in the same direction as that in which line 100 in general differs from line 101, which we regard as our most typical wild-type line in re- spect of duration of life. That is, line 100 is, as com- pared with 101, a shorter-lived line. Its mass culture re-test is still shorter lived. i The variability in respect of duration of life, whether measured in absolute or relative terms, is uniformly higher and in two cases out of the three by a significant amount in the mass culture than in the original inbred tests. This is, of course, exactly what would be expected on general genetice grounds. One brother X sister mating, as has been shown by Pearl (30), Jennings (31) and others, reduces the heterozygosis in the strain by only 50 per cent. It is interesting to note, in connection with the explanation suggested above for the difference in the means in the ease of line 100, that the variability in the mass re-test on that line is very much bns than in the original inbred test. A mass re-test was carried out on two of the lines from the second brother X sister matings. The results from these experiments are presented in Table IX. 185 No. 643] STUDIES ON THE DURATION OF LIFE TCR -F IZ'0g£— 64 = e0'6 ai rI E eot + jo tom &vuvp I “SOUL £I AL e a "4 9 «49 * 9n x t 4 9». kc a» io DN E a Vo 99Uu919]g(T L2°S E: £c'9 PIE 69° nA ost mam 46 E YO'I + poe a s&up $86 ‘SOUI 9 a S E R^. 4 954.» 9 99.4 à S w^ c". 9 9 Ah aa as a ee o YS 99u919gt(T TUS + 69c9 cL = P0'06 LOT = BOGE 191 Z AVJN-SI idy 061 194818 X JOYyZOIG FEUIBEIQ | 77707770707 ee £08 $6'6 * I6'8€ 99' e P8Gl 16° * 09ER 06 I '^oN-S86 990 OZ6I e dvd g 1oqsts X 10gj04q puooog |'^^^^* Jd D EAT. 64' - BEZE te = GOTT re ct POPE SOP GZ-8T ABI imr o qme rm ‘oes emapno ssvyy [0777 pee e ip 608 aril = 67g — 96° * Z8 I — SS" + 890 — eri sAep 2% ‘soul Il Le Rhe as RE RM ees E ua eoa ra 1 yg oouodogid £9'I ES 98°F ee Eit ES t'e üi OUI ES 00°F yr dis Sep p souo pe eei E aie PLI M UR Ri O, parr t t n ng YS əsuə yig I£TI = IO'I& Og = Id SE S8' æ- c0'08 6ST @ SvVW-LI "dy 0c6I $ bob . O l9j8I8 x 19g304q PuB Li 2 Er d AN E. E R . ee TOIT PIT = BEGE 94 F OF ZL LOT = PLES Tél y '^oN-S6 "090 OZ6I oig e X Jeq)oq puooeg [577077070 000000077 101 to + 6926 ST + 69°ET Sc = PL 6F SEE£'I cc 4dy-6I "dy IZ6I Seas Senn ccr t yang sep ee 101 UOTPELIB A (sAeq) uon (sAuq) Sol]. JO vus ‘ON Jo yuan yoo -1A q pieputjg u*valN sequin N jo pqa LEOR A our[ WNIT GXUZSN[ WOLM,, ‘SISAL-AY AWALIAO SSV "O4 SLNVISNOD AONSODN,[ XI W'IdViL 186 THE AMERICAN NATURALIST [Vor. LVI The substantial constancy of line 101, in both mass and inbred tests, is evident. In respect of variability the line behaved somewhat like 303 diseussed below. In line 303 again the constancy of the line in respect of mean duration of life is as definite as could be expected. Over periods of approximately 7 and 13 months, the mean duration of life has not sensibly changed, having regard to the probable error involved. The results respecting variability are somewhat anomalous. Both the second inbred and the mass re-test show variability of a dis- tinetly lower order than was exhibited by the progeny of the original brother X sister mating. It seems probable that the original test by accident gave a variability re- sult higher than was really characteristic of the line. But the mass culture re-test exhibits a lower variability, not - certainly significant, to be sure, than the first test on line 309. Of course it is to be expected that with continued brother x sister mating the variability of mass cultures from the line would come nearer and nearer to that of a further inbred lot of progeny from the same line. Prob- ably the results of Table IX are an expression of the realization of such expectation, obscured by the fact that the numbers are small and the errors of sampling conse- quently relatively large. Discussion AND SUMMARY The data presented in this paper appear to demon- strate, with comprehensiveness and accuracy, three broad facts. A. That there exist in a general population of Droso- phila melanogaster (or its mutants) genetie differences in respect of duration of life. | B. That these genetic differences are capable of isola- tion, by appropriate selection and inbreeding. C. That within an even moderately inbred line, the gen- etic differences in duration of life remain constant over periods of at least 10 to 25 or more generations. No. 643] STUDIES ON THE DURATION OF LIFE 187 These facts, based upon the determination experi- mentally of the duration of life of 3,039 individual flies in 18 experiments, under constant environmental conditions, place this character ‘‘ duration of life’’ in the category of genetically definite and workable characters, and indi- cate that it will just as well repay careful analytical study as the characters more usually dealt with. Furthermore, duration of life is a character of great general biological significance. LITERATURE CITED 27. Pearl, R. and Parker, S. L. Experimental Studies on the Duration of lafe, L cater ype Discussion of the Duration of Life in Droso- phila. AMERICAN NATURALIST, Vol. 55, pp. 481-509, 1921. 28. Johannsen, W. Ueber Erblichkeit in Populationen und in reinen Linien. Jena (Fisher), 1903. 29. Johannsen, a Elemente der exakten Erblichkeitslehre, 3d Edit., 1913. 30. ge R. On the Results of Inbreeding Mendelian Population; & Correc- and iie nsion of Previous Conclusions. AMERICAN NATURALIST, 31. Jennings, H. S. Formule for the Results of Inbreeding. AMERICAN NATURALIST, Vol. 48, pp. 693-696, 1914 SHORTER ARTICLES AND DISCUSSIONS NOTE ON A CASE OF HUMAN INBREEDING? THrouGH the kindness of a friend the following pedigree is presented. It is that of a family of English stock, which has been in this country since the early eighteenth century, and during that time has been one of the principal families of a rural community. ee 4i 4T: 3 4 r 4 ] F 8 OTHER CHILDREN sT? dT? du d | $ gT? ENS IO OTHER CHILDREN i Í zl TUE ER REN dJ? 2 LMNG 90+ YEARS AG T f E tf 1 6 OTHER CHILDREN dT? 3 d LIVING , NORMAL DIED TBG DIED TBC DiED TBC ABOUT 20 YEARS 46 YEARS ^— 25 YEARS QA dB gc DIED INFLUENZA IB YEARS Fic. 1. Pedigree of an inbred family. To quote from my eorrespondent's letter, ‘‘A was a fine young girl. She had graduated from high school but did not go to eollege as her mother had died in the summer and she wished to take charge of the home. B is about 16 years old. A splendid young man, bright and apparently healthy. He is in school standing about average. C is 10 years old or thereabouts. An exceptionally bright child and one that is very much alive and full of spirit." Assuming that the line of descent represented in the figure by a broken line, indicating that the number of generations is not © known, includes the same number of generations as the other . lines.the coefficients of inbreeding * for the children in the last 1 Papers from the Department of Biometry and Vital Statisties, School of Hygiene and Public Health, Johns Hopkins University. No. 41. 2 Pearl, ‘f Studies on Inbreeding," I-VIII, AMERICAN NATURALIST, 1913-17. 188 No. 642]. SHORTER ARTICLES AND DISCUSSION 189 generation (viz, A, B, and C) are as follows: Z, — 0, Z, =25, Z. —25, Z, —3431, Z = 25, 47, == 27.1, ie. in five generations of ancestry the inbreeding is about a quarter of the possible maximum. There is no deleterious effect of inbreeding apparent in this pedigree. The three children in the last generation, the most inbred of any, show no signs of abnormality. In their father’s fraternity, for which Z, —2Z,—2,—0; 2,—12.5; Zr,—4.1, or in four generations of ancestry there is 4%; of the possible maximum inbreeding, one of eight died of tuberculosis; the other seven have attained adult age. In the mother’s fraternity, for which Z2, — Z,-= 2%, ==0;; 2, 2=6.25:. 24, 29, two. out of the three have died of tuberculosis. The least inbred, there- fore, show the greatest susceptibility to tuberculosis. The num- bers are, of course, too small to draw any certain inference, but so far as they go, they accord best with the view that there is no harmful effect of inbreeding per se. Jonn Rice MINER ON COLOR VARIATIONS IN CHITONS ` THE question was raised by Bateson (''Materials," 1894, p. 307) as to whether variation occurring in serial parts whose repetition is not strictly speaking of a metameric sort, would be found to simultaneously affect each of the parts involved in sueh a series. With this point in mind he examined a collection of ehitons, the 8 shell plates of these animals providing an exeellent opportunity for such observations. He found color variations affecting all the plates of an individual to be of rather rare occur- renee, but that plates 2, 4 and 7 seemed, on the other hand, to exhibit a deeided tendeney to vary together (in several species of Chiton). Although the problem of metamerism, so far as it concerns variation, has perhaps lost some of its original attractiveness, I have thought it worth while to point out that in Chætopleura several curious types of shell variation are apparent, involving either simultaneous variation throughout the series or variation in a single shell-valve alone, or both. a 190 THE AMERICAN NATURALIST [Vor. LVI The commonest type of color pattern, in Chetopieura apiculata, is one which involves a double band of blackish or grayish pig- ment running the length of valves 2 to 8, the bands joined on 2 and on 8 by continuous semicircular blotches (cf. Fig. 1, d). In 3 out of 219 specimens, however, a central dark stripe, clearly marked on valves 2 to 8 inclusive, accompanied a much fainter double band extending to valve 1 (Fig. 1, b), and made up of regular triangular grayish blotches on the posterior bor- ders of valves 2 to 7. In one ease, a bright central band of white E Fic. 1. Chaetopleura apiculata (x 2). was marked by a median grayish blotch on valves 3 to 7 (Fig. 1, a) ; a paler example of the same kind is shown in Fig. 1, e. These are instances where variation from the more usual color type clearly has taken place simultaneously in the whole series of valve-plates, this condition being seen not only in the form of the central stripe, but also in the shapes of the individual pig- ment blotches comprised in the double band (ef. Fig. 1, a and b), as well as in a few eases where three small but distinet pigment flecks were noted on the lateral field of each plate. In addition, however, two sorts of pattern variation occur which are quite different from the foregoing. In three instances a definite yellow or orange central bloteh appeared on valve 2, and nowhere else (Fig. 1, e). And in five further instanees there was found a marked blackish bloteh at either lateral margin of valve No. 642] SHORTER ARTICLES AND DISCUSSION 191 4 (Fig. 1, d, e). In two further cases, this type of lateral marking was continued in the form of less distinct marginal blotches on valve 3. The marginal blotches on valve 4 may accompany an otherwise ‘‘normal’’ pigment pattern (10 examples in 219 ex- amined; Fig. 1, d), or may be present where there is evidence of a tendency for the formation of a distinct axial stripe (five ex- amples; Fig. 1, e). | It is evident, then, that in Chitons eolor pattern variations may oeeur in sueh a way as to affect single valves only (and, in Chetopleura, specifically valve 2 or 4); and either quite inde- pendently of this type of variation or accompanying it, may also affect all valves in the series simultaneously. Such variations are quite independent of age. W. J. CROZIER RUTGERS COLLEGE FUGITIVE NET-VEINING IN THE CICADA (HEMIPTERA) TILLYARD has lately noted that, besides the chitinized veins which serve for the support of the insect wing, there exist in some eases at least fugitive blood-veins during the expansion of the wing, which later collapse and more or less completely dis-- appear when the wing dries. In the Lepidoptera 1st A and the base of M are veins of the same character, and possess trachese like other longitudinal veins in that order. In every particular except the absence of chitinization these appear to be true veins, and in forms where the veins are provided with speeial series of sete, as in Acrea, they are often similarly supplied. In watehing a eieada expand, recently, I saw appear, as the 1 Proc. Linn. Soc., New S. Wales, 44, 621; 1919. 192 THE AMERICAN NATURALIST [Vor. LVI expansion approached completion, a regular system of blood- veins in the spaces between the permanent veins. These show plainly only in the few minutes when the wing has become partially transparent, but in the adult wing they produee a characteristic waviness of the membrane, and a few of them may be seen in a favorable light as faint white lines. The arrange- ment is perfeetly definite: the narrow cells are filled by a series of simple, evenly spaced, cross-veins, while in cells R, 1st M, and M they form a double series of cells alternating with each other. On the narrow margin beyond the ambient vein they are evenly spaced, the regular longitudinal veins each ending opposite the middle of a marginal cell. Toward the costa there are two veins opposite each definitive cell, while opposite cells M, and M, there are three, and opposite cell Cu, there appear to be four. The margin of the hind wing is similar, but the dise of the wing was not observed. In the large triangular anal cell (3d A,), instead of cross-veins there is a series of closely spaced parallel longitudinal veins, which remain visible in the dried wing. It seems possible that these structures are the relic of a net- veining such as occurs in the Neuroptera. The different arrange- ment in the anal region is especially suggestive, as it would correspond to the plaited portion of the wing in the Orthoptera, where there exist numerous parallel longitudinal veins. The figure is drawn from memory so far as the fugitive veins are concerned, checked up by the few that could be traced in the dry nne it ean be trusted only approximately. WM. T. M. FonBEs CORNELL UNIVERSITY, ITHACA, NEW YORK THE AMERICAN NATURALIST Vor. LVI. May-June, 1922 No. 644 IS THERE A TRANSFORMATION OF SEX IN GS? PROFESSOR W. W. SWINGLE OSBORN ZOOLOGICAL LABORATORY, YALE UNIVERSITY Tuis paper is a reply to the recent article of Dr. Emil Witschi which appeared in a late issue of the NATURALIST (Vol. LV, No. 641). Witschi is quite convinced that the problem of sex development and differentiation in frogs has been settled, and that nothing further remains to be said. However, the writer feels that instead of being solved, the time has come for a revision of the entire question of sex development in Anurans, and that the subject is ripe for a reinterpretation upon a more ra- tional basis than that accorded to it heretofore. The first portion of the paper will be devoted to a brief exposition of the writer's interpretation of sex in frog larve based upon data obtained from a study of the bull- frog. The second part of the paper is a reply to certain questions raised by Dr. Witschi. In larval males of the bullfrog two gonads are formed, just as there are two kidneys formed, a pro-testis or em- bryonie sex gland destined to degenerate and disappear in ontogenetic development and a definite or functional testis which replaces it. The germinal elements of the pro-testis arise in the entoderm and migrate into the germ ridges early in embryonic life. The cells multiply rapidly and together with the mesodermal elements of the germ glands form paired ridges projecting into the eclomie cavity. While the tadpole is very immature and has yet a year of larval life before metamorphosing, the 193 194 THE AMERICAN NATURALIST [Vor. LVI germ cells of the pro-testis undergo a precocious and abortive sexual cycle culminating in degeneration and resorption. Beautiful cysts of spermatocytes are formed, but the first maturation division rarely proceeds past the m - $5. VN x X- b prs fi 8 E Fic. 1. Transverse section pro-testis R. catesbeiana tadpole. Animal has a year of larval es remaining. A, Spermatogonia Showing tse ipe polymor- phism due to incomplete fusion ehromosomal vesicles; B, Final seil tia gonad which develops as a core within pro-testis. No. 644] TRANSFORMATION OF SEX 195 anaphase owing to fragmentation of the centrosome and consequent formation of polyasters (Fig. 1). Sometimes aberrant spermatids are formed by suppression of the first and second maturation divisions and growth of axial filaments from the centrosome. Practically all the germ cells of the pro-testis degenerate and disappear while in various stages of maturation—some undergo an oviform type of degeneration, i.e., hypertrophy enormously and take on the superficial characters of oocytes. The ovi- form type of degeneration, however, is more character- istic of the short larval-lived frogs than of R. catesbeiana, for in many animals these large cells appear rarely and in others not at all and this is an important point to keep in mind. This type of degeneration will be discussed in detail in a later paper; suffice to say it gives no clue to the sex of a cell. (See Plates 1 and 2.) Some cells of the pro-testis fail to take part in the abortive sexual cycle persisting through the phase of ma- turation and degenerate as spermatogonia. These ele- ments migrate into the sex cords (Fig. 1, g) which have formed meantime, and form a core of germinal tissue extending through the center of the pro-testis. This core of tissue plus the sex cords is the anlage of the definitive testis and is quite distinct from the pro-testis, the cells of which are maturating and degenerating, whereas the cells of the forming functional gonad remain as primitive spermatogonia. The definitive testis by rapid growth completely supplants the pro-testis which usually disap- pears some time before metamorphosis. The functional gonad is generally fully formed at metamorphosis when the larve are two years of age. Some tadpoles, but not all, develop ripe spermatozoa in the gonad at metamor- phosis due to a second sexual cycle of the germ cells of the definitive gonad. (Swingle, '21, Jour. Exp. Zool., Vol. 32.) In the frogs with short larval-life the same succession of gonads oceurs, but in these forms the developmental processes are greatly accelerated and the pro-testis ma- 196 THE AMERICAN NATURALIST [Vor. LVI turation cycle is eut short by the cells early becoming senescent and undergoing oviform degeneration t.e., hy- pertrophy to such an extent as to superficially resemble oocytes. This oviform degeneration occurs to an even more marked degree in the progonad of the toad which has a still shorter larval-life, e.g., in Bidder's organ. In male anurans the entire pro-testis or larval gonad is the homologue of the male organ of Bidder in Bufo. The pro-testis of the short larval-lived frogs has been misinterpreted as an ovary owing to the oviform-type of degeneration characteristic of many of its senescent cells, and hence tadpoles are said to develop first as females, fifty per cent. later transforming into males. The normal embryological process by which the definitive testis de- velops as a central axis through the degenerating pro- testis or larval Bidder’s organ, has been described by Witschi as the transformation of female tadpoles into males. In R. catesbeiana, where the larval life is pro- longed over two years, the true nature of the pro-testis is revealed, for relatively few of the cells are of the ovi- form type and all transition stages between such cells and normal spermatocytes occur. The evidence pre- sented by this material will be published in due time, and is too clean-cut to admit of any doubt that the entire larval gonad of male anurans is simply an embryonic male sex gland rudiment and not a temporary ovary. Witschi’s Fig. 6 (this journal, Vol. LV), which he sup- . poses is an ovary transforming into a testis is simply a transition stage in the development of the definitive testis, and degeneration of the pro-testis or Bidder's organ in a short larval-lived frog. Compare his Fig. 6 with Fig. 1 of this paper and note how the true male character of the cells of the pro-testis comes out in Rana catesbei- ana tadpoles. | When the facts are considered it is evident that th transitory gonadie rudiment of male frog larve is an or- gan of Bidder which degenerates and is replaced by the de- finitive gonad. Any one who has studied the oviform-like No. 644] TRANSFORMATION OF SEX 197 cells of the so-called sexually intermediate tadpoles and compared them with the cells of Bidder's organ in male toads, is at once struck by the remarkable similarity in their origin, development, structure and fate in the two groups. They are identical. The crux of the problem is the nature of Bidder’s organ in male Bufonide and of the oviform-like cells of the pro-testis. The advocates of sex transformation have assumed that such cells are undoubtedly female, but no proof has ever been advanced that they are. Their ultimate fate is the same as that of the first year spermatocytes in the bullfrog tadpole — degeneration (see Plates 1 and 2). The sex-transforma- tionists have been misled by the idea that everything superficially resembling an oocyte is necessarily such, or that any cell in tadpoles and first-year animals undergo- ing the early growth stages, leptotene, pachytene, etc., is to be regarded as female. These are fallacious criteria. Enormously hypertrophied oocyte-like cells which have passed through the early growth stages and entered the ** germinal vesicle ’’ period so characteristic of oocytes, occur as normal features of the male sexual cycle of certain animals, e.g., myriapods (Figs. 5-8). These animals were at first regarded as hermaphrodites by Blackman. (1905, Bull. Mus. Comp. Zool., Harvard, Vol. XLVIII, no. 1) who found upon examination, however, that these ‘‘ oocytes ’’ were in reality spermatocytes of giant proportions, and developed into spermatozoa. The writer has examined some of Professor Blackman's ma- terial and the oocyte-like charaeter of the male sex-cells is remarkable. In the material examined these cells prac- tically fill the gonads. Firket, 1920, working on the chick embryo, describes and figures spermatocytes undergoing oviform degeneration, i.e., enlarging to such an extent as to resemble oocytes. There are many other cases re- ported in the literature. How does Witschi know that the transitory oocyte-like cells he deseribes in the future male tadpoles or so-called hermaphrodites, are female cells and not senescent organ of Bidder cells occurring 198 THE AMERICAN NATURALIST [Vor. LVI in the eourse of the abortive and degenerate sexual cycle of an embryonie pro-testis? The work of Witschi on the problem of sex in anurans ean be summarized thus: He has described in great de- tail and with admirable exactness the process of develop- ment of the pro-testis or Bidder's organ in the short larval-lived frogs, its degeneration and final replacement by the definitive gonad. This process he calls transfor- mation of females into males. The experimental investi- gations of Witschi upon sex transformation by environ- mental influences consists of this: By means of such agents as heat or cold, etc., he has simply modified the normal course of development of the pro-testis — Bid- der’s organ, thereby accelerating or delaying the devel- opment of the definitive testis. The experimental results show that it is possible to modify the developmental rate of the embryonic testis. Similar experiments carried out with regard to other larval structures would unquestion- ably give similar developmental modification. Cold hin- ders metamorphosis and all the normal structural changes metamorphosis implies. All of these environmental influ- ences are interferences with the normal cycle of the go- nads, by which the development of the definitive gonad out of the pro-testis is accelerated, retarded, or possibly prevented entirely. The following quotation from Wit- schi '14, page 10, is significant in this connection: Bei seinen Untersuchungen war es Hertwig aufgefallen, das unter dem Einfluss verschiedener Aussenbedingungen sich nicht nur die Geschlechtsziffern, sondern oft auch in ganz auffilliger Weise der Rhythmus, in Welchem die Keimdriisen und manche andere Organe sich anlegen und entwickeln. It is probable, judging from certain experiments re- ported, that the degree of development attained by the larval gonadic rudiment, its position in relation to the definitive gonads, its period of persistence, non-formation in some forms, and such like, may vary in different frog species and is determined by heritable factors. For ex- ample, in Bufo, the structure persists throughout life in No. 644] TRANSFORMATION OF SEX 199 males, disappears after two years in females, and is an- terior to the funetional gonads. In frogs it forms the outer husk of the germ gland enclosing the centrally de- veloping functional testis and may or may not show the oviform type of degeneration, e.g., R. catesbeiana. If sex is so labile in tadpoles and young frogs, and females so readily transform into males under environ- mental stimuli, why is it that such sex reversals do not occur in adult frogs after the degeneration of the pro- testis and the formation of the definitive testis has oc- curred? All investigators are agreed that the sex ratio of adult frogs of all species reported is approximately 50-50. If environment (ever changing in the same lo- cality, and never the same in different regions), plays such an important sex transforming róle, why do male tadpoles never transform into females — all investigators agree that they do not. Why do only fifty per cent. of the so-ealled larval females transform into males if they were not zygotie males from the beginning, and why do not all female frog larve transform into males instead of only fifty per cent. if such transformation is possible? Appeal eannot be made to Professor Hertwig's well- known late fertilization experiments because in these ex- periments the influence of the over-ripeness of the egg upon the zygotie conditions determining sex are unknown. Hormones! To date there is no positive evidence that such secretions have ever aetually changed a female germ cell into a functional male germ-cell. Cases of hermaphroditism in adult frogs are thought by some to furnish evidenee of a sex transformation in frogs. However, true hermaphroditism in adult frogs is as rare a phenomenon as it is in mammals when we con- sider the few recorded cases, and the enormous number of frogs annually dissected the world over. Crew (’21), Journal of Genetics, Vol. II, no. 2, has summarized the recorded cases of abnormal sexual organs in frogs and states that there are forty cases. To this number should be added a recent case described in the bullfrog, making 200 THE AMERICAN NATURALIST [Vor. LVI forty-one. Among these forty-one cases, there are but twenty-seven true hermaphrodites. Crew's cases, twenty- one to thirty-three, inclusive, are not hermaphrodites, nor is case thirty-eight, as none of the animals possess ovotestes and some are entirely without gonads. True hermaphroditism in frogs is a permanent and pathologi- eal condition, probably due to a mix-up in the genetic constitution of the individual, and is not to be confused with the present problem which has to do with a normal but transitory embryological process. Much has been written about the marked ‘‘ sex poten- cies ’’ of various portions of the gonads in so-called sex- ually intermediate frogs, i.e., females transforming into males. It is claimed that the outer rind of the gonad exerts a profound female sex influence, while the inner portion exerts a purely male influence. Germ-cells re- maining in the outer husk (the main portion of the larval gonad by the way) of the gland are female, those migra- ting into the central part among the sex cords become male. All such speculations are based upon misinter- pretations. The outer portion or husk of the larval male gonad is simply the pro-testis, the cells of which are un- dergoing a precocious maturation cycle just as they do in the organ of Bidder in Bufo, the inner portion or sex cord region is where the definitive gonad begins develop- ment and as it spreads and grows the embryonic male gonad degenerates and disappears. It is in the region of most marked *' female " tendencies that the writer finds in the bullfrog entire cysts of unmistakable spermato- cytes, and occasional spermatids (Fig. 1, e). In other words, the pro-testis — what Witschi regards as an ovary — can in the bullfrog, where its development is greatly prolonged, give rise to practically mature male sex prod- ucts. Recently, the writer made an observation of con- siderable interest. In the degenerating Bidder's organ (pro-testis) of a two-year-old male larva in which forma- tion of the definitive testis had been delayed until meta- morphosis and in which the oviform type of degeneration No. 644] TRANSFORMATION OF SEX 201 was the most marked of any animal yet observed, several cysts of unmistakable spermatocytes and spermatids were observed. They arose from the maturating cells of what Witschi regards as the female part of the gonad — in reality the pro-testis, and were of the cell type charaeter- istic of the adult frog. This observation shows two things elearly: (1) That the direct descendants of the male primordial germ cells (pro-testis elements) can pro- duce practically mature. germ cells; (2) that the sper- matocytes of the structure regarded by the writer as a pro-testis are really male cells, and that the structure in so-called sexually intermediate frogs and tadpoles is in no sense to be regarded as female in character. Another point is of interest here— the writer has never observed direct testicular development in R. cates- beiana, though it probably occurs in some strains; the indirect method alone has been found, e.g., first a pro- testis is formed which is later supplanted by the defin- itive gonad. In the bullfrog, which has the longest larval life of any anuran, the pro-testis persists longer than in other forms, sometimes two years before giving place to the definitive gonad. What the writer calls a pro-testis of so-called sexually intermediate tadpoles is according to Witschi a transitory ovary. If this is true why is it that despite its persistence for such a long time, relatively few oocyte-like cells are found in R. catesbet- ana and in many individuals none, throughout a two- year period, but instead the structure produces sperma- tocytes and sometimes spermatids? Why is it, if this structure is an ovary in the so-called females that later transform into males, that the shorter the larval life of male anurans, the more the pro-testis in its structure and behavior resembles the Bidder’s organ characteristic of male toads, due to rapid oviform degeneration of its cells; the longer the larval life, e.g., Rana catesbeiana, the more the germinal elements undergo a normal sexual cycle characteristic of male cells? The answer is, because in forms with extraordinary prolonged larval lives the 202 THE AMERICAN NATURALIST [Vor. LVI true nature of the embryonic male gonad has sufficient time to manifest itself before being supplanted by the definitive gland. We come now to a discussion of the nature of Ridder S organ in Bufo, for this is the classical example of ovi- form degeneration of racially senescent germ cells. Here- tofore, this embryonic sex gland rudiment has been re- garded as characteristic of toads, but such is not the ease. In frogs the pro-testis or larval gonad is a Bid- der’s organ, destined to be replaced by the definitive male gonad developing within; in male toad larve on the other hand, the functional gonad arises behind the pro-testis or Bidder’s organ, consequently this structure persists as a degenerate gonadie rudiment attached to the func- tional gland. According to the writer’s view, Bidder’ s organ in Bufo is simply a vestigial larval gonad persisting throughout life and has the same sex as the definitive gonad behind it— male in males, female in females. It is just as though the pro-nephros of tadpoles persisted as a non- funetional and degenerate rudiment at the end of the mesonephros. That many such larval and embryonic rudiments do persist through adult life in various ani- mals is a commonplace of embryology, and their per- sistence in one species and total disappearance in another related one, is also well known. Bidder’s organ in Bufo then, is a persisting, in frogs a transitory, embryonic sex gland rudiment, a relic of a phylogenetically earlier sexual condition. The functional gonads are more re- cently acquired structures (like the larval mesonephros) superimposed upon the older degenerate glands. Briefly stated, the evidence for the view that Bidder’s organ is homologous to the pro-testis of frogs and that it is not a rudimentary ovary except in female animals is as fol- lows: 1. The cells of Bidder’s organ in Bufo are unquestion- ably germ cells. The gland appears very early in em- bryonic life (two weeks after hatching) and its cells far No. 644] TRANSFORMATION OF SEX 203 outstrip in development the cells of the definitive gonads located posteriorly. 2. The cells of Bidder's organ extremely early in de- velopment undergo a precocious and abortive matura- tion cycle and become senescent and degenerate oocyte- like structures when the germinal elements of the func- tional gonads have barely started to multiply to form the definitive glands. This occurs in individuals of both sexes. 3. The larval maturation cycle such as occurs in the bullfrog, and in other anurans, throughout the entire larval gonad is confined to Bidder’s organ in Bufo, and the changes occurring in this structure do not affect the normal developmental cycle of the definitive germ glands behind. 4. The so-called transformation of female animals into males, claimed by Witschi and others to be the normal course of development in frogs, does not occur in toads. Why? Because in Bufo, the definitive gonads are from the beginning located posterior to Bidder’s organ, and it is not necessary in order that they may develop that this structure degenerate and disappear as is the case in frogs where the definitive testis starts development as a core within the pro-testis or Bidder’s organ, neces- sitating its complete destruction. ew have ever claimed that sex in toads is labile and easily reversed by environmental influences. Why? Because the sex of the definitive gonads is definitely fixed and clear cut at an early stage of life. The separation of Bidder’s organ and the gonads has precluded the possibility of confusing the pro-gonad and the definitive gonad. ' 6. Bidder’s organ is merely a persisting embryonic gonad whose cells have undergone oviform degeneration. It is not a rudimentary ovary except in female animals. This is indicated by its presence in both sexes in toads; its presence in Spengel’s case of true hermaphroditism; by the fact that neither in male or female of toads do 204 THE AMERICAN NATURALIST [Vor. LVI its cells develop into true funetional eggs; and by its de- generate structure from its inception in both sexes. In a recent paper (Zoologischer Anzeiger, Dec., 1921) Harms deseribes marked hypertrophy of Bidder's organ following testis removal. He considers that castration of males causes Bidder’s organ to develop into an ovary. However, it should be noted that such operated animals with hypertrophied Bidder’s organ (ovary according to Harms) retain all their male secondary sex characters, and their normal mating instincts and that these male characters and instincts undergo a normal cyclical de- velopment in such induced ''females." When Harms removed both testes and Bidder’s organ the somatic sex characters and instincts failed to develop, showing clearly that Bidder’s organ in male toads acts like a testis in maintaining the secondary sexual characters. This is excellent evidence for the writer’s view that in male toads Bidder’s organ is simply a persisting embryonic male sex gland rudiment and not an ovary. If it is an ovary why should it develop and maintain the secondary sex charac- ters of the male in absence of the testis? 7. Recent investigators have inclined to the view that this structure is a hermaphrodite gland, ?.e., in male toads a rudimentary ovary, in females a rudimentary testis. If this is true then the admission is made that large, senescent, oocyte-like germ cells are not necessarily female cells — the crucial point for which the writer is contending. 8. Bidder's organ in Bufo corresponds to the larval gonad of frogs which in these forms disappears in the male and is replaced by the definitive testis. In the case of female anurans so far as the writer is aware no one has carried out a thorough investigation of the germ cycle from larval to fully adult life to see whether or not such a degeneration occurs in the female line. In mam- mals and birds such degeneration of the female embry- onic line of germ-cells is quite well established as the work of Winiwarter, Firket and others shows. No. 644] TRANSFORMATION OF SEX 205 The writer is of the opinion that it is only by adopting the view advanced here regarding the homologous nature of the larval male gonad of frogs, and Bidder’s organ in Bufo, that the problem of sex differentiation in anu- rans can be placed upon a rational basis. The theory accords with the embryological facts, covers the experi- mental finding of Witschi and others, accords with our own cytological data in the bullfrog, accords with the embryonic sexual conditions of other vertebrates, i.e., the degeneration of the embryonic line of germ cells in birds and mammals, and lastly furnishes an explanation of Bidder’s organ in Bufo. The key to the puzzle of sex development in frogs is simply this: every cell that superficially resembles an oocyte is not necessarily a female cell especially when occurring in an otherwise male individual, and that the larval male gonad of anurans is an organ of Bidder — a rudimentary embryonic sex gland with the same sex as the definitive gonad arising out of it. Misinterpretation of oviform hypertrophy and degeneration of racially senescent sex cells has rendered chaotic the problem of sex differentiation in anurans (see Plates 1 and 2). Witschi regards the development of certain somatic sex characters such as the Müllerian ducts as very positive evidence for his theory of sex transformation. e says: In males which show a typical development of the testicles, no Müllerian ducts of any significance are formed. On the other hand, such animals as first develop ovaries and later undergo the trans- formation of sex, also show regular oviducts; and these continue to grow just up to the time when the transformation of sex begins. This parallelism in the behavior of the Miillerian ducts and the gonads furnishes definite proof that the “eggs” and “ ovocytes,” described by the writer, are in fact really eggs and ovocytes and that the transformation of sex is a well-established fact. After the trans- formation of sex, when the ovocytes have disappeared, the Müllerian ducts begin to shrink but they do not disappear completely, etc., ete. The following data shows that in reality such so-called parallelism in the behavior of the Miillerian ducts and 208 — THE AMERICAN NATURALIST [Vor. LVI the gonads does not exist and that evidence based on such parallelism is worthless. In the normal males of adult Rana pipiens the Mül- lerian duets are remarkable for their size and degree of -= development. They arise as cellular cords in the peri- toneum at the time of metamorphosis and only acquire full development long after transformation when they come to resemble to a striking degree the oviducts of females (Fig. 2). In the larva of R. pipiens the so-called 2. Urogenital apparatus of adult Rana pipiens a the normal sees of the Müllerian ducts (md) in males of this species transformation of females into males (degeneration of the pro-testis and formation of the definitive testis) occurs very early in larval life, before the Müllerian duets ap- pear, and in this species the ducts undergo practically their entire development after the definitive testis has formed. In other words, while subjected to the influence of the fully formed testis and its ripening sex products the duets undergo the most marked development known in the males of any anuran species. Moreover, in Rana catesbeiana, where if we accept Witschi’s interpretation of femaleness, the so-called transformation of female in- No. 644] TRANSFORMATION OF SEX 207 dividuals into males is a prolonged process requiring two years, and where the future male larve are subjected to the so-ealled female influence during the entire period, the Müllerian duet does not appear. At metamorphosis when the definitive testes are fully formed and sperma- tozoa are beginning to appear the cellular cords repre- senting the vestigial Miillerian ducts of the male form but do not develop. If Witschi’s interpretation were correct, one would certainly expect to find marked de- velopment and hypertrophy of the Müllerian ducts in R. catesbeiana because of their being so long exposed to female influence. As a matter of fact, these structures in the male bullfrog are less developed than in other forms. The same criticism applies to the so-called develop- mental correlation of the Miillerian duct with the gonad of the same side in cases of lateral hermaphroditism. What Witschi terms lateral hermaphrodites are nothing more than larve or young frogs which show the defini- tive testis developing out of the pro-testis (larval male Bidder’s organ) faster on one side than on the other. (See Witschi, Am. Nar., page 533.) In the end such ani- mals develop into definite males with testes symmetri- cally formed. True lateral hermaphroditism in adult frogs is an exceedingly rare phenomenon. In the wri- ter’s material it is rare to find both definitive testes de- veloping out of the pro-testes at the same rate, one gland may be the finished gonad, the other the pro-testis un- dergoing degeneration. Such larve are in no sense to be regarded as lateral hermaphrodites. There is no de- velopmental correlation of the Miillerian ducts with the gonad of the same side in R. catesbeiana and R. pipiens, because there are no ducts formed until after the defini- tive testes are formed. Regarding the other somatic sex characters such as seminal vesicles and thumb cushions, it should be pointed out that the thumb pad in R. cates- beiana is not formed until after metamorphosis when Ev Tal Ww. Cep MAS MANT. Post SUM » n. we Pee D s m i ree PLATE I Fic. 3. So-called oocytes cecurring in the degenerating pro-testis of larval bullfrogs. These cells according to a's the writer’s view are merely hypertrophied iform erat o; set of degeneraticn. At X e giant sperma atocytes lopoda). The ce i rm functional atezoa and make up the pocius part of the testes. Note ie * germinal Tuto " condition of the nucleus PLATE II ad 6. a a of Scolcpen "IG permatocytes of eias (Chilopoda). The resemblance to oocytes m the germinal vesicle stage is remarkable. Sections of the testes ook like All photographs on Plates I and II made at a magnification d Er diameters, No reduction. Figures 5-8 are from Professor Blackman’s mate 210 THE AMERICAN NATURALIST [Vor. LVI the fully formed testes are present, and the seminal ves- ieles are absent or rudimentary in the males of many frog species, and exceedingly well developed in others. In the few cases reported of true lateral hermaphro- ditism in adult frogs there is not always a developmental correlation of the Müllerian ducts with the gonad of the same side. Crew (’21), (Journal of Genetics, Vol. II, no. 2) has summarized the known cases of sexual ab- normality in amphibians — see Figs. 7, 8, 9, 12, 14, and 16 of this paper, also the report of cases 21, 22, 23, 24, and 39. These are exceptions to any rule of develop- mental correlation. In several eases, Figs. 25 and 31, the duets are quite as well developed in total absence of ovarian tissue as when such is present in large amounts, this, of course, being the normal condition in Rana pipi- ens. Crew also gives a list of frog cases reported where both gonads were entirely missing and yet the Müllerian duets were well developed in such individuals. . Because of these facts it is fair to conclude that the appeal to the somatic sex characters completely fails as proof of the transformation of female frogs into males. In closing, it should be pointed out that Witschi has made but one original investigation of sex in anurans (Witschi "14, no. 1). His later papers on the subject con- tain no new observations or experiments but are purely speculative endeavors to interpret his early work in ac- cordance with Mendelism (’14, no. 2), later (720, no. 3) in accordance with internal secretions. THE SEX-LINKED GROUP OF MUTANT CHAR- ACTERS IN DROSOPHILA WILLISTONI REBECCA C. LANCEFIELD AND CHARLES W. METZ STATION ron EXPERIMENTAL Evotution, Corp Spring Harpor, N. Y. INTRODUCTION Tus present work was undertaken for the purpose of comparing the genetical behavior of the fruit-fly Droso- phila willistoni with that of Drosophila melanogaster and other members of the genus. It deals with the 28 sex- linked mutant characters thus far studied. The non-sex- linked characters will be considered in another paper. Drosophila willistoni Sturtevant (D. pallida Williston)! is not unlike the well-known D. melanogaster in habits and superficial appearance, but a detailed examination reveals numerous features in which it differs from mela- nogaster. Among these are the following: (1) absence of sex combs in the male, (2) six instead of eight rows of acrostichal hairs on the thorax, (3) smaller size and more slender form, (4) vermilion instead of red eye color, and (5) narrow instead of broad bands on the abdomen. This species has been chosen for the present study be- cause it is one of the species having the same general type of ehromosome group as D. melanogaster. It will be recalled that within the genus Drosophila at least eleven different types of chromosome groups are represented (Metz, 1916). The most common type is that called type A (Fig. A, present paper), which is found in 13 of the 29 species studied. In these 13 species (which include melanogaster and willistoni), the chromosome groups are so much alike as to suggest that similar chromosomes are homologous and earry homologous groups of genes throughout. On the other hand, the species themselves do not form a restricted taxonomie group, but seem to be 1 See Sturtevant, 19215, for description, ete. ait 212 THE AMERICAN NATURALIST [Vor. LVI scattered more or less at random through the genus — which does not conform to such a view unless this type of chromosome group be considered primitive and the forerunner of several other types. These eonsiderations indicated the need of a compara- tive study of different species possessing this type of chro- mosome group, in addition to the studies already being made on species having different types of groups. Since the species can not be hybridized (or have thus far refused to hybridize — with one exception considered below), it is necessary to make cytological and genetical studies of them individually. This of course limits the comparison to a very few species. The -present paper supplements our previous one (Lancefield and Metz, 1921) on the sex chromosome re- lationships of willistoni and melanogaster, in which it was shown by means of non-disjunctional flies that the sex chromosomes are not strictly homologous in the two species. In melanogaster the short, rod-like pair is the sex chromosome pair (Bridges, 1916), whereas in willis- toni we find that the rod-like pair is an autosome pair, and that one of the large V-shaped pairs is the sex chro- mosome pair. This relationship is shown in Figs. A and B. on i A B Fics. A and B. Diagrams of female chromosome groups of Droscphila me- lanogaster and Drosophila willistoni 1espectively. The X-chromosomes are rep- resented in solid black, the autcsomes in outline. In D. willistoni the small, dot-like pair may be absent. The genetical study considered here is for the purpose of comparing the constitution of the sex chromosomes by No. 644] MUTANT CHARACTERS IN DROSOPHILA 213 means of the sex-linked characters and their linkage re- lations. The stock of D. willistoni which we have used was brought from Cuba in 1915, and was kept in the labora- tory without being studied until 1919 when the present work began. In making the tests for linkage and in calculating cross- over values, the usual procedure has been followed? We have been concerned particularly with determining the relative order of the genes and the approximate amount of crossing over between them, but not with obtaining exact crossover values. In consequence, the ‘‘ chromo- some map °’ given here is to be considered as indicating only the approximate location of the respective genes. In presenting the data, the mutant types are described, not in ehronologieal order, but in such a way as to follow the serial order of the genes on the chromosome map. All of the sex-linked characters are recessive. The data on the origin of mutants are necessarily imperfect, and in some cases are very meager, owing to the fact that many of the mutants appeared in stock cultures, mass cultures, ete., for which no complete records were taken. In such cases the available data are given briefly under the appropriate headings. We are indebted to Mr. D. E. Lancefield for carrying out some of the early experiments, and for finding the mutants ‘‘rimmed’’ and ‘‘nicked.’’ Similarly, we are in- debted to Miss Ruth Ferry for the mutant ‘‘ yellow ’’ and for carrying out the experiments involving ‘‘ yellow.” To Dr. A. H. Sturtevant we owe many valuable sugges- tions regarding the comparison of mutant characters in D. willistoni with those in D. melanogaster and D. simu- lans. We are also indebted to the following persons for making the accompanying drawings: Miss Ruth Lincks — Figs. 1, 3, 7, 8; Mrs. D. B. Young — Figs. 2, 4, 5, 6, 17; Miss E. M. Wallaee — Fig. 9, and Miss E. D. Mason — Figs. 10 — 16. 2 This has been deseribed in earlier papers by Morgan and others, and is to be found in eurrent books on geneties. 214 THE AMERICAN NATURALIST [Vor. LVI DESCRIPTION AND ORIGIN? or MUTANT CHARACTERS Stubby (sy) Description.—Stubby is a bristle character, manifested by all of the thoracic and head bristles (Fig. 2). These are usually shortened, thickened, and somewhat curled, and often are split or forked at the tip. The two pos- terior scutellar bristles are frequently tightly twisted to- gether and point anteriorly. The short, thick appearance Lh d Vs. j | ih \ a E ‘Fig. 1.: Drosophila willistoni, “normal.” of the bristles is never apparent in combination with small-bristle, but the charaeter ean always be distin- guished by the forking of the sternopleural bristles. Both sexes are fertile. Stubby looks very much like forked in anelanogaster. |. Origin.—One stubby male was obtained from a pair mating. No complete record of this culture was kept, however, and it is not known whether others appeared previously or not. 3 See Table I. No. 644] MUTANT CHARACTERS IN DROSOPHILA 215 Orange (o) Description.—The eyes are orange colored. In newly hatched flies, the color is a pale lemon, which deepens into orange as the fly matures and may become very dark in old age. The color resembles garnet or coral of D. melanogaster. Fie. 2. Stubby, bristles. Origin.—One male appeared among the offspring of a mating of three normal females by an unrelated male. The other offspring were all normal, but their number was not recorded. Presumably, the mutation occurred in one of the P, females and affected only one or a few germ cells, although it is possible that this female was hetero- zygous and produced a very small number of the flies in the culture. Small-bristle (sb) Description.—All the bristles are more slender and somewhat shorter than in normal flies. The character is extreme when orange eye color is also present. 216 THE AMERICAN NATURALIST [Vor. LVI Origin.—One small-bristle, forked male was obtained from a mass eulture carrying rough and forked. More than a year later, a seeond mutation to small- bristle occurred in an entirely unrelated stock. In this case the single small-bristle male found among the off- spring of one pair was crossed to a female from a ho- mozygous orange small-bristle stock and produced only small-bristle flies. Bent (bn) Description.—The wings of bent flies are slightly spread out, and are bent at the base so that they slope down toward the body (Fig. 3). They are often slightly crum- pled. The flies hatch as well as their normal sibs but do not breed as readily. Fic. 3. Bent, wings. Origin.—Many bent males were found in a culture of five orange females from stock mated to a single male of different parentage. At least one of the females carried bent, but the exact origin is uncertain. No bent flies were ever observed in orange stock. Forked (f) Description.—In forked flies, all the bristles are wavy with the ends sometimes forked. The females are sterile. This character is similar to, but less extreme than, stub- No. 644] MUTANT CHARACTERS IN DROSOPHILA 217 by. It recalls singed, of melanogaster although singed, is slightly more extreme. an ae Fig. 4. Forked-2, bristles. Origin.—Seven males were found in a stock bottle. Since no forked females were obtained, it is possible that this was the original appearance of the mutant and that all of the forked flies were from one mother, heterozygous for forked. Forked-2 (fə) Description.—This character is much more extreme than its allelomorph, forked, or the similar mutant, stub- by. The bristles are twisted and thickened with their ends often split (Fig. 4). The twisting also affects prac- tically all the hairs on the fly, including those on the inner margin of the wing. The hairs on the antenne are forked. Forked-2 resembles the melanogaster singed. The fe- males are sterile. Origin.—Several forked-2 males and one female ap- peared in a stock bottle of small-bristle flies. 218 THE AMERICAN NATURALIST [Vor. LVI Tiny (t) Description.—Tiny-bristle flies usually have small an- terior dorso-central bristles. Sometimes the two anterior scutellar bristles are also small. Occasionally all the bristles may be small, so that the individual may be in- distinguishable from a true ‘‘ small-bristle’’ fly. The character is rather variable and, in many cases, is very hard to separate from normal. This difficulty was so great that the stock was finally discarded. Origin.—A single male was found on the last count of the offspring from a pair mating. It is possible that other tiny males were present among the previous off- spring and escaped observation since the character is very inconspicuous, Square (sq) Description.—The wings are about two thirds the nor- mal length with the ends almost square instead of pointed (Fig. 11). A characteristic slight wave extends through- out the length of the wing. The females are sterile and the viability of the males is rather poor. The description of square suggests that of rudimentary melanogaster but square is much less extreme than rudi- mentary; the wing is not shortened so much and is not eut off so squarely. Origin.—Among the offspring from a pair mating (rough female by orange rough stump male) several square males were found, indicating that the mother was probably heterozygous for square. Reduced (re) Description Reduced flies regularly lack the two an- terior dorso-central bristles; occasionally, they also lack one of the posterior dorso-centrals; and less frequently, all four are absent. In combination with seute-2, how- ever, the more extreme condition of reduced is frequently found (Fig. 7). The reduced gene also affects the shape No. 644] MUTANT CHARACTERS IN DROSOPHILA 219 of the abdomen, which is blunted, or apparently com- pressed along the anterior-posterior axis. The abdominal bands are slightly irregular. Reduced flies, especially females, are hard to breed in pairs, although those which do produce offspring seem normally fertile. i Origin.—Many males were found among the offspring from a pair mating from orange stock. The mother of the eulture was apparently heterozygous for the gene. THE Scute ALLELOMORPHIC SERIES 1. Scute (sc) Description.—The two anterior scutellar bristles are usually lacking, although occasionally only one may be gone. Rarely the combination of one anterior scutellar bristle and one posterior one may be found. The remain- ing bristles are normal in size. The character almost always manifests itself in homozygous flies. Only one exception to this has been detected up to the present time. Origin.—Fifteen scute males and eleven normal males ` were obtained from a normal pair. The female offspring were all normal (number not recorded). It is almost cer- tain that the mother was heterozygous in this case, and that the mutation occurred in a previous generation or else very early in her own ontogeny. 2. Scute-2 (scz) Description.—Seute-2, an allelomorph of scute and seute-3, involves the same seutellar bristles as scute, but varies toward a more extreme condition than this allelo- morph. The two anterior scutellar bristles are always missing, frequently one of the posterior scutellars is gone, and oceasionally all four are lacking. The bristles on the scutellum are fine and small. In a stock homozygous for reduced and scute-2, both characters are more extreme than either is alone (Fig. 7). Such flies often entirely lack dorso-central and scutellar bristles, and lack one or more orbital bristles. 220 THE AMERICAN NATURALIST [Vor. LVI The compound seute seute-2 females are either some- what intermediate between the two, or they may look en- tirely like one or the other component. On the whole, they are more apt to resemble scute-2 than scute. Origin.—F rom a normal female mated to a scute male, the following types of offspring were obtained: males — - one half normal, one half seute-2; females — one half normal, one half somewhat intermediate between scute and seute-2. From this it was concluded that the parent female was heterozygous for the new character and that this character was allelomorphie to scute, a conclusion subsequently verified by direct tests. Six sisters of this female were also tested and none gave scute-2. 3. Scute-3 (sc) Description.—Scute-3 is an allelomorph of seute and seute-2. All four seutellar bristles, the two sterno-pleural bristles, and a varying number of head bristles are ab- sent. On the head, all three pairs of orbitals are usually. missing, and occasionally some of the others are gone. The compound females involving scute-3 and scute-2 are more apt to be like scute-3 than like seute-2, although ` in general they are intermediate. Such females could be distinguished from the homozygous seute-2 females in all the eases observed by the absence of at least one sterno- pleural bristle and generally by the absence of all seutel- lar bristles. Seute-3 males are sterile. | Seute-3 strongly resembles the scute of melanogaster. In both cases scutellar and head bristles are affected. Two stocks of melanogaster scute kindly examined by Dr. Sturtevant agree with scute-3 in lacking scutellar bristles and orbitals, and in having small ocellar bristles. They both possess post-orbitals, however, and one stock occasionally shows the middle orbital present. The other usually lacks the postverticals. : Origin.—Scute-3 was first observed in the offspring of an F, female from a cross of a seute female from stock by two rough rimmed stump males. This female seemed to No. 644] MUTANT CHARACTERS IN DROSOPHILA 221 be heterozygous for the new factor, and it was found that the character was also present in males in sister eultures which had been used for stocks. Yellow (y) Description.—In '* yellow ” flies the body, wings and legs are deep yellow. The bristles and hairs are all yel- lowish or bronze instead of black. In the latter respect yellow differs from the yellow in Drosophila virilis which has black or dark brown bristles and hairs. Origin.—A single yellow male appeared in a bottle of scute rough stump stock. Yellow was found after the main part of this paper was prepared for publication, and since the experiments involving it have not added materially to the data given in the tables they are omitted from the latter and are given briefly here. The original yellow seute rough stump male was mated to normal females giving a normal F,. The latter, inbred in pairs, gave 1354 normal daughters and the following classes of sons: normal 488; yellow scute rough stump 466 (non-erossovers 954); yellow scute 25, rough stump 23 (single crossovers in region two 48); yellow scute rough 49, stump 46 (single crossovers in region three 95) ; yellow scute stump 3, rough 2 (double crossovers involv- ing regions two and three, 5). In addition, two yellow rough stump males and one yellow stump male were ob- tained. Of the former, one proved to be genetically scute when tested and hence should be in the non-crossover class. The other gave no progeny, but presumably was also a non-erossover. The third fly likewise failed to breed, but since it lacked rough as well as seute it pre- sumably represents a double crossover in regions one and three. Itis this fact which leads to the tentative location of yellow above rather than below seute on the map.* 4 This is supported by subsequent data. 222 THE AMERICAN NATURALIST [Vor. LVI Peach (p) Description.—Peach eye color is practically indistin- guishable from orange eye color, although, as a rule, it is a trifle darker than orange and does not have the range of shades due to age which are observed in orange. In the same culture, it is impossible to distinguish the two with certainty. The double recessive of peach and orange is probably indistinguishable from either eye color alone. Homozygous peach rough flies have darker eye color than orange rough flies, and are hard to separate from rough alone. Peach eye color is similar to ruby and garnet of melanogaster and to rubyoid and carmine of simulans. Origin.—A. single male with peach eye color was found in a double recessive forked stump stock. Beaded (be) Description.—Beaded refers to the condition of the wings, which have the marginal hairs clumped in irregu- lar patches, especially on the posterior half of the outer margin. The wings are pointed at the ends due to a long notch, extending from the tip of the third vein to about Fie. 5. Normal, wing. Fie. 6. Beaded, wing. No. 644] MUTANT CHARACTERS IN DROSOPHILA 223 the region of the posterior eross-vein, and to the loss of a section from the outside of the wing between the distal ends of the second and third veins (compare Figs. 5 and 6). Beaded flies have poor viability, and the females are sterile. Beaded is similar in appearance to the melano- Fic. 7. Reduced scute-2 compound. gaster cut, although the latter is slightly more extreme than beaded. While cut, flies are vigorous and fertile, some of the cut allelomorphs are not completely fertile and have poor viability.® Origin.—An out-crossed female, known to carry small- bristle rough on one X-chromosome, and small-bristle orange short-3 on the other, produced offspring in which the small-bristle rough males were also beaded. This fe- male was almost certainly heterozygous for the new gene. Nine sisters were bred separately but no beaded flies ap- peared in their offspring. 5 Unpublished data for which we wish to thank Drs. Mohr and Bridges. 224 THE AMERICAN NATURALIST [Vor. LVI Rough (r) Description.—Rough eye affeets mainly the surface of the eye (Fig. 8). When the outer portion of the eye is mounted and examined under the high-power microscope it is seen that the ommatidia are irregular in shape and size with uneven surfaces which are more convex than the normal. The normal eye has regular hexagonal faeets with a bristle at every alternate intersection of the sides (See Carnegie Publ. 278, Plate 10, Fig. 3c for the normal eye of D. melanogaster which has the same arrangement). The bristles of rough eye are irregularly distributed with groups collected in one place and no - bristles at all in another. These bristles are about one and a half times the length of the normal ones. The roughened condition is similar to that found in star eye of melanogaster (Carnegie Publ. 278, Text-fig- ure 83). The eyes of willistoni rough are also somewhat glossy in texture, and the wing veins are slightly heavier than in the normal flies. Origin.—Several rough males and females were found in one of the bottles of a stock that had been kept in the laboratory for approximately four years. It is probable that the mutant gene had been present in the stock for . some time. : Triple (tr) Description. — Triple causes four variable wing changes, one or all of which may be present in either or both wings (Fig. 10). (1) The second and third veins may be fused for a short distance at their origin. (2) The wings, slightly tilted up at the ends, are held away from the body at an angle which varies up to about 90°. (3) The third veins fail to reach the distal margin of the wings by amounts which vary from almost nothing to one third the length of the vein. This is particularly evident in the females, where a large section may be missing from the central part of the third vein. (4) An extra cross-vein is present between the second and third No. 644] MUTANT CHARACTERS IN DROSOPHILA 225 veins at a level about half way between the anterior and posterior eross-veins. This vein, when not wholly formed as a cross-vein, is often indicated by short pieces of dis- connected vein. The fusion of the second and third veins and the ex- tension of at least one wing are constant characters as far as observed. The latter forms the easiest basis of distinguishing this mutant. The females are more ex- treme than the males in all four of the changes involved. Triple suggests the melanogaster mutant, bifid, by its extended wings, fusion of the veins at the base of the wing, and the shortening of one of the veins, although the short vein is the fourth in bifid and the third in triple and the third vein of bifid is thickened at the distal extremity. ! Origin.—Triple was first noticed in the offspring of a female out-crossed from wild stock. Half the males were triple. On investigation it was found that several bottles of wild stock contained similar males. Triple males were found six months later in stump stock. Crossed to the original triple stock, these males produced triple female offspring in the F, generation. The possibility of contamination of the stump stock can be eliminated since the triple males found in the stock were also stump, and there were no cultures containing triple stump flies in the laboratory. Tug DEFORMED ALLELOMORPHIC SERIES 1. Deformed (d) Description.—Deformed, which involves many parts of the body (Fig. 9), shows sexual dimorphism. In the male the eye is about two thirds the normal size and very rough; in the female the eye is normal in size and only slightly roughened. In both sexes the bristles are ab- normally long and irregularly bent. The thoracic hairs are badly disarranged in both sexes, but the effect is very much more exaggerated in the female than in the 226 f THE AMERICAN NATURALIST [Vor. LVI male. In the former, the hairs are often clumped in a compact mass on the anterior half of the thorax; while in the male, the hairs are irregularly scattered over the Fic, 8. Reugh (eye), rimmed (wing margin), stump (second wing vein) compound. whole area. The scutellum in the male (sometimes in the female also) is blunt instead of pointed, posteriorly, and the under portions of the thorax are consequently visible. The wings are extended at an angle of about 45° to 90° in both sexes, and the veins are often faint, short, and irregular, especially in the male. These flies are rather feeble and breed poorly except in mass culture, probably on account of the many physi- eal defects present. Origin.—Many males and females were found in a stock culture of orange forked rough. Sister bottles made up at the same time did not produce any deformed flies. No. 644] MUTANT CHARACTERS IN DROSOPHILA 227 2. Serrate (st) Description.—Serrate is an allelomorph of deformed, but involves only a part of the characters modified by deformed. The changes in the eyes of the two sexes are exactly the same as those caused by deformed. On the other hand, the bristles and the shape of the scutel- lum are almost normal and the thoracic hairs are only slightly disarranged. The wings may occasionally be held at an angle with the body, but the venation is prac- tically normal. The only strikingly noticeable effect of serrate is the change in the eyes. Fic. 9. Deformed q. The compound deformed-serrate females are inter- mediate between the two allelomorphs but tend to re- semble serrate more closely than deformed. Serrate flies are more viable than deformed and breed more readily. Origin.—A single male was found in an F, mass eul- ture from a mating of two females by one male from seute stock. No other serrate flies appeared in this cul- ture or in a sister culture. Rimmed (ri) Description.—In rimmed flies, a heavy rim of marginal hairs surrounds each wing and the wings curve down 228 THE AMERICAN NATURALIST [Vor. LVI over the abdomen as if the margins were constricted (Fig. 8). 'The depression between the seutellum and thorax of normal flies is eradicated, leaving a smooth surface at the junction. The thick marginal rim of hairs is the most constant of the effects of rimmed, but the other changes are usually apparent. Origin.—Several males were found in wild type stock. Pale (pa) Description.—The post-vertical and all the thoracic bristles are pale yellow. Occasionally, a few other head bristles are also yellow. The bristles are thin, and the entire fly is weak and small with the wings often not un- folded. None of the original pale flies could be induced - to breed. The heterozygous females produced a few pale offspring, but the stock was soon lost. Origin.—'The mother of the culture in which pale ap- peared was heterozygous for seute rimmed on one X- chromosome and for pale morula on the other. Five sis- — ters of this female were bred, but no pale offspring were obtained from any of them. It is impossible to tell whether the mutation to pale occurred in the mother of the eulture in which pale was found or whether it took place in her mother. Stump (s) Description.—The distal portion of the second vein is lacking, leaving only a stump at the base of the wing (Fig. 12). This stump varies in length from one quar- ter to two thirds that of the normal vein. Origin.—Six stump males were obtained from a mass culture in which the mothers were heterozygous for forked and the fathers were normal. Four of the stump males were also forked, the others were not. THE SHORT ALLELOMORPHIC SERIES 1. Short (sh) Description.—Typieally, all the veins of the wing fail to reach the margin in short flies, although No. 644] MUTANT CHARACTERS IN DROSOPHILA 229 sometimes the third vein is entire (Fig. 18). The second vein is shorter than any of the others. Ordinarily, the distance between the ends of the veins and the margin is not great. The posterior cross-vein is broken occasionally. 15 M 16 Fies. 10-16: Fig. 10, triple. Fig. 11, square. Fig. 12, stump. Fig. 13, short ¢. Fig. 14, short-2 P Fig. 15, short-3 j. Fig. 16, nicked. Origin.—Short was first observed in half the sons of a single female indicating that the mother was hetero- zygous for the short gene. This female carried orange * 230 THE AMERICAN NATURALIST [Vor. LVI rough on one X-ehromosome and stump on the other. The mutation to short evidently affected a locus of the orange rough ehromosome not far from the stump locus. Several sister eultures were examined, but no short flies were found. 2. Short-2 (sh,) Description.—Short-2 -is the most extreme of the series. In the females the second, fourth, and fifth veins are very short, the fourth and fifth often-not reaching as far as the posterior eross-vein. In eases in which they extend beyond the éross-vein, this vein is broken. In the males the fourth and fifth veins are about three quarters the normal length, and the posterior cross-vein is broken (Fig. 14). The males are indistinguishable from those of short-3. Origin.—A. single male was found in small- bristle rough Stock: Con 3. Short-3 (shay. Description.—Short-3 is about the same in both sexes (Fig. 15). The second vein is very short, and all the others are about three quarters the normal length. The posterior cross-vein is usually broken. Females containing any two of these three allelomorphs are intermediate between the two used, with perhaps a closer resemblance to the more extreme member of the pair. Origin.—Several males were found in seute stock. ` Morula (m) s Desoto —Morula involves a partial ut of the eyes which is due to a consolidation of a group of facets, especially in the central area of the eye, suggest- ing the lozenge of melanogaster. The viability of this stock is very poor, and the double recessives of morula and any other mutant character rarely survive. Origin.—At least five males were obtained from a mass | culture of three pairs which carried rough. The classi- No. 644] MUTANT CHARACTERS IN DROSOPHILA 231 fication of rough and morula was not accurate at its first appearance. : Nicked (nk) Description.—Nieked is characterized by irregular notches in one or both wings (Fig. 16). The indenta- tions vary in size and location, but tend to show on the posterior and inner portions of the wing. In certain lines, the character shows regularly, while in certain other lines it overlaps normal a great deal. Flies in which nicked is combined with other mutant characters have rather poor viability. Origin.—Several individuals of both sexes were found in a mass culture. Rosette (ro). Description.—In this mutant race, a large number of characters are affected (Fig. 17). The eyes are slightly roughened due to disarrangement of the facets; the thoracic hairs are disarranged, looking as if they had been brushed in the wrong direction; the bristles may be bent; the distal tarsi of the legs may be twisted; and the wings are generally held at an angle with the body, and one or both may be small and circular. The rough eyes and disarranged hairs are constant characters. Rosette flies have very low viability and are hard to breed, especially when other mutant characters are pres- ent also. Origin.—Four rosette rough males were obtained among a large number of offspring from one morula male by three rough females. CONSTRUCTION or THE X-CHROMOSOME ‘‘Map’’ ' With one or two exceptions the usual procedure? has been followed in constructing the chromosome ‘‘map.’’ The order of the genes was determined by means of erosses involving three or more loci (Tables III-VI), and that order adopted which, in the consideration of any three points, made the double crossover class the small- ? See footnote, p. 213. 232 THE AMERICAN NATURALIST [Vons LVI est. In most cases the decision was confirmed by several subsequent experiments made for other purposes. Tiny and square have not been definitely located with refer- ence to forked since they proved unsuitable characters for use in linkage experiments. Similarly, the loci of triple and deformed are known to be between rough and . rimmed, but the relative positions of the two could not be determined on account of the impossibility of using the two characters together. Pale, morula, and rosette are also only tentatively placed. Fic. 17. Resette. With the order of the genes established, the ‘‘dis- tances,’’ or crossover values, between them were obtain- ed by combining the data from all the experiments given in Tables II-VI. Table VII gives the summary of all data between any two loci in the ‘‘map’’ from those ex- periments in which no intermediate genes were concerned. This differs from the usual method of summarizing the data in that it includes only experiments in which no intermediate point was used. No. 644] MUTANT CHARACTERS IN DROSOPHILA 233 As far as possible the po- sitions of the genes on the ‘‘map’’ have been determined by summation of the ‘‘dis- tances’ between neighboring loci taken in pairs, using stubby arbitrarily as the zero point. In several cases, how- ever, the locus of a gene has been assigned by reference to some main well-established point; notably, orange, forked, seute, rough, or stump. In Table VII, the starred data are those on which the ‘‘map”’ is based. No correction for data involving non-adjacent loci has been made, since the degree of numerical accuracy does not warrant such a com- putation in this ease. No cor- rection has been made for double crossing over in long regions in which no mutant loci are known. Owing to the possible par- allelism between the yellow and seute in willistoni and the yellow and scute in mela- nogaster a second set of readings has been given on the ‘‘map’’ using the position of yellow as the zero point and plotting the others in opposite (+ and —) directions from this. Comparison with the X-chromosome map of mela- nogaster is thus facilitated. +427] 84? rosette (ro) +30 —L.72 nicked (nk) +28? _ 1.70? morula (m) +26 | 68 short etc. (sh) 418 —] 60 stump (e) +142__| 56? -pale (pa) ib um u LES Ze a de wn v) eh ~ beaded (be) LAS P (p) 42 At =11? i31? square (eq). -12 —f—30 forked etc» (f) =14? |28? tiny (t) -32 ——l]-10 bent (bn) dT uus 9 enall-bristle (sb) -5 0.7L. 1.3 orange (0) 2 —} O stubby (sy) Fic. 18. Map showing linkage relations of sex-linked genes in Drosophila williston 234 |^ THE AMERICAN NATURALIST [Vor. LVI Discussion The previous work on the comparative genetical study of different species of Drosophila has been concerned largely with species having different types of chromo- some groups. It has involved mainly the species melanogaster, virilis, funebris, simulans and obscura. Of these, only melanogaster and simulans have the type of chromosome group with which we are concerned here. The published data on the first four of these species have recently been summarized by Sturtevant (1920) and may be passed over briefly. The data on obscura are in press and our references to them are made with the kind permission of Mr. D. E. Lance- field. A In melanogaster, virilis, funebris and obscura, the evi- dence suggests a tendency on the part of each species to give mutants paralleling those in the others, although the extent of this tendency ean not be ascertained ac- curately because of the impossibility of proving the ho- mology of similar characters. In the case of melano- gaster and simulans the parallelism extends to nearly all of the known simulans characters and certain homol- ogies are established by means of hybridization (Sturte- vant, ’20, *21a, 21b). To be sure, the two latter species are almost identical and would be expected to give similar genetical results; but it is of interest to note that there is a close resemblance between the proven cases of par- allel characters in these, and the apparent cases of par- allel characters in the other species. This tends to in- crease the probability of actual parallelism in the latter where a series of linked characters is involved. Upon comparing the mutant characters of willistoni with those of the others it is evident that only a few striking cases of resemblance are found. Of these the most significant involve the characters yellow and scute. Their morphological resemblances to the yellow and scute in melanogaster have already been noted in the No. 644] MUTANT CHARACTERS IN DROSOPHILA 235 descriptive section. But the evidence for their being parallels is made particularly strong by the fact that their genes are completely or almost completely linked in both species. In melanogaster yellow and scute are - located at the extreme end (zero point) of the chromo- some map, while in willistoni they are approximately in the middle. A situation similar to this has already been found in D. obscura (according to unpublished data of D. E. Lancefield). Here the characters yellow and scute also bear a close resemblanee to those in melanogaster and are very closely linked. As in willistoni, the factors for yellow and seute are near the middle of the chromosome map. It will be recalled that obscura, like willistoni, has a large V-shaped X-chromosome—although the other chromosomes are different (Metz, 1916). In the two species having. V-shaped X-ehromosomes, then, yellow and scute are ‘‘located’’ near the middle of the chromo- some map, while in melanogaster with its short, rod-like X-chromosome, yellow and seute are at one end. As Lance- field has pointed out in his discussion of obscura, this sug- gests that one end of the rod-like X in the one ease cor- responds. to the middle of the V-shaped X in the other. And this suggests that the rod-like chromosome itself may correspond to one arm of the V. The only evidence in willistoni on the latter hypothesis is that furnished by the characters forked and stubby. These are possible parallels of the singed and forked in melanogaster. They are similar in a general way in the two species (see description above), and the serial order of the genes is the same (Fig. 18), although the linkage relations do not agree exactly. In this connection it may ‘be recalled that **yellow,"" **singed" and ‘forked’? have also been found in Droso- phila virilis (Metz, 1918, and unpublished data), and may, likewise, be considered as possible parallels to those in melanogaster. Virilis has a rod-like X-chromosome resembling that of melanogaster, and the relative posi- 236 THE AMERICAN NATURALIST [Vor. LVI tions of the three genes on the chromosome map re- semble those in melanogaster. Yellow is about three units from the end instead of at the end; singed is at about 35 instead of 21 and forked is at about 58 instead of 56.5. The evidence is not sufficient to warrant the conclusion that these are actually homologous series, but the fact that such series exist suggests that by the present means it may eventually be possible to obtain reliable data for a comparison of the chromosomal make-up of the different species. Among the other characters in willistoni which show some resemblance to characters in melanogaster or sim- ulans the following may be noted as a matter of record, although there is little indication of their being actual parallels: orange and peach (which look alike) resemble coral or ruby; beaded is similar to the cut allelomorphs both morphologically and in respect to its sterile fe- males and poorly viable males; triple suggests bifid, and morula looks like lozenge. The small bristle of willistoni may be comparable to the tiny bristle-2 of simulans. The fact that the X-chromosomes in willistoni are mor- phologically like the large autosomes and not like the X-chromosomes of melanogaster suggests that we ought to compare, not only the sex-linked groups of the two species, but also the sex-linked group of willistoni with the non-sex-linked groups of melanogaster. This has been done, but without revealing any significant indica- tion of parallelism. In conclusion it may be noted that although the evi- dence is not yet clear on the genetic relationship of the sex chromosomes in melanogaster and willistoni, yet if the above suggestion is correct, that the X-chromosome of melanogaster corresponds to part of the X-chromo- some in willistoni, then the resemblance ‘between the chromosome groups of the two species is only super- ficial. It may also be noted that the genetic ‘‘map’’ of No. 644] MUTANT CHARACTERS IN DROSOPHILA 237 the X-chromosome of willistoni at present is only slightly longer than the map of the melanogaster X-chromo- some (84 as contrasted with 70 units), whereas the wil- listoni X-chromosome itself appears to.be about twice the length of that of melanogaster. This suggests that cross- ing over is less frequent in willistoni than in melano- gaster. SUMMARY 1. Twenty-eight recessive sex-linked mutant charac- ters in Drosophila willistoni are described and their linkage relations considered. 2. In general, the genetic behavior of willistoni (as re- gards crossing over, etc.) is similar to that of D. mela- nogaster and the other species of Drosophila whose ge- netic behavior is known. 3. There is a strong indication of parallelism between the mutants yellow and scute in willistoni and yellow and scute in melanogaster. 4. In both species these characters are completely or very closely linked. 9. There is some indication of series between the characters forked and stubby in willistoni and singed and forked in melanogaster. 6. In melanogaster the genes for yellow and scute are “located”? at one end of the chromosome map, and singed and forked are 21 units and 56.5 units respec- tively from this end. In willistoni yellow and scute are near the middle of the map, and forked and stubby are on one side at 12 units and 42 units respectively. 7. Since the X-chromosome of melanogaster is short and rod-like, while that of willistoni is approximately twice as long and is V-shaped, this relation of the chro- mosome maps suggests that the melanogaster X-chromo- some corresponds to one arm of the V-shaped X-chromo- some of willistoni, with the locus of yellow correspond- ing in the two cases. This agrees with the suggestion made by Lancefield in the case of D. obscura in which the X-chromosomes resemble those of willistoni. 238 THE AMERICAN NATURALIST [Vor. LVI 8. The comparative lengths of the X-chromosome maps in melanogaster and willistoni suggests that there is less crossing over in the latter than in the former. TABLE I ORIGIN OF MUTANTS Ezplanation of ‘‘records’’: W indicates R. C, Lancefield records on numbers 1-100 whieh fadiesin D. E. Laneefield; L indicates C. W. R indicates ius Ferry. Mutant Sym- First Record Parts Affected bol Found 1. Stubby.......:.| sy | March, 1920, W 1128 |Bristles. COCA ORAE. ois ve ok o April, 1919 L 37 Eye color. Š July, 1919 | L 336 8. Small-bristle sb Uus: 1920 | W 1745 Bristes 4. Bent. :. s. bn | Nov., 1920 W 1687 D. Forked -> oc f March, 1919| L1 ed fertility of females. 6. Forked—2.....| f» March, 1920 | W 1177 |Bristles, hairs; fertility of 6 TINI Ursa t Jan., 1920 W 856 |Bristles. 8. Square.........| sq |Feb., 1920 | W965 |Wings; fertility of 9 9. 9. Reduced. ...... re | Oct., 1919 W 288 Usher abdomen. 10. Bonto i1 sd e May, 1919 L 231 Bri stles 11. Scute—2....... Sc? | Feb., 1920 WwW ristles 12. Seute—3....... sca | May, 1920 W 1346 Bristle; fertility of oo. 1d. Yellows; Sits Oct., 1921 R2 Color of body, wings, etc. I4. Poeh. Se a. es p May, 1920 W 1384 |Eyec ilr 15. Beaded........ be | Nov., 1920 W 1964 "uer viability: 16, Rough... ... r March, 1919| L 9 Texture of eye. : Dec., 1919 l6 TBI V I.e. tr May, 1920 | W 1498 HR i 18. Deformed...... d Nov., 1919 W 360 (Almost every part o of body. 19. Serrate........ t March, 1920 | W 1146 Texture and size of eyes. 20. Rimmed....... ri May, 1919 wi Wings and scutellum; vi- ability. 21: Phe su. pa |Feb., 1920 980 |Bristles; viability 22. Hang; veces 8 June, 1919 L 254 ing vein. 28; Short. ies a sh Feb., 1920 W 1110 Wing veins. 24. Short—2.......| she | May, 1920 W 1440 | Wing veins 25. Short—3....... shs | March, 1920 W 1164 Wing veins. 206. Morul. o is m L411 Texture of eyes; viability. 27. Nioked. -i nk | June, 1919 Wil 28. Rosette........ ro | Nov., 1919 L 438 Almost every part of body. In Tables II-VI parentheses indieate data omitted from the regional summary on account of poor viability of one class or else inability to classify one class. The two eolumns under the respective headings represent complementary classes, the one to the left that includ- No. 644] MUTANT CHARACTERS IN DROSOPHILA 239 ing the normal allelomorph of the gene farthest to the left. E.g., in Table II, experiment 3, under crossovers in region 1 there are 48 normal bristle rough eye flies, and 55 stubby bristle normal eye flies. The plus sign (+) indicates wild-type or normal. TABLE II Two-PoINT CROSSES | | Crossovers in Experi- | Re ment Nature of Cross Non-crossovers | Total Number | | 1 Ned ea syr X + 73 | 48 55 247 x Rep qr pute osb X + 1,031 796 i 34 28 1,886 a obn X + 2 275 | 36 25 621 MU c Lo HUN xn 66 70 | 26 20 182 Bey MEI sc X ri 3 291 | 31 29 655 zi Wise ue e prx 131 7 224 Jue ONE psx + 1 111 | 20 21 252 SE. cil oes rri X + 127 H 1 202 a e. woe ae rxs 179.4470 20 98 398 407. Ga ee ssh X + 495 (560) | 37 (0) 532 Zi. vl. 126 (169) | 14 (0) 140 TABLE III THREE-POINT CROSSES Crossovers in Region Experiment Nature of Non- Total Number Cross crossovers 1 2 1,2 r lcu xs sy sb X bn 704 582 |28 29| 40 38 0 0| 1421 8: 2: 1. obn Xf» 35 6 16 sud 0 142 SG oc. fr Xre 943 238 |40 35] 12 25 |0 O: 593 BR III negli fs X sc 48 10 © u ui 9 146 os aD TRA Deae seri Xr 0. 180 120 .HT 3 30 0| 462 ss MONEE ay pede are sers X + 510 26 45 |4 2| 1130 vo IR o sxd obe. 994 |33 25 | 2b Apn 2 691 B vv. seri Xm 98 28 | 60 26/0 6| 625 NE V cT se s sh; X + 174 (146) (54) 55 | (8 19 3 (0| 251 [2 MEME A NU pr X be (20) 68 (0 | (1) 0 0i IU 7.7. E rrixtr 312 281 2 0 1 56 10 0| 601 TABLE IV Four-POINT CROSSES THE AMERICAN NATURALIST [Vor. LVI 2E on- Crossovers in Region = EE Nature of | cross- $ Ez overs E a 1 2 3 h211;8]2,3 12,1 l.|spyr Xosb |321 325|9 4 |22 12244 2050 Ol 1/7 4/00 1,156 Tore Xfs 40. 5B0H5- 20|.3 13|15 19i 26 5| 3,00 187 Gof? xri 161 15070 88 {68 51| 10 218 102 412 1.0.1 618 l0.lorg Xt 48 -57180 27 115. 23 7. 15/5 13 3.|1 11.2 2-1; 240 12.lors X tr 37 56/40 50) 0 1 6. 710 08 2/0 0 01? 208 13.j0od X rri 1 19) 6 Lee 4 0 2/0 0,0 0/0 O00 48 15./f re sce r 11/54 3,9 4 | 43 260 01 00 000 781 21.|re se» X be r|(89) 268 1 (0)| 4 (0|(3 60 00 OlO 0 00| 279 22.scrs Xd 220 160/16 20] 8. 0| 14 19310 11 11i 0/00] 455 24.|scri Xrs |348 306 41 18518. 12 125 ae 13 210--0|]001] 779 27.scrs X nk |250 306/31 23 42 27| 28 10|l 43 9|1 19/00, 754 30.|sc ri X pa ym 10810 (22,7 (0)| ©) 70 00 (00 0|00]| 182 3l.scrim Xs |14 43|2 315112 ..61 12 13/2 21 010 0/00, 257 TABLE V FIVE-POINT CROSSES Be Crossovers in Region £3 Nature of | Non- E EE Cross cross- i BZ 113] | 4 |L2531,4/532,43,4/1]| © [el 3,4 8 of ris| 64 5827 319 911 2 212 2/1 3/3 2/1 01 211 2/0 0| 239 16 rs 183 3 8 6 17 28000 12 110100 000 480 17...\fre se r Xro|232 222/41 271 121 14/190 75/0 00 0/6 210 01 02 111 0| 866 25...|sc ri Xr s nk 40 10 72011 0,27 6000 2,0 5/0 0/0 3/0 1410 1| 350 TABLE VI SriX-POINT CROSSES = Crossovers in Region LET o9 on- LE Nature of | cross- F Cross overs . TRI a 2 3 4 5 11,2]1,3]11,4/11,5|2,4/2,5]3,4|3,5|4,5| 2, | 2, 5 ll.o są r sXsc ri| 92 60415019410 42 1.9 9332312160111/0120000102/331 14./f re scz rXp 8/180 152 16 18 11, 3 311 822 18/0 00 00 1/2 3/0 0/0 0 0 0/0 2/1 0 0 0 0 0,442 No. 644] MUTANT CHARACTERS IN DROSOPHILA 241 TABLE VII SUMMARY OF ALL AVAILABLE DATA BETWEEN CONSECUTIVE LOCI Cross-| Num- Cross-| Num- Cross-| Num- Region over |ber of| Region over |ber of| Region over |ber of Value | Flies Value | Flies í Value | Flies 8 E 13 Li t. 7 2 711 20.8 240 49 is oa sy-sb...... 4.0 | 1,421 |sq-sc5. . ... 10.6 9891 [rt ll. 2.3 12,242 ge SER EE 41.7 re-sc$...... 0.95 | 2,848 [r-s9........ 11.1 | 3,444 o-sb$...... 3.5 | 3,042 ]re-r....... 6.2 593 |r-ro*....... 35 o-bB.. v. 10.2 FOU I To 23.0 187 [tr-ri$....... 0 601 aera 29.0 | 1,044 |seo—p®...... 1.8 PW ee 11.5 208 Oot. 8 30.4 240 |sce—be..... 1.4 79 REA rn B m. bn 34.7 331 S. T. 6,988 Id-8........ 7.9 | 1,146 USC 44.5 256 |sc-d5. .....| 7.2 691 |ri-pa*......| 5.3 132 sb-bn* ,421 |sc-ri.. Tí 1,908 jri-s........ 7.5 | 1,956 sb- 40.0 | 1,156 [sc-s. ...... 21.9 397 Him... 14.7 625 bn-fi...... 19 142 |p-be* ..... i4 176 |pa-m. ..... 5.3 132 f-r 11.2 .| 3,349 Ip-r........ 5.0 654 |s-sh*.. ..... 7.9 923 TABOOS uu. 11.9 385 |p-s... 16.3 252 |s-m$....... 10.1 257 Te usd 21.2 618 |be-r....... 2.4 455 |s-nk*. ..... 11.5 | 1,100 I4. 1... 25.3 ISP... 0.5 809 LITERATURE CITED Bridges, C 1916. ida as Proof of the Chromosome Theory of Here- dity. Genetics, 1: 1-52, 107—163. Bridges, C. B, and Morgan, T. H. 1919. The Second Chromosome Group of Mutant Characters. Car- negie Inst. Wash. Publ. 278: 123—304 Lancefield, R. C. and Metz, C. W 1921. Proc. Nat. Acad. of 8c., August, 1921. Metz, C. W. 1916. geen hien on the Diptera. III. Additional Types of ge romosome Groups in the Drosophilidæ. AMER. NAT., 50: 1918. The P RAS of Eight Sex-linked Characters in Drosophila virilis, Genetics, 3: 107—134. Morgan, T. H. and Bridges, 1916. Sex-linked Inherifance in Drosophila. Carnegie Inst. Wash. Publ. 237: 1-87, Sturtevant, A. H. 1920. Genetic Studies on Drosophila simulans. I. Introduction, Hy- ‘brids with Dr lano, Genetics 5: 00. 1921a. Genetic Studies on Drosophila inulat. II. Sex-linked p nes. Genetics 6: 43-64, 1921b. The North American Species of the Genus Drosophila. Car- negie Inst. Wash, Publ. 301, 150 pp., 3 pl., 49 text-figs. $ These data were used in the construction of the chromosome ‘‘map.’’ INHERITANCE OF PLUMAGE COLOR IN CROSSES OF BUFF AND COLUMBIAN FOWLS DR. L. C. DUNN: As. a part of a search for material suitable for use in measuring the linkage strength of several sex-linked characters in poultry, some preliminary experiments have been undertaken on the inheritance of the Columbian plu- mage pattern. The results of these experiments have confirmed those of Sturtevant (1912) in establishing the sex-linked nature of one of the genes involved in the pro- duction of this pattern, and have demonstrated the rela- tionship between it and the buff plumage coloration. The inheritance and somatic effects of the chief factor in- volved appear to be clear enough to make it useful in genetic investigations on poultry. A short description of the experimental results is therefore given here, to be followed by a more detailed report when further evidence is at hand. i The Columbian pattern, sometimes known as the Er- mine coloration, is characteristic of several standard varieties of a number of breeds of poultry of which the Light Brahma, the Columbian Plymouth Rock and the Columbian Wyandotte are perhaps the most familiar. Although subject to some variation the pattern consists in general of a pure white surface color in all parts of the plumage except in the hackles, wings, and tail feathers, in which black is present either as a central stripe (hac- kles); as a solid color covering somewhat more than half the feather (primaries) or as a solid color covering the whole feather (tail). In typical Columbian fowls the undercolor or fluff at the base of the body feathers is generally lead or slate, which is sometimes so pronounced as to show through at the surface especially on the back. 1 Contributions in Poultry Genetics, Storrs Agr. Exp. Station. 242 No. 644] INHERITANCE OF PLUMAGE COLOR 243 This pattern is alike in both sexes except for the slightly different appearance caused by structural differences in the hackle and saddle feathers. The down color of newly hatched Columbian chicks is white or a yellowish white like the down characteristic of the chicks of many white varieties of poultry, e.g., White Leghorns. Black or gray markings appear on most Columbian chicks as a spot on the head, or as dor- sal stripes on the head or back and in the developing wing quills. This pattern varies in different individuals from an entire absence of dark pigment to the presence of rather heavy dark dorsal stripes. | The color variety chosen for crossing with Columbian was buff, since this offered a clear contrast in the absence of white and the uniformity of coloration and because the plumage color of both sexes is practically the same. Moreover, buff is known to be recessive to many other plumage colors and patterns and for this reason is less likely to earry other factors which might complicate the results. . The first crosses were made between a Columbian male extracted from the second generation of a cross between Light Brahma and White Leghorn,? and purebred Buff Orpington females. Twenty-three chicks were hatched from these crosses. Of these, twelve were predominantly white in the down, and eleven were buff. Of the whites four were pure white, five had a black streak or spot on the head and saddle, and black pin feathers appearing in the wings, and closely resembled purebred Light Brahma chicks in color; one was smoky white and two were white with a buff spot or streak on the head and neck. Of the buffs eight were clear buff in color, one was a very light buff 2 The cross of Light Brahma by White Leghorn (quoted by permission from unpublished data of Sinnott and Warner) when made reciprocally produced white birds in F,, generally with some ticking with black and occasional brassiness or tinging with buff. The White Leghorns used were apparently pure for the dominant inhibitor of color (I) while the Light Brahmas contained the recessive allelomorph of this gene (i). 244 THE AMERICAN NATURALIST [VoL. LVI and two were buff with blaek down on heads and saddles and with black feathers in the wings. All of the white chicks developed adult plumage resembling the Colum- bian pattern except that the black in the hackles, tail and wings was a dingy gray, occurring as stippling on the white ground rather than as a solid color. The buff chicks which survived developed adult plumage in which the hackles, tail and wing feathers were gray or black, while the feathers over the rest of the body were buff. These resembled mosaics of buff and Columbian in which the Columbian pattern was imposed on a buff ground.’ In tabular form the results of this cross were as follows: TABLE I Columbian Male ^ Buff Female Whi ta Buff Down Colors 12 11 d X T Adult Colors 6 6 8 6 The appearance of two kinds of offspring in equal num- bers from this cross indicated that one parent was prob- ably heterozygous in a factor causing the difference be- tween white and buff. Later work showed this to be the male. When a purebred Light Brahma male was mated to purebred buff females (Orpingtons and Plymouth Rocks) the thirty-seven offspring were without exception white in the down and developed into white Columbians as adults. The dominance of white over buff was prac- tieally complete although one or two buff feathers were noted on one hybrid and a slight buff tinge on another. 3' The resemblanee of these hybrids to deseriptions and illustrations of early buff varieties (Tegetmeier, 1872) is quite striking. At present the only buff variety characterized by a considerable amount of black in hackles, wings and tail is the Buff Brahma, although this is not yet recog- nized by poultrymen as a standard t The coloration characteristic of the Rhode Island Red breed is essenti- ally ermine or Columbian pattern with a red ground substituted for the white of true Columbians. A very useful discussion of the relation- ships between these patterns from the standpoint of a breeder and fancier of long experience is given by Robinson (1921) see esp. pp. 55 and 56. No. 644] INHERITANCE OF PLUMAGE COLOR 245 The offspring of the purebred Light Brahma male by buff females were much whiter than the chicks from the first male. Only one of the thirty-seven showed the dark head spot characteristic of Light Brahma chicks and all were of a clear dead white, lacking even the yellowish tinge characteristic of most white chicks in the down. adults these birds were very similar to the first lot. The amount of black in hackles, tails and wings was about intermediate between the amount present in the Colum- bian parent and the absence of black in pure white birds. The first generation hybrid chicks were crossed in two ways. The F, Columbian females were backcrossed to a purebred Buff Plymouth Rock male; the F, buffs were bred inter-se. The results of these matings are presented in Tables II and III. TABLE II RESULT OF CnossiNG F, COLUMBIAN FEMALES WITH PURE Burr MALE: White Buff Total Down Colors 36 38 74 Columbian Buff 9 3 Adult Colors 14 0 0 21 354 TABLE III RESULT OF CROSSING E. BUFFS INTER-SE Buff and White Buff Total Down Colors 9 74 83 Columbian Buff 3 a. 2 Adult Colors 0 0 1T. J5 354 From the backcross of F, Columbian females with a buff male equal numbers of buff and white chicks re- sulted, a clear monohybrid segregation. Evidently one factor determines the difference between white and buff, and from the F, results it is clear that white is the domi- nant allelomorph. This factor is however sex-linked, since 4The differences between the numbers of chicks and the numbers of adults indicate the number of birds which died before definitive plumage or secondary sex characters were developed. 246 THE AMERICAN NATURALIST [Vor. LVI all sons of the F, Columbian females are white (Colum- bian) while all the daughters are buff. The mating of F, buffs inter-se produced only buff chicks, indicating that buff is recessive and breeds true. The nine chicks recorded as buff and white all had buff heads or wings or both and those which lived developed buff adult plumage. Genetieally they were probably ex- tremely light buffs. As regards only the difference between white and buff, we may conclude that the Columbians contain a dominant sex-linked gene for the inhibition or restriction of. buff from the plumage. The first male was evidently hetero- zygous for this factor; the second male was homozygous for it; the Columbian females contained but one dose of it, and this was located in the single sex chromosome; while all the buffs lacked it entirely. This is evidently the same gene (I) which Sturtevant (1912) found in Co- lumbian Wyandottes, although its effects were somewhat obseured by other factors in his crosses with Brown Leg- horn. The presence of this gene in some White Wyandottes which I have studied strengthens the homology between the gene with which Sturtevant was dealing and the gene whieh is present in the Light Brahmas used in these ex- periments. I have recently crossed two White Wyandotte males with purebred Buff (Orpington) females. The white males were known to be recessive white (cc), ?.e., they lacked the gene (C) for the development of color in the plumage. The results of this ¢ross are shown in the table following. RESULT OF CROSSING WHITE WYANDOTTE MALES wiTH BUFF ORPINGTON FEMALES : White Buff Total Down Colors 13 13 26 Columbian Buff T d 3 Adult Colors 3 2 1 3 T 16 In addition to the types noted above three unclassified No. 644] INHERITANCE OF PLUMAGE COLOR 247 chicks were born from one mating. These were chiefly white in the down with black spots on the erown and neck and black quills in the wings. They resembled very dark Columbian chicks. These developed adult plumage dif- ferent from the other chicks in this cross and could not be classified either as Columbians or buffs. Additional factors which have not been identified were probably contributed by one of the White Wyandotte males and further reference to these birds will therefore be post- poned until more information is obtained. Omitting these, the salient fact concerning this cross is the pro- duction of only two classes of chicks, white (Columbian) and buff in equal numbers. The White Wyandotte males bred, therefore, like F, hybrids between Columbian and buff and were undoubtedly heterozygous in the gene for the restriction of buff. As adults the offspring of this cross were indistinguishable in color from the offspring of the heterozygous Light Brahma male first used in crosses with buffs. The amount of black in the hackles of these birds appeared to be somewhat greater than in the offspring of the Light Brahma cross, but it was not sufficient to serve as a distinguishing mark. White Wy- andottes, therefore, may carry a gene for the restriction of buff which is probably the same as the gene found in Light Brahmas and Columbian Wyandottes.* It is not demonstrable, of course, except in crosses of White Wy- andottes with colored fowls which supply the dominant gene C for the development of pigment. In addition to these three instances of the occurrence of a gene for restriction of buff there are numerous other cases in the literature in which the difference be- tween buff (or red) and white in certain parts of the plumage is apparently due to the same gene or one with similar effects. Davenport (1912) found a sex-linked dif- 5 Professor W. A. Lippincott has called my attention to this statement in Robinson (1921), p. 42: ** The pattern (i.e., Columbian) was also produced by erossing the Rhode Island Red (which has really the same pattern with the blaek—on a red ground—redueed to a minimum) with a White Wyan- dotte.’’ 248 THE AMERICAN NATURALIST [Vor. LVI ference between Dark Brahmas and Brown Leghorns, the gene ‘‘ W’’ inhibiting the appearance of buff or red in the hackles and saddles of Dark Brahmas. Jones (1914) was probably dealing with a similar sex-linked gene in his crosses between Silver and Golden Campines, and Hagedoorn’s (1914) evidence indicates that the same or a similar sex-linked gene differentiates Silver and Golden Assendelvers. Punnett (1919) distinguishes a sex-linked gene ‘‘S,’’ which in crosses of Silver and Golden Campines inhibited the development of buff or gold in the plumage, leaving certain portions of the feather ‘‘ silver" or white. Most recently evidence pre- sented by Haldane (1921) indicates that black and white barring such as characterizes the Barred Plymouth Rock variety is differentiated from black and buff (or red or gold) barring by the same gene ** S" for the inhibition of buff. This sex-linked gene ‘‘S,’’ Haldane found to be linked, as was to be expected, with the sex-linked gene B” (barring). In each of these cases a dominant sex-linked gene was found which restricted or inhibited the development of buff (or red or gold) in certain parts of the plumage. Although in the absence of data on erosses between the varieties mentioned it is impossible to assert that the restrietion of buff in the silver or white-patterned vari- eties is in each ease due to the same gene, the presumptive evidence in favor of such a view is strong. It appears probable that Columbian and buff varieties of several breeds (Leghorns, Plymouth Rocks, Wyandottes, etc.) are differentiated by the presence in the Columbians of a gene for the inhibition or restriction of buff pigment; and in view of the history of the various color varieties that this gene has been introduced into and now differ- entiates Golden from Silver-laced Wyandottes, Golden from Silver Spangled Hamburgs, gold pencilled (or part- ridge) varieties from silver-pencilled ones and other golden varieties from silver varieties which differ only in the distinction between buff and white in the plumage. No. 6444] INHERITANCE OF PLUMAGE COLOR 249 Data on the results of crosses involving these color vari- eties are urgently needed, and the generalization offered above is put forward as a temporary simplifieation in lieu of but as an aid to more extensive research. Tue BLACK COMPONENT OF THE COLUMBIAN PATTERN When the experiments with buff and Columbian fowls were begun it was supposed that at least two alternative characters distinguished these varieties; viz., ground color (white as opposed to buff) and pattern (black in hackles, wing, and tail as opposed to self coloration). The results of these experiments, a reexamination of the parent types and a cursory review of poultry literature indieate the error of this assumption. 1. Experimental.—The first generation of the cross Columbian X Buff consisted of birds intermediate between the parental types in the amount of black pigment pres- ent. If four arbitrary grades (1-2-3-4) in the reduetion of the amount of black in hackles, wing and tail are made between typical Columbian and entire absence of blaek (white or buff self) then the first generation is found to consist of the following grades. Columbian —1 —2 —3 —4 Self 0 8 10 2 0 0 If the buff parents are classified as self then the hy- brids resemble the Columbian parent more closely. But a careful examination of all the buff females used re- vealed the presence in each of them of a small amount of black pigment usually as broken patches or fine stipp- ling in the tail and primary feathers, and occasional traces in the hackles. The buffs, therefore, can not be regarded as selfs and most of those used in these ex- periments were assignable to grade-4. The amount of 6 It is the experience of farmers and poultrymen, as evidenced in poultry literature, that buff fowls with no admixture of black pigment have been rare. Black in wings and tail is being rigidly selected against and is being gradually reduced in modern breeds. 250 THE AMERICAN NATURALIST [Vor. LVI black in the F, fowls is only slightly above the mid point between Columbian and grade-4. An F, generation has not yet been raised from the cross of typical Columbian X Buff,but some data are avail- able on the F, generation from the cross of the hetero- zygous Light Brahma which was first used in crosses with buff. This male was lighter than typical Columbian, about grade-1. His offspring were not graded but had slightly less black than the offspring of the typieal Co- lumbian male, averaging about grade-3. The amount of black was similar in the Columbian and buff progeny. When these F, buffs were bred inter-se the grades of the F, adult fowls were as follows: ' Columbian —l —28 —3 —4 Self 0 7 13 10 rE 0 The variation in amount of black pigment was practi- eally continuous, except that the Columbian parental type was not recovered. No buffs were obtained which were entirely free from black in tails or wings. The F; Columbians were backcrossed with a pure Buff Rock male which showed only faint traces of black stipp- ling (mealiness) in the tail. The progeny of this cross were of the following grades: ) Whites (Columbians) and buffs combined: Columbian mi —2 —3 —4 Self 0 2 3 9 13 4 The amount of black in these fowls was obviously much less than in the F, generation although the same grades were represented. In four fowls (three buffs and one white) no trace of black pigment could be detected. It is obvious from these facts that as regards the black component of the Columbian pattern, the Light Brahmas and the buffs used differ only in amount. A blend oc- curs in the first generation followed by segregation in the second and backcross generation. It is probable, there- No. 644] INHERITANCE OF PLUMAGE COLOR 251 fore, that the two types crossed differ from each other by multiple factors affecting the amount of black pigment produced. The number of these factors can not be esti- mated because of the small numbers of animals involved and because a second generation has been bred only from an original eross in which the Columbian parent did not have the amount of black normal to that variety. Fail- ure to recover the typieal Columbian pattern in later generations is probably due to the last named circum- stance rather than the absence of segregation. The two varieties probably do not differ by any single factor determining the presence or absence of black pig- ment, but only in the degree to which black is produced, the degree probably being governed by accessory or modi- fying factors. This fact attaches especial interest to the appearance of several birds in the backcross generation which show no trace of black pigment. Do these repre- sent loss of a factor determining the ability to develop any black pigment at all or are they segregates in which factors limiting the exercise of the black-producing fune- tion are at a maximum? Since they are few in number and since variation in the amount of black grades im- perceptibly into the self condition I am inclined to the latter view. If this is true it should be possible to reduce the amount of black in Columbian fowls by rigid selection against it to such a point that birds might be produced which were phenotypically white, but which as regards restriction of buff would breed like Columbians.‘ Such a character would be in effect a sex-linked self white and the absence. of a sex-linked white in the many breeds investigated points to the probability that none has been produced in this way. Much of the interest in the case presented here inheres in the apparent simplicity of the results. The crossing of Light Brahmas and Buff Orpingtons or Buff Plymouth Rocks produces in the first, second and backcross gen- erations only two easily distinguishable types, white (Co- 7 One such bird has appeared in the course of these experiments; see p. 244. 252 THE AMERICAN NATURALIST [Vor. LVI lumbian) fowls and buffs—in which to be sure there is some variation in the amount of black pigment present in certain parts of the body, but apparently no epistatic pattern factors are introduced from either side of the cross to obscure the visible segregation of the main fac- tors. Restriction of buff as found in Columbian fowls is, therefore, a valuable sex-linked gene for use in meas- uring linkage or in other Mendelian experiments with poultry while buff appears to be the best color variety to be used in studying the inheritance of unknown plu- mage characters. The Brown Leghorn or game type plu- mage pattern, although it resembles the supposed wild type form, is in the writer's opinion less valuable than buff because of the often evidenced? presence in the ge- netie constitution of the Brown Leghorn of epistatie pat- tern factors for extension of black pigment, stippling, ete. The general results of the experiments reported have been to confirm and extend the previously known facts regarding the inheritance of the Columbian variation. The genetic relationships of this pattern and the buff coloration also throw some interesting light on the evo- lution of these two color varieties. They differ, it has been shown, only in one main gene which determines the production or restriction of buff in the plumage. Both are able to develop black pigment in certain parts of the plumage, while they differ quantitatively in the de- gree to which black may be produced. The former single factor difference probably arose as a single mutation, while the latter and less important difference is one which could be brought about by selection of small variations which had already arisen in a common parental stock. The variation which produced the chief difference be- tween these two color varieties, i.e., the restriction of buff, undoubtedly took place at least 75 years ago and probably in China, although there is no evidence that the same variation has not occurred several times. The first known Columbian breed was probably the Gray Shang- 8 Sturtevant, A. H. (1912); Lefevre, G. (1916). No. 644] INHERITANCE OF PLUMAGE COLOR 253 hae, from which the Light Brahmas were derived. These fowls were imported into the United States from China in the decade before 1850? and into England shortly afterward. At about the same period and often in the same shipments were imported certain buff birds which eventually became the foundation stock of the Buff Co- chin breed from which practically all buff varieties of the present day received their color. These two varieties were practically identical in characters other than plu- mage color’ and in the matter of plumage the chief dif- ference was the difference in body feathers, being white more or less stippled with gray in the Shanghaes and buff similarly stippled with gray (mealiness) in the Buff Cochins. In China, observers have regarded the buff as the older color variety while the gray was noted as sep- arate about 1840.1! "The Chinese apparently paid little attention to color in breeding their fowls and the vari- ation from buff to white (or the reverse) in the plumage may have occurred many years or even centuries pre- vious to this date. The further differences between Columbian and Buff breeds have taken place since their introduction from the Orient, chiefly under the selective breeding of English and American poultrymen. The buffs were at first char- acterized by a great deal of variation in the shade of the principal color—ranging from lemon to red; while the wings, and tails, and tips or margins of the hackles va- ried from solid black through stippling and blotching to an absence of black in any one of these parts? All subsequent selection has been against the black? and the Ameriean Standard of Perfection now specifies ** buff in all parts of the plumage." In the Shanghaes or Light Brahmas on the other hand the object of the breeder 9 Weir, Johnson and Brown, ‘‘The Poultry Book,’’ N. Y., 1912, p. 528. 10 Tegetmeier, W. B., loc. cit., p. 63. 11 Weir, Johnson aid Brown, loc. cit., p. 528. 12 Weir, Johnson and Brown, loc. cit., p. 527; p. 630, p. 540. Tegetmeier, loc. cit., " 18 With the exeeption of the selection for black in hackles, wing and tail whieh was employed in developing the Buff Brahma variety. 254 THE AMERICAN NATURALIST [Vor. LVI has been to preserve the black in the hackles, wings and tails and to heighten the contrast with the white body by selecting against grayness or mealiness in the body feathers. Two principal processes were apparently in- volved in the production of buff and Columbian vari- eties; a discontinuous change or, mutation producing the chief difference, and the accumulation by selection of minor factors producing the minor changes. It is im- possible to say whether the buff and Columbian varieties which exist at the present time in the principal breeds were derived from these original types by crossing or whether the principal mutation and the minor changes and selection have recurred in the different breeds. The probabilities are in favor of the first alternative. SuMMARY 1. The Columbian plumage coloration in domestic fowls. is distinguished from buff coloration by the presence of a gene S which determines the restriction or inhibition of buff pigments from the feathers. This gene is sex- linked, and dominant over its allelomorph s, which per- mits the development of buff pigment. 2. Fowls with the Columbian coloration do not differ from buff fowls in any single gene governing the develop- ment of black pigment. Multiple genes appear to de- termine the difference in the amount of black pigment developed. 3. Columbian and buff fowls are genetically alike in plumage pattern, that is, in the ability to develop black pigment in the feathers of certain areas (hackle, wing and tail feathers). 4. The buff coloration appears to have diverged from the Columbian coloration, or the reverse, by a single gene mutation affecting the development or inhibition of buff pigment; and by the accumulation through artificial se- lection of multiple genes for the development of black pigment in the Columbian varieties of fowls, and by the reverse selection in most buff varieties. No. 644] INHERITANCE OF PLUMAGE COLOR 255 BIBLIOGRAPHY Baur, E. 1914. Einfuhrung in der experimentelle Vererbungslehre— Berlin, ea E. B. 1912. ag Exp. Zool. Vol. 13, pp. 1-18. Hagedoorn, A 1909. utei by Davenport, 1912. Hagedoorn, A 1914 E data quoted by Baur 1914, pp. 202-3, Haldane, J. B. S. 1921. Science, N. S., LIV, p. 663. dy 1914. The Campine Club, 1914, Year Book. Quoted by Morgan. 1921. Utility Powtry Jour. Vol. VI, No. 4 Lefevre, - 1916. Anat. Rec, XI, p. 499. Morgan, T 1919. pee Inst. of Washington, Pub. No. 285. Punnett, R. C. 1919. Mendelism, London, pp. 83-87. Sturtevant, A. H. 1912. Jour. Exp. Zool., Vol. 12, pp. 499—518. Robinson, J. H. 1921. Fundamentals n Poultry mim laee Poultry Journal Pub. ney, Il en pp. (Contains fine ph rh ‘end color plates o fst pe. feather patterns, ete.) Tegetmeier, W. B. 1872. The decani Book, London, p. 57. Weir, Johnson and Bro 1912. The Poultry I Book, New York, pp. 527-528. FURTHER NOTES ON THE PALEONTOLOGY OF ARRESTED EVOLUTION DR. RUDOLF RUEDEMANN State Museum, ÁrBANY, N. Y. Tue writer has in a former paper! endeavored to fol- low up the eauses of persistence as seen from the side of the paleontologist. Using as a basis the genera which appear in Zittel-Eastman's Textbook of Paleontology (1913) and defining as persistent all genera which pass through more than two periods, the following data rela- tive to number of persistent genera (4), total number of genera cited (B), and percentage of persistent genera (C) were obtained: A B C Forsuabuerh.. 20.1. 1.525.705 8L 48 86 56 po TE E c uu do O 149 6 do e deu Ea a aN 46 294 15 aries e POON g il a PISA 5 277 2 CVNKDERIERE. ee a 0 96 0 Pup ett] WEM ee a e 1 23 4 Diumuden -orea re ea eI 0 25 0 AStATONIGR ae es 5 43 11 BebinoHon o I ose s 19 191 10 ON CI CLIE Uc Tus 68 306 22 BracbioDodB.... e ee eva cvs 33 384 9 Mollusca: Peleeypoda... vise rein 78 446 16 isst Bx A CNN REUS V E MEE 5 27 M SUES ee las E ees 126 420 30 Jiu ‘oad i ee ee 5 1 29 Pulsonifa. . 1 3 a ss 7 65 11 Are veu (a) punt 7 b cti 12 170 7 ) Ammonoidea...... 0 455 0 (c) Dibranchiata Dole. 0 CRUS 0 cea TUODA C UR D IUE M TE 131 4.5 a ert ike ea S 18 68 26.5 hive Cae he eee ee IS 4 20 Fei tri Linc Se I PUERILIS. 7 134 4.5 Arachnida... os a Ren 3 66 4.5 DHS OQ A ose eEE 16 168 9.5 (in first edition) 1R. Ruedemann, ‘‘The Paleontology of Arrested Evolution.’’ Presi- dential Address. Albany, 1916. New York State Museum Bull. 196, 1918, pp. 107-138, 256 No. 644] ARRESTED EVOLUTION 204 Dipnoi, Teleostei, Reptilia each have one in the 1896 edition of Zittel-Eastman. The vertebrate volume had not yet appeared of the second edition. From an analysis of the percentages we drew the fol- lowing inferences: l. The lower classes tend in general to have more per- sistent types than the higher. |. 2. Within each order and class, again, the lower sub- classes tend to furnish the greater percentage of per- sistent forms. 3. Frequently the persistent genera form a primitive central stock from which numerous shorter lived genera branch off. . 4. The stable conditions of the open ocean and deep sea (as in the Foraminifera) and the subterranean con- ditions favor persistence of types, the latter condition including the burying and boring forms. 9. Sessile forms eontain more persistent types than the vagile benthos. 6. Persistent types prevail in much greater number among the marine forms than among the land and fresh- water animals. Among the eontinental forms again the limnal and fluviatile forms appear to be more persistent than the terrestrial forms. 7. Most persistent types are small and inconspicuous forms. 8. Many persistent genera show a slow development, a distinct climacteric period and a long post-climacteric period. Connected with this observation is the other that persistent genera which slowly develop never pro- duce many species during a single geologic period. 9. Minor factors of persistence are seen in (a) extreme individual vitality (as in Lingula and Crania), (b) im- mense broods (as in Ostrea and Limulus), (c) extreme restrietion in the matter of food, as in the eaters of carrion and refuse (Capulid:m, oyster, ete.). The same criteria were found to hold, on the whole, in regard to the persistent species and the higher groups 258 THE AMERICAN NATURALIST [Vor. LVI (families and orders). In the latter case superior sets of offensive arms and defensive armors, early developed, appear to have helped to give stability to some, as in the seorpions (pineers and poison glands), limulids (leathery armor combined with burrowing habit and enormous broods). Some, as the turtles, have successfully special- ized for protection. In trying to reduce the multiplicity of factors to a few controlling agents, it was found '' that these are the fixa- tion of the ‘ over-taken’ and post-climaeterie types, the presence of stable physical conditions, and withdrawal in various ways from the fields where the struggle for ex- istence is fiercest. The stable physical conditions have been found by many in the open ocean, by some in the deeper littoral regions of the oceans, by others again in subterranean fields, by some in the rivers and lakes of continental regions that remained undisturbed by fold- ing. Withdrawal from the struggle for existence with other organisms has been accomplished by a variety of means, as by isolation, burrowing life, small, inconspicu- ous size, superior, often deadly, offensive and strong defensive arms, through restriction to poor fare, great power of endurance, ete.” In an analysis of the biologic factors that have per- mitted persistence, two entirely different groups of per- sistent types must be distinguished: (1) The post-cli- maeterie types; (2) the primitive central stocks. The former rely on stable physical conditions and withdrawal from the arena of the struggle for existence, as far as possible; while the latter are frequently dominant in the very seats of war. We have termed the first persistent terminals, the others persistent radicles. The persistent terminals were considered to have be- come so fixed in all their characters as to make them persistent partly by the factors of progressive fixation and partly by the fact that they have in various ways avoided the opposing factor of natural selection; their conservation thus being in fact due in part to their ge- No. 644] ARRESTED EVOLUTION 250 rontie condition and in part to the peacefulness of their surroundings. The persistent radieles, on the other hand, were thought to owe their persistence to the fact that through their primitive nature they are still adapted to a greater vari- ety of conditions and that while there may be consider- able variation, it is around a still unspecialized, primi- tive form and thus diffieult of recognition. Or, expressing the same difference in terms of the four processes of heredity, ontogeny, environment and selection, around which, according to Osborn, the life and evolution of organisms continuously center, we found that * the difference between the two groups of persistent types, the relatively rigid terminals and the more vari- able radicles, consists in the fact that in the former all factors have become fixed and unresponsive to stimuli, only the selection still slowly acting, while the latter are so well adapted to a variety of conditions that no changes readily originate through any of the processes of envi- ronment, ontogeny and selection, which affect the whole stock, while at the same time no changes in the germ plasm are induced through hereditary tendencies."' The following notes are written with the intention partly to add certain new factors that appear to con- tribute to the persistence of forms, and that had not been taken into account in the first essay; and partly to enter deeper into the analysis of the ultimate causes of per- sistence made possible through more recent investiga- tions into the nature of phylogenesis. © 1. ADDITIONAL FACTORS or PERSISTENCE The new factors here mentioned have all to do with the methods of reproduetion whose influence had not been recognized, in the first paper, in the percentage table of persistent genera. (a) Reproduction by Simple Division.—In the Proto- zoa reproduction takes place by division without any 260 THE AMERICAN NATURALIST [Vor. LVI loss, so that there is no distinction between parent and offspring. There is no death and thus it is that Weis- mann and others have spoken of the ** immortality of the Protozoa." It is certainly significant, in this connection, that among the Foraminifera 56 per cent. of the genera were found to be persistent and many were found to exhibit tremendous persistence, ranging from the Ordo- vician, Silurian, Carboniferous and Triassic to recent times, and that even species (see Ruedemann, op. cit. p. 126) are known to extend from Silurian, Devonian, Carboniferous and Triassic times to the present. These forms the writer designated as actual ‘‘ immortal types ”’ in contrast to the theoretically immortal protozoans of Weismann. There occurs, however, among the protozoans besides this asexual mode of reproduction a group of processes that are clearly the primitive beginnings of fertilization. In these forms of conjugation different stages may be dis- tinguished, viz., the mere congregation of cells in groups without visible exchange of plasms (cytotropy); the ex- change of substance taking place only through osmotic processes; further conjugation, where real fusion of plasmas occurs but the cell-nuclei remain separate (plas- togamy); and finally such modes of conjugation, where also nuclear fusion of the conjugating cells takes place (karyogamy); and here again, the pairing cells may be either similar in size (isogamy), or even markedly dis- similar in size (anisogamy). It is, however, to be remembered that the usual re- productive process among protozoans is simple fusion of ordinary vegetative cells and conjugation as a rule occurs at rare intervals in most forms, often only when unfavorable conditions arise, or as Maupas’ experiments indicate, the individuals in the course of numerous suc- cessive asexual generations grow old. (b) Reproduction by Budding.—This mode of asexual reproduction differs from that of division originally in the protozoans merely in the different sizes of the daughter- No. 644] ARRESTED EVOLUTION 261 cells and the mother-cell, but develops into a complex pro- cess in the multicellular forms. Distinct budding occurs already in the protozoans as in Arcella, where a number of small buds are constricted off all round. In sponges it is developed to such a degree that no one can fail to recognize the impossibility of drawing any rigid line be- tween growth and asexual reproduction.? In the celen- terates asexual reproduction runs riot, as Geddes and Thompson state. It is, further, by far the prevailing mode of reproduction among the stock-building bryozo- ans; it also is common among marine worms, as with the famous palolo-worm off the coast of Samoa, and fi- nally it is also frequently found among the tunicates. The primitive character of this mode of reproduction ean not be doubted. It probably in all cases is an in- herited character that persisted from the ancestral pro- tozoans. It has by many zoologists been considered as an acquired character among the tunicates, but Van Name? has lately advanced good reasons for the conclu- sion that it is also a primitive character among the as- cidians inherited from their remotest ancestors and that it is not a faculty that can be acquired secondarily. Budding leads to the formation of colonies or stocks. These as a rule are not favorable to a swimming or va- grant mode of life, hence by far the majority of budding forms are sessile, although there are a considerable num- ber of exceptions in the swimming siphonophores, cteno- phores, floating graptolites, and compound swimming worms and ascidians. Since most of the colonial stocks are sessile, budding has often been considered as having been induced by a sessile mode of life and thus held to be a function that could be acquired. Its absence among the sessile cirripedes seems, however, to support Van 2 Geddes, Patrick, and Thompson, J. Arthur, ‘‘The Evolution of Sex,’’ London and New York, 1914, p. 205. 3 Van Name, Willard G., ‘‘Budding in Compound Aseidians and other Invertebrates, and its bearing on the Question of the Early Ancestry of the Vertebrates,’’ Bull. Amer. Mus. Nat. Hist., Vol, 44, art. 15, 1921, pp. 275—982. 262 THE AMERICAN NATURALIST [Vor. LVI Name's contention that this function ean not be acquired when once lost. The faet that the sessile forms contain more persistent types (corals have 15 per cent., bryozoans 22 per cent.) than the vagile benthos would suggest that budding may be a mode of asexual reproduction favorable to the per- sistenee of types; and that it may be the cause of the large percentage of persistent types among the sessile forms. It must here, however, be considered that also the sessile Cirripedia which lack the function of bud- ding, have furnished 20 per cent. of persistent types; and further that in all the classes here considered bud- ding is associated with sexual reproduction, often, as in many celenterates, in a regular alternation of genera- tions. Moreover, the sexually reproducing brachiopods, gastropods and pelecypods have furnished large percen- tages of persistent types, a large number of which are sessile forms. While thus budding would not seem to be the control- ling factor in the persistence of the sessile forms, it is, nevertheless, true that budding may have a distinctly retarding effect upon the evolution of such forms, prin- cipally by the material decrease of the cases of sexual reproduction. As in the ease of the corals, the number of new stocks that originate from sexual reproduction and finding a new lodging place, start new colonies, is very small when compared with the number of asexually pro- duced individuals on the stocks. There are therefore many more generations of asexually than sexually pro- duced individuals. (c) Reproduction by Hermaphrodites.—Another factor that possibly may have contributed to the persistence of forms is hermaphroditism. Claus has pointed out that hermaphroditism finds most abundant expression in slug- gish and fixed animals. ‘‘ Among sponges, sea-anemones, corals, Polyzoa, bivalves, etc., we find frequent illustra- tion of the association of fixedness and hermaphroditism "' (Geddes and Thompson, op. cit., p. 83). The origin of No. 644] ARRESTED EVOLUTION 263 hermaphroditism is still a matter of dispute (see Geddes and Thompson, pp. 83, 84) for while some, as Simon, attribute it to a plethora of nutrition (as especially in parasites), others are ‘‘ content to interpret it as an adap- tation to ensure fertilization, for the possibilities of pair- ing between separate sexes are certainly lessened if the animals are sluggish, sedentary or parasitic." There is likewise difference of opinion as to whether the stage of hermaphroditism is the lower, and the condition of distinct sexes has been derived from it (Gegenbaur), or whether it is a secondary condition, derived from primitive uni-sexuality as claimed by Pelseneer who con- siders it grafted on the female sex in Mollusca, Crustacea and Pisces (Geddes and Thompson, p. 84). Considering its prevalence. among the lowest classes with sexual reproduction, notably the sponges and corals, and again among the Cirripedia, we believe that herma- phroditism is in the former an inherited primitive char- acter and in the latter an acquired one. At any rate, since it is so frequently and distinctly associated with sessility, as in the just mentioned Cirripedia, and in many pelecypods (oyster) and with sluggishness in other pelecypods and many gastropods, and since it is exactly these same groups which contain numerous persistent types, it seems probable that hermaphroditism is a fur- ther reproductive condition contributory to persistence. (d) Reproduction by Parthenogenesis.—Parthenogene- sis is the mode of propagation in at least one typically persistent genus, viz., Apus; but it has also become a confirmed physiological habit in other archaic types of crustaceans among the branchiopods, as notably in Ar- temia, the brine-shrimp, in Branchipus, and in Limna- dia; further in the equally primitive water-fleas (Daphnia and Moina) and finally, among the ancient ostracods, also in some species of the common Cypris. Of the whole class of Branchiopoda, which through paleontology, and notably through the recent amazing 264 THE AMERICAN NATURALIST [Vor. LVI diseoveries of Waleott* in the Middle Cambrian of British Columbia, are proven to reach back to the oldest fossili- ferous beds (in Protocaris marshi Walcott to the Lower Cambrian), Apus is the most remarkable and most often cited form in paleontologie literature. The writer has in a paper, now in press, shown that true Apus, identical in form of carapace and ‘‘ shell glands has been found in Permian beds of Oklahoma. It was before known from the Triassie Buntsandstein of the Vogesian Mountains. Its more than 70 pairs of gill-bearing feet and other- primitive characters have made it the model of compar- ison for Paleozoic crustaceans, especially the trilobites. The Lower Cambrian Protocaris marshi is so closely allied to Apus that it was termed Apus marshi by Ber- nard. There is hence no doubt of the immense age of this type. Apus is now so parthenogenetical in its reproduction that the males were not discovered until a hundred years after the description of the first and best known species (A. cancriformis Schäffer) ; and ** von Siebold repeatedly investigated every member of a colony of Apus, once over 5,000 in number, without finding a single male. At other times he found one per cent. while in certain un- known conditions (probably when food is scarce and life generally unfavorable) the males may be developed in crowds ’’ (Geddes and Thompson, p.189). Similar condi- tions prevail in the brine-shrimp and the other branchio- pods, cited above, as shown by Lereboullet and Nowikoff. Parthenogenesis is associated with other strange habits in the three branchiopods, Apus cancriformis, Limnadia hermanni, and Branchipus stagnalis, which occur together in Europe. These creatures occur only after very wet seasons in puddles, road-ditches and other small pools, where their eggs have lain for decades in the dry mud, exposed to heat and frost. They develop with amazing 4 Walcott, Charles D., ‘‘ Middle Cambrian Branchiopoda, Malacostraca, Trilobita, and Merostomata,’’ Smithsonian Miscellaneous Collections, Vol. 57, No. 6, 1912. No. 644] ARRESTED EVOLUTION 265 rapidity, 4pus cancriformis reaching in two weeks a full size up to five inches,? produce an enormous number of eggs and die. The origin of parthenogenesis in these forms as well as in the rotifers and certain insects has been fully dis- cussed by Geddes and Thompson, and they are certain that it has originated as a degeneration from the ordi- nary sexual process (ibid, p. 198) and is no direct persistence of a primitive ideal state. "Their theory of parthenogenesis is that the ova that develop partheno- genetically ‘‘ are to be regarded as incompletely differenti- ated female cells, which retain a measure of katabolic (relatively male) products, and thus do not need fertil- ization ’’ (they form only one polar body). ‘‘ Such a successful balance between anabolism and katabolism is indeed the ideal of all organic life. In parasitic fungi, sexual reproduction disappears, and surrounding waste products presumably help the purpose otherwise effected by sexual organs, so peculiarities in the conditions of parthenogenetic ova may explain the retention of the normal balance which makes division possible without the usual stimulus of fertilization. Abundant and at the same time stimulating nutrition (Rolph), early differentiation of the sex-cells (Simon), the general preponderance of reproductive over vegetative constitution (Hensen), their liberation before the anabolic bias has carried them too far, are among these favoring conditions." Parthenogenesis thus appears as a degenerative asex- ual process arising from peculiar conditions, the most important of which appears to be temporary over-nutri- tion. As in the other asexual modes of propagation, in division and budding, the inference suggests itself readily that this suppression of fertilization must induce persistence, for as Geddes and Thompson point out (ibid., p. 193) the establishment of parthenogenesis and the ab- 5See Bruno Weigand, ‘‘Mitteilung über das Auftreten der Limnadia Hermanni Ad. Brgt. bei Strassburg im September 1912,’’ Mitt. der Philo- mat. Gesellsch. in Elsass-Lothringen, Bd. 4, Heft. 5, Jahrgang 1912; 1913, p. 730. 266 THE AMERICAN NATURALIST [Vor. LVI sence of fertilization probably involves some diminution in the frequeney and range of variability and thus the establishment of parthenogenesis will be a handicap to evolution. In the case of Apus, and its other associated branchio- pods as well, it is probable that the successful adaptation to special conditions is a strong contributing factor in . the establishment of persistence, as pointed out by the writer in the former paper.. It is possible that Apus has existed under these conditions from very early times. Summing up the evidence on persistence of types from the habits of reproduction, it seems that simple division, budding, hermaphroditism and parthenogenesis have each contributed to this persistence and in their way acted as factors that arrested evolution, and that thus help to explain the relatively large percentage of persistent types in the protozoans, sponges, corals, molluscs and the just mentioned branchiopods among the crustaceans. While the facts thus seem to indicate that these modes of reproduction, other than the normal process of fer- tilization, were favorable to persistence in fossil types, it is, in the present stage of our knowledge of the mean- ing of fertilization, not so simple to recognize the under- lying cause of their arresting influence on evolution. The simplest explanation would obviously be to see in fertilization the principal cause of variation, as such authors as Treviranus, Brooks, Galton, Weismann and Oscar Hertwig have done. Weismann has insisted that the intermingling of two ‘‘ germ-plasms "' is an impor- tant fountain of congenital variation. It ean be readily seen that, under this view, the retarding effect of fission, budding and parthenogenesis consists in the exclusion, or restrietion to long intervals, of fertilization, thereby redueing variability and the possible action of selection. It is also plausible under this view that mutual fertiliza- tion between hermaphroditie individuals tends toward equalization of characters; and this tendency towards equalization is still more increased by fertilization within No. 644] ARRESTED EVOLUTION 267 the same colonial stock or neighboring colonial stocks of plantations. The most important of the disadvantages resulting from hermaphroditism would then be to reduce the variability which is necessary to progress in the struggle for existence. While, however, the possibility is not denied that fer- . tfilization may be a controlling factor in variation, as stated, e.g., by William E. Kellieott in his ‘‘ Text-book of General Embryology,’’ 1913, p. 216, it is also obvious, according to the same author, that the evidence for this view is still seanty and uncertain and, moreover, there : are two exactly opposed views as to the nature of the relation. While Hertwig maintains that the effect of fertilization is to limit variation within a species, Weis- mann asserts that the effect of syngamy or ‘‘ amphi- mixis "' is to cause or promote variation. Kellieott (op. cit., p. 214) states: There is little direct factual evidence for or against these views, either one of which can be maintained upon theoretieal grounds. In a few cases it is known that the amount of variability is not significantly different among sexually (gametically) or asexually (parthenogenetically) produced individuals of the same species. from the standpoint of more recent studies upon heredity and varia- tion the evidence is chiefly either negative or opposed to the idea that this relation constitutes an important element in the origin or present function of fertilization. The present aspects of this reld- tion between fertilization and variation merge in the larger question of the relations with heredity. While among the higher classes fertilization has be- come a stimulus to reproduction and a means of heredity, evidence from the lower groups tends to show that fer- tilization in its results has undergone evolution like every other organic function. The view is widely accepted today (see Kellicott, p. 209) that among the Protozoa the processes of reproduc- tion and fertilization are not fundamentally related, and the primary significance of fertilization must be sought in some other direction. The observations made on protozoans have led to the 268 THE AMERICAN NATURALIST [Vor. LVI rejuvenation hypothesis, chiefly represented by Biitschli, Maupas and Richard Hertwig. ‘‘ It has been found that protoplasmic activity tends gradually to diminish in in- tensity, and that associated with this diminution are certain morphologieal alterations in the structure and composition of the cell ’’ (Kellieott, p. 209). These modi- fications are known as senescence, the senescent condi- tion of the cell eonsisting frequently in the relatively large proportion of cytoplasm as compared with nuclear substanee. Conjugation is assumed to restore the senes- cent protoplasm to its original condition of vigor, bring- ing about rejuvenation. It follows from this that pro- toplasmie activity is cyclic and that periods of senescence would lead to death unless fertilization should occur. The real evidence for this cyclic character of the life processes of the Protozoans has been furnished by the observations of Maupas and Calkins on Paramecium. But observations of Jennings have shown that in differ- ent forms of Paramecium conjugation and rejuvenation may occur at very different intervals, and Woodruff has been able to prevent cyclic relations by substituting nor- mal conditions for the artificial and more uniform ones of the laboratory. ‘‘ By continually altering the char- acter of the food, and by imitating in other ways the naturally variable conditions of pond life, he has been able to continue a single race of Paramecium for over five years ” (quoting from Kellicott), during which peri- od more than 3,000 generations were formed by simple fission. It follows from these observations that proto- plasmie aetivity among the Ciliata may not be cyclie in character under certain conditions, and that when cyclic periods of depression or senescence do occur, the proto- plasm may be restored to a condition of normal vigor. either by physieal or chemieal stimuli, or by fertilization (Kellieott, p. 212). Fertilization is in these cases a form of reaction that takes place when external conditions become too uniform to bring forth the normal vegetative activities, and that No. 644] ARRESTED EVOLUTION 269 leads to an internal disturbanee, thereby correcting the eonditions of uniformity. Applying these conclusions to our case of the persis- tent types it could be conceived that the reduction of fer- tilization to rare intervals, or its entire suppression, in the numerous persistent types that reproduce by fission, budding, parthenogenesis or hermaphroditism, produces a perpetual senescent condition that while not leading to death as in the rapidly dividing and sensitive Para- mecium, finds its expression in the rigidity of the forms, recognizable in their lack of response to external stimuli and of further evolution, i.e., in their persistence. Or in other words, infrequency or entire lack of rejuvenation through fertilization favors the persistent condition, at least among those persistent terminals that do not live under stable physical conditions. Those living under stable conditions may require fertilization as a necessary rejuvenating process counteracting progressive senes- cence and final extinction through lack of external stimu- li. It would then appear that these lower modes of re- production and very stable external conditions could not very well exist together. However, as pointed out by Kellicott, there has been an evolution both of the process and of the consequences of fertilization, and the various possibilities as to the significance of fertilization are not mutually exclusive. It is therefore possible that the large percentage of per- sistent types among forms with more or less suppressed fertilization finds its explanation in some cases in the resulting lack of variation, in others in the resulting senescent and rigid condition of the race, and in still others it may be sought in the process of heredity, con- nected with fertilization. This last possibility will be dealt with in the following chapter. REDUCTION or Factors To FUNDAMENTAL CAUSES The investigation of the various groups of persistent types has indicated that there are a variety of factors 270 THE AMERICAN NATURALIST [Von LVI involved in their produetion. Many of these were found to be connected with the environment, others, acting through variability, or its lack, with selection, and still others with the processes of heredity and ontogeny. While of these four fundamental processes of evolution, viz., heredity, ontogeny, environment and selection, that of selection may account for the cases of persistence where variability has been reduced to a minimum, possi- bly by the lower modes of propagation mentioned above, and that of environment accounts for persistence in those cases where the environment has become so stable as to lack the aetual stimulus for further development, it is obvious that still more important factors are involved in heredity and ontogeny that make for persistence in organisms, especially as it is seen in the post-climaeterie forms, or persistent terminals. Both the conservative proeess of heredity and the much less rigid one of on- togeny appear to become more or less fixed and inacces- sible to changes in persistent types. None of these four processes gives any clue to the actual mechanics of the factors that induce persistence. In trying to trace the latter to its ultimate causes, it be- comes, therefore, necessary to go beyond these processes, and to appeal to the important conclusions that have been obtained by modern experimentation and observa- tion regarding the methods of inheritance and production of new characters by means of the genes or character- determiners of the heredity-chromatin. Among these conclusions especially suggestive in regard to our problem, are the views advanced by Diirken and Salfeld.5 These authors have, one through an analysis of all recent zoological experiments on evolutionary prob- lems, the other through a corresponding analysis of the evolution among the fossil ammonites, arrived at the view that variability or the appearance of new characters, and of new combinations of characters is produced in differ- 6 Dürken, B., and Salfeld, H., ‘‘Die Phylogenese. Fragestellungen zu ihrer exakten Erforschung,’’ Berlin, Gebr. Borntrüger. 1921. No. 644] ARRESTED EVOLUTION A ent ways, by the genes; and not only through internal faetors, as claimed by the Neo-Darwinian school, but also through external ones as demanded by the Neo-La- marckians. The genes, which are not only actual units, or representatives of definite phenotypic characters, but definitely delimited, material bodies, may not only pro- duce new characters or character-combinations by a cor- relative and a combinative mode of ontogenetic evolution, or by loss of genes, as demonstrated by abundant experi- ments, but undoubtedly there takes place also a new for- mation of genes in evolution. This they hold to come about in successive stages through long enduring ex- ternal influence, which first acts upon the cytoplasm of the cells and especially of the germ-cells. This cyto- plasm in itself has been proven to have certain hereditary possibilities (plasmogenous heredity). Under long per- sistent external influence there form first preliminary stages of genes in the cytoplasm which finally, when a certain ‘‘threshold’’ (Schwelle) of continued strain is passed, become true genes of the heredity-chromatin. When this takes place, mutations appear abruptly (salto- mutations). This view, here altogether too briefly presented, would explain the absence of evolution through salto-mutations ` in cases of persistence under continued stable exterior conditions, and since the cytoplasm is known also to in- fluence directly the heredity-chromatin, also the absence of flucto-mutations or variations under stable conditions through lack of external stimulation. However, in the cases where no new genes are formed by external influences, new characters could still appear through loss of genes or correlative or combinative modes of production of new genes from the old ones within the germ-plasm. This, however, leads to a restrictive cone of divergence (‘‘ Streuung’’) of the characters and through ‘‘ self-differentiation’’ by a combinative mode of gene-production to the excessive characters of many terminal series (e.g., dinosaurians); and to the rigid Zia THE AMERICAN NATURALIST [Vor. LVI persistent terminal types, on the other hand, through the gerontic rigidity of the remaining stock of genes. The principal causes of the persistence of terminal forms would then be the failure of production of new genes arising from the cytoplasm, through external influences, and the senescent rigidity of the remaining genes. The persistent radicles, on the other hand, correspond to the extreme development of what Salfeld terms ‘‘ Kon- servativreihen.’’ There are series in which the salto- mutations appear in very long intervals, while the nu- merous side-branches (which furnish the index-fossils) develop by rapid salto-mutations. These persistent radi- cles are therefore able to undergo new periods of explosive and climacteric development (‘‘ Virenz-perioden " of Wedekind) and are thus still less absolutely persistent than the persistent terminals. In these conservative series, according to Salfeld, flucto-mutation is so prevalent that sharply defined ‘‘ species,’’ or better mutants, can not be separated, as notably in the phyla of Phylloceras and Lytoceras which range, qualitatively unchanged in their characters, through Jurassic and Cretaceous time. They thus represent true persistent radieles. "This fact, combined with the observation of the vitality, relative primitive simplicity and adaptation to a variety of condi- tions of persistent radicles, pointed out by the writer in his former paper, suggests that the complex of genes is able to remain relatively undisturbed through external influences (only flucto-mutations appearing) in one part of these groups which persist as radicles, while those parts which become changed through the addition of genes by way of the cytoplasm turn into the side-branches by salto- mutation. EXPERIMENTAL STUDIES ON THE DURATION OF LIFE III. Tue EFFECT or SUCCESSIVE ETHERIZATIONS ON THE DURATION oF LIFE or DROSOPHILA ! PROFESSOR RAYMOND PEARL AND SYLVIA L. PARKER PurRPosE AND PLAN oF EXPERIMENTS In any experimental work of a genetic character on Drosophila, it is often necessary to anesthetize the flies which are to be used in an experiment for a sufficiently long time so that they may be sexed and sorted into dif- ferent groups for the purpose of making matings, etc. It has been shown by Morgan (33) that this procedure has no effect upon the causation of morphological mu- tations, the inheritance of which he has studied (9). The effect might, however, conceivably be quite different in the case of a physiological character like duration of life. Any one who has undergone a major surgical operation feels that anesthetization is at least immediately a rather profound physiological disturbance. Unfortunately, so far as we are aware, no accurate determinations have ever been made to show whether in man one or more anes- thetizations changes the expectation of life. As a mat- ter of fact, there are presumably no human data on the point available in any such amount as would be necessary for actuarial determinations, because in man anestheti- zation is, generally speaking, only undertaken in connec- tion with surgical operations of greater or less severity, so that if we did have statistics of expectation of life of persons who had been anesthetized, there would always be involved the two factors of anesthetization and oper- 1 Papers from the Department of Biometry and Vital Statistics, School of Hygiene and Public Health, Johns Hopkins University, No. 54. For description of the method of numbering bibliographic citations see the second paper in the series (32). 273 274 THE AMERICAN NATURALIST [Vor. LVI ation. In Drosophila these two factors can be separated. It has seemed important, in an early stage of our ex- perimental work on the duration of life in this form, to make a careful and extensive experimental test of the question of whether anesthetization singly or repeated changed in any way the expectation of life or form of the life curve, so that if this factor does have any sig- nificant influence, either favorable or unfavorable, due allowance may be made for it. It is the purpose of this paper to report the results of such a test. The flies used in this experiment were flies of line 107 (generation 8 since January 14, 1921, line bred from a single brother and sister mating for approximately 30 generations). The characteristics of this line relative to duration of life have already been described (cf. Pearl and Parker (32)). The 4,330 flies used emerged between 10 a.m. April 18, 1921, and 4 p.m. April 22, 1921, from thirty-five mass cultures started in half-pint milk bottles April 7, 1921. The regular procedure in these experi- ments was to collect the flies from all 35 breeding bottles in one empty bottle and then to count the flies through a counting tube into 1-ounce vials, allowing 50 flies to each vial? Ten vials were used for each series, except the control series, which had 18. For two of the series only two hours were allowed between successive empty- ings of the mating bottles, to get flies at an average age of one hour, assuming that they emerged uniformly over the interval. - One series was etherized as soon as counted out, and the other series kept as a special control group to see if the handling when the flies were so young and soft had any effect on the duration of life. For the rest of the series the flies ` were allowed to emerge over a 24- hour interval. - Each day’ s hatch was divided randomly and as equally as possible among the | different series. Sg It will be noted that the totals shown in Table I do not accord exaetly with this statement. The discrepancies are due to the fact that a few flies were lost in ehanging to fresh bottles in the eourse of the life duration de- terminations made aceording to the technique deseribed in (27), and oc- easionally a bottle was broken by accident and all its contained flies lost. No. 644] THE DURATION OF LIFE 275 The counting tube referred to above is a device in- vented in this laboratory which we find extremely useful in a great deal of the experimental work. It was devised and first used in connection with studies of the growth of experimental populations of Drosophila (cf. Pearl (7), and Pearl and Kelly (34)). Its construction is shown in Fig. a STS em li p E A. eT ae 1. Diagram showing Vise arty of Boab at sng: counting tube. The. aperture at a is just large enough to allow one fly to pass through at a time. The essential dimensions are as iem length over all 25 cm., diameter of main tube 2 cm., diameter of funnel mouth 6 c When it is desired to eount a definite number of flies the small aperture a is temporarily plugged with a bit of cotton wool, the plunger P is removed from the tube and flies are shaken into the counting tube by inverting the open bottle containing them over the funnel mouth of the counting tube. Then the plunger is inserted and gently moved forward to concentrate the flies in the lower end of the counting tube. Then the counting tube with enough cotton around it to close up the mouth of the bottle is inserted into the bottle into which it is de- sired to place the counted flies and the plug removed from the aperture a. Then as the flies come out of the tube, one by one, through the aperture a, they are counted as they pass this point, with the aid of a tally register, such as is used by doorkeepers at theaters, ete. The plun- ger is gently moved forward as necessary to keep up an even flow of flies through the mouth of the tube. The ether dose used was constant for all the flies throughout the experiment. The group to be etherized was shaken into a clean half pint milk bottle; 5 c.c. of ether was poured onto a piece of absorbent cotton fas-. tened to the under side of a cork stopper; the bottle with 276 THE AMERICAN NATURALIST [Vor. LVI the flies was stoppered tightly with the cork and left for two minutes. Then the flies were turned out on a tile and sexed and counted (since that operation corresponds in extent of handling to what we need to do in making up matings, éte.), then emptied into a vial with fresh food, where they recovered from the ether in about half an hour. For each successive group of flies a fresh bottle and fresh cotton for the ether were of course used. In all other, here unspecified, particulars the technique used in these ether experiments was uniformly that de- scribed in detail in the first paper of this series (27). Seven series of experiments were conducted, differing in respect of the number of times the flies were etherized, and in their age at the time of etherization. The seven series were as follows: Etherized once when one hour of age. Etherized once when twelve hours of age. Etherized once when thirty-six hours of age. Etherized once when three and a half days of age. Etherized twice when seven, and fourteen days of age, respectively. Etherized three times when seven, fourteen, and twenty-one days of age, respectively. Etherized four times when seven, fourteen, twenty-one and twenty- eight days of age, respectively. $ SSSAeh Data The l- lines for the several series of etherized flies and the controls are given in Table I. These l» distributions are calculated on the basis of 1,000 flies at emergence . from the pupal stage, with the absolute number of flies on which the distribution is based given at the bottom of the column in each case. The l» distributions for all etherized flies and for their eontrols in the ether experiment, and for two tests of the flies in line 101 and its continuation 107, are shown graphically in Fig. 2. The data for the survivorship lines in the two tests of line 107 are to be found in Pearl and Parker (27). No. 644] THE DURATION OF LIFE 277 TABLE I SURVIVAL DISTRIBUTION OF ETHERIZED AND NON-ETHERIZED DROSOPHILA CULTURES | Etherized Series Age | l | | Con- | Con- in | | | | | All | Con- | trols | trols Days A|/B|c|npn|z|£*| Gd | ma] usn Hare | | | | ‘eed Old | Old | | | | eriz )-8 1,000 1,000 1,000, 1,000 1,000 1,000 1,000 1,000 1,000 1,000 1,000 7-12 988} 998| 988| 998 993| 990| 995| 993] 13-18 986 993| 978| 995| 984 995 988 980 987 976 19-24 984) 971 978 984! 971| 971! 964| 975| 974| 981| 970 25-30 970| 953| 966 957 918 910 949| 925 918 928 31-36 942|) 916| 912) 942 4| 910 857 918| 903 | 37-42 902| 880 926 901 879 810 880 843. 43-48 778| 792| 7 793| 763| 698, 771| 735| 753| 724 54 629| 634| 732| 725| 643| 598| 605| 652| 604| 701 550 55-60 453, 492| 625, 581| 465 4 476! 50 | 8 61-66 117| 95} 397| 301| 153| 249| 145| 183| 150| 138| 157 67-72 47} 34| 136| 116 9| 165 71 73-78 2 22, 65} 17| 27| 15| 19 13 0 0 0 | 0 0 0 0 0: Absolute number of flies....] 428] 443 411) 432 445) 413) 420 2,992/ 1,338) 478 860» It is at once evident, from an examination of the fig- ures in Table I, and the diagram, that there was no con- siderable difference in duration of life, or in the form of the life curve, for the etherized flies taken as a class, and the non-etherized groups. It is, however, desirable to examine the results of the experiments in detail in order to see whether there are detectable by biometrie methods any small but still statistically signifieant dif- ferences between the several groups. To this end, Table II has been prepared, giving the usual biometrie constants for the several series. Comparing first the entire etherized group as a whole with those which never had any ether at all in their lives, it is seen that the mean duration of life (expectation of life at emergence from pupa) is 1.82-+.30 days longer in the former (etherized) than in the latter (normal) group. The difference is thus slightly more than 6 times 278 THE AMERICAN NATURALIST [Vor. LVI 1000) M \—_—07 ALL CONTROLS —>\ CN . E. ALL ETHERIZEi ; Joop n c Ren SGE & a s I Cc " sio bcd ck cbr PL |) 4*7. H WM M AR 5 AI 44 5 67 73 79 8 9 DAYS O FLY LIFE The 7, lines for all etherized and ccntrolled flies, plotted from the Pik y^ ones 3 TABLE II BIOMETRIC CONSTANTS FOR DURATION OF LIFE OF ETHERIZED AND NorMAL DROSOPHILA | Num- | Treatment ber Mean Standard | Coefficient of (in days) | Deviation [n Fli (in days) | Variation cp V. ck PRIVAT C E EVA 2,992 |51.60 = .16 13.30 = .12125.77 = .24 All e Pepe ce Late Wd DEL ET CQ 1,338 |49.78 = .25/13.68 + .18/27.47 + .38 Etherized- V ON is zu 428 (50.82 + :38 11.76 æ .27'23.14 Etherized when 12 hours old......... 443 (50.20 = .39 12.25 + .28/24.41 = .59 Etherized when 36 hours old......... 411 |53.36 = 46 13.91 + .33/26.06 = .65 Etherized when 314 days old......... 43 a .36 = .28/22.68 = .55 Etherized when 7 and 14 days old 44 s .50 = .28/24.31 = .58 Etherized when 7, 14, and 21 days old. 413 .72 = .50|15.01 = .35|29.03 = .74 seer ot when 7, 14, 2L and 28 days Pel. $20. CLE TE TERI ER MI 420 |49.26 + .47]14.42 + .34/29.26 + .74 Coe rols — out of mating bottles i ag Oi ae A 478 |51.11 + .42/13.75 = .30|20.90 = .63 Controls Pw out of mating bottles »whmi 12 bows 0D iiss oy Osa 860 |49.04 + .31/13.58 = .22/27.69 + .48 No. 644] THE DURATION OF LIFE 279 its probable error, and must therefore be regarded as statistically significant. Absolutely, however, the differ- ence is small. It is equivalent to only 3.7 per cent. in- crease of the expectation of life of the controls. In va- riability in respect of duration of life there is plainly no significant difference between etherized and control groups. It is not entirely clear that the small difference be- tween the etherized and control groups in mean duration of life can be regarded as due to the influence of the ether. An examination of the last two lines of Table II shows that an entirely similar difference in the means appears between the two control groups, which differ only in respect of the time when they were taken from the mating bottles, and without either having been ether- ized. The difference in these two means amounts to 2.07 + .52 days, a statistically significant and absolutely slightly larger difference than that between etherized and control groups. Again there is no significant differ- ence in variability in the two groups. Altogether we shall be justified in concluding that there is no evidence from these experiments that the occasional etherization of Drosophila to the extent necessary in sex- ing and making matings alters the expectation of life by an amount large enough to introduce any sensible source of error into experiments on the duration of life in this form, except possibly where the most careful and accurate actuarial determinations need to be made. Then it will be well to have this possible source of error in mind and to plan the experiments in such way as to check it. Examining the results for the different etherized series it is seen that the highest mean duration of life appears in the group etherized once at 34 days of age, and next to this stands the group etherized once at 14 days of age. Both of these give relatively high mean values. There also appears a definite, though not particularly marked tendency for the variability in duration of life to be greater in the groups which were etherized several 280 THE AMERICAN NATURALIST [Vor. LVI times. No great importanee is probably to be attached to these differences between the several groups, however, though some of them appear significant statistically. CONCLUSION From the experiments herein described, involving the determination of the total duration of life in 4,330 indi- vidual flies, it may be safely concluded that no sensible error will be introduced into duration of life experiments on Drosophila as a result of completely anesthetizing the flies with ether, at least up to as many as four times in the course of their lives. LITERATURE CITED 32. Pearl, ea and Parker, S. L. Experimental Studies on the Duration of Life II. Hereditary Differences in Duration of Life of Line- bred p een of Drosophila. AMERICAN NATURALIST, Vol. 56, pp. 174-187, 1922. 33. Morgen, T. H. The Failure of Ether to Produee Mutations in Dro- sop AMERICAN NATURALIST, Vol. 48, pp. 70 , 1914. 34. ipis R. and K ae Y . C.. Forecasting the Growth of atis. The Future Population of the World and its Problems. Harper's Mag., Vol. gv pp. 704—713, 1921. SHORTER ARTICLES AND DISCUSSION A TEACHING NOTE ON THE ARRANGEMENT OF THE TUBE-FEET IN ASTERIAS SEVERAL summers ago while direeting the laboratory work on Echinodermata in the Invertebrate Course at Woods Hole, a question was raised by one of the students as to the correctness of the account which had been given of the arrangement of the tube-feet in the common starfish, Asterias forbesi Desor. An examination of the point in question revealed the occurrence of a rather interesting irregularity which is here briefly re- ported on. The source of the conditions here described is entirely unknown, but inasmuch as this starfish is commonly used as material for laboratory study, it occurred to me that the facts themselves might be of interest to teachers of inverte- brate zoology. It will be recalled that in Asterias the tube-feet are arranged in four longitudinal rows, two of which are on each side of the radial eanal (the mid-ventral line of the arm). The tube-feet in these rows of two are arranged in an alternate manner, nearer or farther from the mid-line, thus allowing for the accommoda- tion of more tube-feet in a given linear space. The tube-feet are connected with the radial canal by short transverse canals, which are thus longer or shorter according as they pass to tube- feet in the inner or outer series. This arrangement of the tube- feet can be clearly made out in properly dried specimens from which the remnants of the tube-feet themselves are all removed. Their position is clearly marked in such a preparation by the perforations between each pair of ambulacral ossicles, and the whole topography of the ambulacral groove is well demonstrated. It is usually stated that? ‘‘Each pair of transverse canals con- sists of a short canal on one side and a longer canal on the opposite side of the radial canal. The short and long canals of each side are alternating.’’ This arrangement of the tube-feet is shown in a diagrammatic way in Fig. 1, in which the tube-feet are represented as black ovals situated in the perforations be- 1 Quoted from Petrunkevitch, ‘‘ Morphology of Invertebrate Types," New York, 1916, pages 177 and 178. 281 282 THE AMERICAN NATURALIST [Vor. LVI tween adjacent ambulaeral ossicles. This, the common arrange- ment, may be designated Type I. It will be noted that as one runs along the arm the transverse canals of succeeding pairs are long-short, short-long, and so on. Figs. 1 AND 2 However, it appears from my experience in this laboratory that some teachers give a different description of the arrange- ment of the tube-feet. According to these teachers the length of the transverse canals does not alternate in a single pair, but is the same on both sides of the radial canal. This would lead to an arrangement which is shown diagrammatically in Fig. 2. According to this account as one runs along the arm the trans- verse canals of succeeding pairs are long-long, short-short, and so on. It seemed worth while from a teaching standpoint to de- termine which of these descriptions is the correct one. For con- venience, the arms of the starfish will be named in the conven- tional manner a, b, c, d, e—a being the first arm to the right of the madreporie plate (as seen from the aboral surface), the others being named in a clock-wise direction around the dise. In all, seventeen specimens of Asterias forbesi have been ex- amined, some more completely than others. The first two or three pairs of tube-feet at the very base of the arm are usually rather crowded by the abrupt narrowing of the ambulacral groove, so that it is rather difficult to say exactly to which type No. 644] SHORTER ARTICLES AND DISCUSSION 283 of arrangement they belong. They seem usually to be more like Type II than Type I. The groove widens rapidly, however, and the four characteristic rows are quickly established. In the majority of cases the arrangement is undoubtedly like Type I, and this is obviously the source of the usual text-book description. However, in at least nine specimens of the seventeen examined, one or more arms have the Type II arrangement. This may occur in any arm, but in my specimens is most frequent in arm € (five cases). Sometimes the Type II arrangement is estab- lished from the very beginning of any regularity at the base of the arm; in my specimens there were three (probably four) cases of this kind in arm e and one in arm d. In such eases the Type II arrangement may persist throughout the entire length of the arm. More commonly, however, the Type I arrangement is first established and after persisting for a longer or shorter distance abruptly changes to Type II. The number of Type I pairs in such cases seems usually to be small. The transformation is made by a slight irregularity on one side such that two long or two short transverse canals are adjacent—and thereafter the arrange- ment is again entirely regular. Sometimes the region of change is more irregular, but never strikingly so. In one case, in arm a, the arrangement was first like Type I, soon changed to Type II, and in the distal part of the arm changed back again to Type I. In the seventeen specimens examined the Type II arrangement has been found to occur (in some part of the arm) | as follows: BI ee ay ee Ck Rha CERES RR AE KA 3 eases etl Sa UT AR bb Rael trices ine Mere OE oy 4 3 cases SE ork rs Coe alc di WM Ra ed care PE 2 cases CE ee AXE CERE EE Aedes CERTES l ease Oe Te cw ee wy bd tcr EDU. C M Rd 5 eases The number of arms with Type II arrangement in any one *, individual varies eonsiderably, the results for my specimens being as follows: unn affócbod ui seu aeri ee ee 7 cases $ ating affected) ..cluu dedo cut ia Raw ES 1 ease B arma affected. ii. ecco vu stris duod exeo E 1 ease In two cases, both arm a, Type II was found to occur in re- generating arms, though near the base the Type I arrangement 284 THE AMERICAN NATURALIST [Vor. LVI occurred. Whether the injury to the arm was the source of the change is not apparent. It will be seen, therefore, from the above account that both deseriptions of the tube-feet arrangement are correct, but that the one usually given in text-books (Type I) is by far the more eommon; furthermore, that the one type may change to the other with no apparent struetural reasons for the transforma- tion. : The faets here presented furnish, I believe, a complete explana- tion of the differenee in the laboratory accounts as given by different teachers. Rosert H. BOWEN MARINE BIOLOGICAL LABORATORY, Woops Hore, Mass. THE MICRO-FILTER FOR MINUTE FLAGELLATES ITr is frequently desirable during the study of the minuter protozoa, and espeeially of the small flagellates, to concentrate the organisms. This the writer has been able to do in a very simple and satisfaetory manner by means of the deviee shown in Fig. 1, whieh may be ealled the miero-filter ; a name applied not only beeause of its office, but also because of the minute piece of filter-paper used. The contrivance consists of a standard, either of wood or of metal, which supports a burette tube, a minute circle of filter- paper, and a vessel beneath. The water containing the protozoa to be concentrated is introduced into the burette from above, by means of a funnel, and the pinch cock (O) opened sufficiently to allow the liquid to drop into the small funnel or circle of filter-paper beneath (P). The filter is supported by means of stout copper wire. The flow of water from the burette can be nieely regulated by means of the pineh cock, which, to give the best results, should be of the screw variety. The water drops through a glass tube, drawn out into a fine point (T). It was found eonvenient to have several of these tips of different diam- ers. Considerable experimentation is neeessary before the exaet balanee between the flow of water from the burette and that from the base of the filter-paper funnel ean be seeured. When this balance is reached, the burette is filled and the water allowed No. 644] SHORTER ARTICLES AND DISCUSSION 285 to filter into the vessel on the base of the stand. It is necessary, at approximately fifteen-minute intervals, to thrust into the burette, as far down as the shoulder, or point of taper (just above the rubber tube on which the pinch cock rides), a straight gu e P Fic. 2. Pipette with flattened tip for Scraping filter paper, to rémove filtered organisms. L tan: V Fic. 1. The Micro-filter. Simple gee: id pf the micro-filter, sup- F), burette (B), clamp eie seb putet (C), pinch cock (0). capillary tip (7), filter paper (P), and vessel for catching filtered water beneath. eopper wire rod, holding in its lower end a bit of eotton. This serves to stir up the material which it is desired shall be de- posited upon the filter paper, to prevent it from settling and adhering to the sides of the glass, on the slopes of the taper. 286 THE AMERICAN NATURALIST [Vor. LVI When the entire amount of water has passed through the filter- paper, the latter is removed, spread out, and immersed in a bath of water, in a watch erystal. The water should just cover the filter-paper. The deviee shown in Fig. 2 is now brought into play. This eonsists of a glass pipette, flattened and spread at its tip, and serves admirably for gently scraping and sucking the surface of the filter-paper, as it lies in the wateh erystal. This withdraws into the pipette the organisms which have been filtered out. These ean now be transferred to a glass slip and examined under the microscope, or injected into culture media as inoculations. The writer has found that, with practice, the possibilities of the micro-filter may be extended to aid, in many ways, in the study of the protozoa. Leon A. HAUSMAN CoRNELL UNIVERSITY COMPLETE LINKAGE IN DROSOPHILA MELANO- GASTER: In 1917 a mating appeared in the cultures of the authors, the flies from whieh showed no erossing over in the region seute to forked of the sex chromosome, although the faetors echinus, eut, vermilion and garnet, were between the extreme points. This culture appeared spontaneously; selection played no part in it. The stock from this culture has now passed through not less than 80 generations and numbers over 3,000 matings. During this time no crossing over has appeared . within the known length of the sex chromosome. In experiments including the second chromosome points, black and purple, it has been shown that no crossing over takes place between these points when complete linkage exists for the first chromosome. Likewise the third chromosome points, dicheate and hairless, have shown complete linkage when the points seute to forked in the first chromosome, and the points black to Uren in the second chromosome show the same phenome The reine cause is genetic, behaving as a recessive. Its 1 Papers from the Biological Laboratory, Maine Agricultural Experiment Station, No. 142. No. 644] SHORTER ARTICLES AND DISCUSSION 287 position is in the region of dicheate hairless of the third chromo- some. It may be noted that such recessive factors effecting the mechanism of segregation show what might be ealled de- layed Mendelian results for the F, flies must be tested for their linkage relations before anything ean be said regarding the stock. Complete linkage has been reported in but one other ease. Thus in 1912 Morgan showed that crossing over did not occur in the second chromosome of the male of this same species, melanogaster. This phenomenon has since been extended to inelude the other ehromosomes. If it be eonsidered that eross- ing over as originally diseovered for the female of this species is the normal, then Sturtevant has shown not less than three dominant factors to materially reduce the normal amount of a crossing over in the second and third chromosomes. A further incompletely analyzed case of the same investigator suggests that a third chromosome dominant partly controls an increase in erossing over in the second chromosome. Crossing over variations have been shown by Bridges in his ‘‘deficiency’’ ease, ete. From this it appears that there are three kinds of effects shown by the crossover mechanism. The first case, that of Morgan, shows no crossing over in the male. No genetic factors have as yet been shown to be responsible for this. The second ease, that of Sturtevant, shows genetic dominant fae- tors responsible for reducing crossing over in the female. The third ease, given here, shows recessive genetic causes allowing no crossing over in the female. It further shows these factors capable of acting on chromosomes of which they are not a part. Detlefesen and Roberts using the sex-linked factors, white and miniature, present another kind of evidence. In a selec- tion experiment they show crossing over to decline from the normal amount (about 33 per cent.) to nearly zero per cent., no evidence being presented as to the causative agent, although the suggestion is made that ''erossing over in the various regions of the sex ehromosome (and the other chromosomes ?) is prose eontrolled by multiple ineompletely dominant fae- tors." From what has been indicated above it seemed more probable that recessive factors, perhaps one, are responsible for these linkage variations. Especially is this true of their results in series A and At, for with delayed Mendelian segrega- tion, recessive autosomal factors effecting crossing over in the 288 THE AMERICAN NATURALIST [Vor. LVI sex ehromosome, mass mating in every other generation, and eomplieations resulting from only being able to test the female, it is to be expected that selection will progress slowly at first and come suddenly to the climax of reduced crossing over. Marie S. Gowen, JOHN W. GowEN ORONO, MAINE THE AMERICAN NATURALIST Vor. LVI. July-August, 1922 No. 645 EXPERIMENTS WITH ALCOHOL AND WHITE EDWIN CARLETON MacDOWELL STATION FOR EXPERIMENTAL EvoLuTioN, Corp SPRING HARBOR, Lona Istanp, N. Y. Since the familiar paper by Elderton and Pearson (710) upon the physique and ability of children from al- eoholie parents, much discussion has taken place on the relation of parental alcoholism to the condition of the offspring. A small proportion of this has been based upon experimental work with animals, as that of Stock- ard (712 and 713), Stockard and Papanicolaou (’16 and 18), Nice (712 and ’13), Pearl (717), and Arlitt (719). From such studies there should be no hope of obtaining an immediate analysis of the human problem. In so far as alcoholism in man is sociological, involving factors of family life, environment and education, no study of laboratory animals ean have significance. The way such studies may have a bearing upon the human problem is through the revelation of general biologieal reactions that may in all the animals available for study, be found so invariable that it becomes safe to conclude that they appear in man as well How far the specific findings herein reported for white rats may apply to different animals is a matter for experiment and not conjecture. But even were such a biologieal analysis secured, the other phases of the human problem would not be solved. From the data at hand are there any indications of general biologieal reactions that may have significance for all animals? Stockard and Papanicolaou, with gui- nea pigs, found that aleoholization of parents gave un- 289 290 THE AMERICAN NATURALIST [Vor. LVI favorable results in the offspring; Pearl reported gener- ally favorable results in the offspring of treated fowl; Arlitt reported unfavorable results from mild doses on rats, while Nice, also with mild doses, found his test mice slightly better in growth and fertility but less active, as measured by the revolutions of the revolving cages in which they were placed, than the controls. Earlier, Hod- ges (703) had found the viability of puppies reduced by the treatment of their parents; the treated dogs were less active and more susceptible to distemper; Laitinen (708) reported high rates of death at or soon after birth of guinea pigs and rabbits from treated parents. Accepting these general statements as correct, there appears to be no obvious uniformity in the results ob- tained by different investigators. But this lack of uni- formity may be only apparent; it is possible that not all the results as presented will be confirmed by subsequent investigations since none of the experiments reported have eseaped unfavorable eritieism from some stand- point. Aleoholism has such a multiplicity of aspects that it is a matter of great diffieulty to arrange experiments concerning its effect on the offspring of treated animals that will be beyond criticism. For technique satisfac- tory to a physiologist may involve serious errors in the eyes of a psychologist, while the experiments of both may, to a geneticist, seem to have weak points. Until aleohol studies meet the requirements of all erities no final conclusions ean be reached. In problems involving comparisons between experimental and control individu- als the nature of the controls is no less important than the comparison itself. However true this appears to be for all experimental work, it is surprising to note that the main adverse criticisms of the experimental studies of the influenee of aleohol upon the offspring have been aimed at the controls. In spite of the general lack of uniformity in the results as they stand, at least one criterion appears to show con- sistency. This is the reproductive capacity of the treated No. 645] ALCOHOL AND WHITE RATS 291 individuals. All the experiments appear to indicate an immediate reduction in the number of offspring. The uniformity of this result tends certainly to increase its value as a general result; but even so, as long as the controls are subject to criticism, the apparent consistency may be due to the controls and not to the regularity of the reactions to alcohol. For a single result can not at the same time prove the reliability of the controls and the results of aleohol treatment. It is hoped that the controls employed in the following experiments will be found to approach the ideal of satisfying all require- ments. METHODS In 1914 an investigation was undertaken upon the in- fluence of alcohol on the untreated descendants of white rats with the primary object of studying the behavior, or learning capacity, in different generations. In the summer of 1917 war conditions necessitated repeated re- ductions of the stocks until, by the end of the next year, the material was completely lost. This calamitous ter- mination of the work must be borne in mind, for, in spite of the final nature of this report, the data come from an investigation that was not completed. Material and Breeding.—The rats employed belonged to four strains; three of these strains originated respec- tively from three pairs of rats in the Wistar Standard Stock, the fourth strain had been bred in this laboratory for three generations. All matings were between full brothers and sisters. When 28 days old the litters used to start these experiments were divided into two lots on the basis of equal weight and equal numbers of each sex; one of these lots was used as controls, the other was treated. All matings were between the original treated males and females or their descendants; or between the original control males and females or their- descendants. In each generation the control matings parallel those of the descendants of the treated animals, so that each group of test animals in each generation had its own particular 292 THE AMERICAN NATURALIST [Vor. LVI group of controls. Since inbreeding was the rule, the closest possible relationship for the tests and controls in the successive generations was secured; they came from a single pair of grandparents or great-grandparents, and were thus raised at the same time, and after the same number of generations of inbreeding. = Treatment.—'The treatment of these rats was by means of the inhalation method, now made familiar by the work of Stockard and Pearl. The rats were placed in closed tanks filled with aleohol vapor; these tanks have been de- seribed in detail elsewhere (MacDowell and Vicari, ’21). Beginning at weaning (28 days) the rats to be treated were placed in the tanks for 30 minutes a day for 7 days. After this the duration of the daily treatment was meas- ured by the reactions of the animals; for the next 14 days the rats were left daily in the fumes until they were obviously under their influence; subsequently the rats were left each day until they were completely anesthe- tized. This required from three to four hours for the : older rats. Criteria.—The term treated is used to indicate rats that were placed in the alcohol fumes after birth. The fol- lowing generations are herein reported: (1) the treated rats, (2) the treated offspring, (3) the untreated off- spring, (4) the untreated offspring of (3) (second un- treated generation following one treated generation). For these rats the following types of data are given: the behavior in the maze, as measured by time per trial; behavior in a multiple choice apparatus, measured by the number of correet first choices; fertility, judged by the size of the litters and the number of litters; body weight, as judged by growth curves based on weekly weighings. Mazxr-BEnavion Apparatus and Training.— The maze used in this study was built according to the details given by Watson (714) ; namely, a concentric arrangement of five alleys with door- ways and blind alleys so arranged that the true path from No. 645] ALCOHOL AND WHITE RATS 293 the outside to the center required a rat to turn alternately to the left and the right at successive doorways. A rat's training was started at the age of 56 days, after prelimi- nary feeding in the center of the maze on each of the 7 preceding days. Three successive trials a day were given. After the first and second trials the rat was removed from the center as soon as it had tasted the food (bread and milk) which was always found there; after the third trial, it was allowed to eat for five minutes. This train- ing was given for eight successive days. The observa- tions were so automatic that there was practically no possibility that the results were being influenced by an unconscious bias on the part of the observer. In the ease of the treated rats the aleohol was given each day following the trials in the maze. Results.—The average time per trial for each day of the training of the different groups of rats is represented in Fig. 1. The test rats, whether actually treated, or the descendants of treated rats, are represented by the broken lines, and their respective controls by the. solid lines. The numbers of rats included in the different curves, beginning at the left, are as follows: 55 treated rats and 62 controls; 46 tests and 48 controls; 25 tests and 25 controls; 8 tests and 20 controls. The broken lines tend to lie above the solid lines. The tests tend to give higher time averages than the controls, that is, the tests took longer time to run a trial The inferiority shown by the treated offspring from treated parents (fourth pair of curves), and by the untreated offspring from untreated parents and treated grandparents (third pair of eurves) is of the same order of magnitude as that shown by the treated animals themselves; untreated offspring from treated parents show less inferiority than their own untreated offspring. Considering the signifi- cance of the differences between the tests and controls for each day independently, the following results are found: the differences between the tests and controls are over three times their probable errors on five days in the 294 THE AMERICAN NATURALIST [Vor. LVI first pair of eurves, on no day in the second pair, on four days in the third pair and one day in the fourth pair. All the significant differences favor the controls. How- ever, more important than the significance of individual MAZE TIME PER TRIAL. TREATED UNTREATED FROM UNTREATED FROM TREATED FROM ENTS ED E D PARENTS SEC. PER AND TREATED © TRIAL GRANDPARENTS PER SEC.PER LAL `. TRIAL SEC.PER AL ae Ee Se a Fic. 1. Comparisons of time averages in four groups of rats—those treated, their treated and untreated children, and their untreated grandchildren. (Data for the third set of curves taken from MacDowell and Vicari, '21, p. 233.) Broken lines tests, solid lines controls. differences as measured by the probable errors, is the agreement in the direction of the differences on succes- sive days. The fact that the differences on eight suc- cessive days lie in the same direction probably has more significance than that half of these taken separately may be significant as judged by their probable errors. Con- sidering the signs alone, in all the curves there are three out of the 32 points of comparison showing the test av- erages lower than the controls. One of these cases is on the third day of training of the untreated rats from treated parents, the other two cases are on the second and third days of training of the treated rats from treated parents. If chance alone is working, the probability of No. 645] ALCOHOL AND WHITE RATS 295 eight days giving differences in the same direction is the same as the probability of eight coins coming down all heads; in the long run this will happen once in 256 tosses. The chances of seven out of eight, 1 to 32, of six heads out of eight, 1 to 9. Carrying this comparison further by considering all the generations together, the chances of finding three cases favoring the tests out of thirty-two are in the neighborhood of 1 to 860,000. From all this it appears that the test rats are different, as a group, from the controls. Apparently the only differ- ence between the tests and controls that could explain this result is the aleohol treatment given directly, or in the ancestry of the test rats; this leads to the conclusion that the difference in behavior is due to the alcohol treatment. BEHAVIOR IN THE MULTIPLE CHOICE APPARATUS The difference in the behavior of the tests and controls in the generation of the untreated offspring of treated parents is further shown by the training on the multiple choice apparatus. This is the only generation from which sufficient data were gathered for the analysis of behavior on this apparatus. Apparatus and Training —The apparatus used in this training consisted of a linear series of nine compart- ments, with front and back doors operated at a distance by the observer (see Yerkes, '21, for history and uses of this apparatus). Different sets of front doors were opened for the successive trials and the rat was given its reward of food by raising the back door when it en- tered the ‘‘ correct’? compartment. The ‘‘ correct ”’ compartment was the one at the extreme right or left (according to the problem) of the series with open front doors. In successive trials, therefore, the correct com- partment was never the same one, and the solution of the problem did not depend upon the repetition of a reg- ular kinesthetic habit. The steps in the training were these: at the age of 65 days the preliminary training 296 THE AMERICAN NATURALIST [Vor. LVI was started; on the first two days the doors were all left.open and food was exposed to view in every com- partment; the rats in groups of five or so were left to run at random in the apparatus. On the second two days the front doors were all open as before, but the food was concealed by covers fastened to the back doors, and when a rat entered any compartment the food was revealed by opening the back door; the rats were run singly on these two days and given ten such feedings a day. On the last two days of the preliminary training only the regular series of doors were opened, but the rats were fed on entering any compartment (20 trials). Right-hand Problem.—In the first problem the rat was fed only when it entered the right-hand compartment of any set-up (those open in any trial); after wrong choices the rat was confined in the compartment for half a min- ute, and then, by raising the front door, was permitted to make further choices (10 days, 100 trials); next, the same problem was given with a different series of open doors (2 days, 20 trials). Further training was given in the form of a problem in which the correct door was the open one at the left end of the open series, but the results from this problem are so complicated that they wil not be treated at this time. "The main reason for this complication is the fact that at the end of the time allotted for the mastery of the first problem the test and control rats exhibited different degrees of perfection; some had made considerable progress in learning, while others had made very little advance. Accordingly, when the reverse problem was given, those that had learned the most were handicapped by the habit already ac- quired, while those that had not formed the required habit in the first problem were able to progress more rapidly in learning the second problem. Results.—From a study of the individual reaetion ten- deneies as revealed in the last two days of the prelimi- nary training before the problem was presented, and in the regular training after the presentation of the prob- No. 645] ALCOHOL AND WHITE RATS 297 lem necessitated the use of the trial and error method of finding the correct compartment, it appeared that the test rats continued the same tendencies in the regular training that were initiated in the preliminary training, but the controls, on the other hand, modified their re- action’ tendencies as soon as the regular training was started. This result is brought out by the curves in Fig. MUL IHOLECHOGICE. NUMBER OF CORRECT CHOICES ESTIS CONTROLS FROM TREATED PARENTS J6 - 10 16L de | A 14| 12 w aL 14 NL IOL IOL 8L sL Stk. 6L 4l. 4L ei. 2L AEB | 1 L L L 1 l l 1 L 1 1 Sere | j PRELIM |-2 3-4 56 78 9401H2 3-14 PRELIM |-2 34 56 78 9-10 IH2 314 DAYS IN TRAINING DAYS IN TRAINING Fig. 2. Showing the pargar for rats from treated parents liminary anā subsequent performance in the nerek choice apparatus. Aver- age numbers of correct first serie are shown for each successive set of 20 trials. The rats have been classified into groups ng to t reliminary records e fi the behavior of the tests in the limin ary trials is a fairly good index of their behavior in the regular training, but the behavior of the controls in the preiiminary trials gives very little indication of the later behavior. d 298 THE AMERICAN NATURALIST [Vor. LVI 2. The test rats have been classified into seven groups according to the number of right-end choices in the last twenty trials of their preliminary training. The first points of the lines given for the tests indicate the average number of right-end choices made by the rats in each of the groups in the preliminary training; the following points give the average numbers of correct (right-end) choices made by these same rats in successive sets of 20 trials in the regular training. Since the procedure in the regular training is essentially different from that in the preliminary trials, the lines connecting the first and second points are drawn as arrows. The numbers at the ends of the lines give the numbers of individuals included in each group. The arrangement of the controls follows the same plan. Whereas the curves for the tests show a general parallelism, those for the controls are, with the exception of the group of four rats whose preliminary training gave between 12 and 14 right-end choices, relatively independent of the preliminary records. This matter can be brought out more clearly by a study of the coefficients of correlation between the preliminary record of each rat and the trials in the regular training. When the correlation coefficients between the preliminary records and the first 20 trials in regular training, and between the preliminary and the second twenty trials in regular training, ete., are calculated, the figures in Table I are obtained. In every case the differences between the coefficients of the tests and controls (fourth column in Table I) show that the tests have higher correlations, and in all but the correlation between the preliminary trials and the last set of twenty trials in regular training, the differences are statistically significant. These re- sults indicate that there is a real difference between the tests and controls in the way they react to the necessity of using trial and error methods; this may be due to a difference in responsiveness to changes in the situation. ‘The tests appear to be less responsive to the changed procedure, since they continue the same general behavior No. 645] ALCOHOL AND WHITE RATS 299 as in the preliminary training, whereas the controls mod- ify their behavior as soon as the change is made in the procedure. TABLE I CORRELATION COEFFICIENTS, SHOWING THE DEGREES OF SIMILARITY BETWEEN THE NUMBER OF RIGHT-END CHOICES IN THE LaAsT 20 TRIALS OF THE PRELIMINARY TRAINING AND THE CORRECT CHOICES IN EACH SUCCESSIVE SET OF 20 TRIALS IN THE SUBSE- QUENT TRAINING IN THE MULTIPLE CHOICE A PPARATUS. Correlation Coefficients Trials Correlated Difference D/P.E. Tests | Controls Preliminary by Ist 20 trials.. ..| .688+.039 .139 4.063 +.549 +.074 7.4 by2d “ “ ....| .628+.045) .072+.075 | +.556+.087 6.3 by 3d * “ ....| .692+.048 .339+.067 | +.253 +.082 el by4th * *“ ....| .444+.060) .070+.075 +.374+.096 3.8 by 5th = “ ...| -432.061 .101+.074 | 4.331 27.095 3.4 by 6th * “ ....| .489+.057; .049 2-.075 +.440 2-.094 4.6 by7ih *. “ ....| .842+.061) .212+.072 -F.220 2-.094 2.3 All the coefficients are positive; the plus sign is used before the differ- ences to indicate that the coefficients for the tests are higher than the corresponding ones for the controls. In view of the above, the direct comparison of the av- erages of the tests and controls in regular training would lead to error unless the average performance in the pre- liminary training happened to be the same for both sets. In the long run this would undoubtedly be the case, but, as it happens, the averages for the tests and controls do not agree in the preliminary training. However, it was found that this difference depended upon the rats with strong right- or left-hand tendencies, for if these (those tests and controls with more than 12 or less than 3 right- end choices in the preliminary training) be omitted, the average of all the rest ‘of the rats was the same for the tests and controls. Using the rats whose preliminary rec- ords showed between 3 and 12 inclusive right-end choices, the averages for the curves in Fig. 3 were obtained. Start- ing with the same average tendency to enter the right-end compartment in the preliminary training, the controls 300 THE AMERICAN NATURALIST . .[Vor. LVI increase the number of correct first choices more rapidly than do the tests, and as the difference between the av- . erages increases it becomes statistically significant. N9 OF CORRECT CHOICES 2 7 VESTS d Ln. b CONTROLS HAYA A EN v qu 5-6 7-8 9-6 ine ET Average numbers of correct first choices in the multiple choice ap- paratus, in the preliminary training and in the following pairs of days, when rats are included which made from 3-12 correct first choices in their preliminary training, i.e., eliminating rats with strong tendencies either to choose or avoid the correct door, before regular training began. In this way the preliminary averages of the tests and controls are brought together and it be- comes possible to compare the averages in the regular training. Granting that the controls are adequate, the data on behavior indieate that a modifieation has been brought about by the aleohol; the generation showing the least absolute difference in maze-behavior is shown to be defi- nitely modified when the tests are made on a multiple- choice apparatus. FERTILITY Compared with the difficulty of measuring the behavior tendencies of rats, the measure of fertility is very simple and definite. However, the great amount of time required by the behavior studies prevented the collection of many of the available data on the purely physiological side. As a result of this, instead of the long list of criteria of fertility that have been given by other authors, it is possible to give only two with any degree of accuracy and completeness. "These are: the number of rats in a litter, and the number of litters. A more detailed report on the data leading to the following conclusions may be found elsewhere (MacDowell, '22a). No. 645] ALCOHOL AND WHITE RATS 301 Size of Litters—A general tendency for the litters of the test rats to be smaller than the controls persists in the summaries of all generations. The difference be- tween the size of the litters from the original treated rats and the litters from the controls is equal to 10.5 per cent. of the size of the control litters. The treated offspring of the treated rats produced litters that were 10.3 per cent. smaller than the litters of their controls. It appears, therefore, that the treatment of the parents’ of the litters as well as the grandparents does not intensify the reduc- tion in litter size found when only one generation was treated. The untreated offspring from treated rats gave litters that were 11.2 per cent. smaller than their controls, and the untreated offspring from untreated parents and treated grandparents gave litters that were 13.1 per cent. smaller than the controls (see Fig. 4). These differences in individual generations are based on too few cases to be significant when compared with their probable errors, but when the numbers are increased by taking all the gen- erations together, the probable error is reduced so that the difference attains statistical significance (3.6 times its probable error). Litter size, then, gives a result not unlike that given by the behavior data: the tests are inferior in each generation, with no apparent relation to the proximity of the alcohol or the number of generations of treatment. Number of Litters —Given equal time, the treated pairs produced 0.72 litter per pair while the controls produced 2.07 litters per pair. This is a reduction of 64.8 + 3.3 per cent. in the number of litters, and as it is 19.2 times its probable error, it is significant beyond all question. The test litters were slower in appearing than the con- trols. The treated rats from treated parents also gave fewer litters than their controls, but instead of a greater reduction than in the previous generation this second treated generation produced relatively more litters. The reduction was 35.4 + 6.9 per cent. of the controls. Com- ing to the rats not directly treated, the untreated rats 302 THE AMERICAN NATURALIST [Vor. LVI NUMBERS: OF RAIS PER LIITER AVERAGES IN FOUR GENERATIONS NUM PER LITTER TESTS TESTS TESTS TESTS 139 - 9 pum - 6 -10% t] 94 10 oZ ^ 7 CONTROLS CONTROLS CONTROLS TESTS * CONTROLS 8 TESTS TESTS TESTS 5 4) 3 2 | FROM TREATEO RATS FROM FROM UNTREATED RATS FROM FROM TREATED RATS FROM FROM UNTREATED RATS UNTREATED PARENTS TREATE ARENTS TREATED PARENTS FROM UNTREATED PARENTS Fic. 4. Average litter size for the controls and tests baing the relation- ships to the alcohol treatment indicated. from treated parents gave 33.3 + 8.2 per cent. more lit- ters than their controls, and the untreated rats from un- treated parents and treated grandparents produced 55.6 + 8.4 per cent. more litters than their controls (see Fig. 5). All of these differences are, without doubt, statisti- eally significant. Discussion.—Two generations of treatment made less difference in number of litters than a single generation of treatment, and two untreated generations following No. 645] ALCOHOL AND WHITE RATS 303 WUMEER OF LITTER S CONTROLS 90 ac NUMBERS OF LITTERS IN FOUR GENERATIONS 70 60 50 iets TESTS TESTS TESTS — 35% + 335 +55% TESTS 30 ITESTS CONTROLS 20 TESTS CONTROLS TESTS CONTROLS ro TREATED FROM TREATED FROM UNTREATED FROM UNTREATED FROM pers erem Se a RA ERNE NTREATED PARENTS ANO TD EDS TREATED GRANII/PARENTS UNTREATED PARENT TREATED PARENTS Fra. Relative numbers of litters produced in equal periods by the test and control rats in different generations. Beginning at the left the test litters irs: 44, 9, 19, 1; in each case the con litters have been given im Pus geh actual numbers of poti pore: involved were: 42, 12 the treatment produced more litters than the controls. The number of litters is strongly reduced when the pa- rents themselves are treated, but when the aleohol is more remote, the reduetion vanishes and the untreated descend- ants of the treated rats produce more litters than their controls. To explain the reduction in the number of lit- 304 THE AMERICAN NATURALIST [Vor. LVI ters in the presence of alcohol along purely physiological lines would be a simple matter, but a genetie explanation appears to be required when it comes to the inerease over the eontrols given by untreated descendants of treated animals. No general depression or stimulation will ac- count for the continuation of small litters together with the increase in number of litters in the generations not given alcohol directly. It seems necessary to assume that there are genetic factors influencing the number of litters; aleohol prevents the reproduction of such females as carry faetors working in the direction of lower reproductive ca- paeity, so that the litters come alone from females carry- ing higher litter-producing capacity; the next generation will produce higher numbers of litters than the unselected controls, for the controls still carry all grades of fertility, while the tests lack the genetically lower grades. The treated offspring of treated rats produced fewer litters than their controls, but genetically they were superior, as shown by untreated offspring giving more litters than their controls; they were superior to the first generation, for, instead of a 65 per cent. reduction, they gave only a 39 per cent. reduction in the number of their litters. Whereas the immediate presence of alcohol reduces the number of litters, it acts to increase the number in the next generation; therefore alcohol may produce two re- sults upon a single character in two generations. This could lead to much confusion were it not so easy to un- derstand the first result as the cause of the second. - This selective action of alcohol will account for the re- sults from the number of litters, but will not account for the uniform results given by litter size. If this is a - correct statement of the situation, it indicates that the number of litters is influenced by genetic factors that are not identical with those influencing litter size. Although such a distinction between genetic bases for the numbers of litters and litter size has apparently not been made, it is not difficult to conceive, for litter size is largely dependent upon the number and constitution of the germ No.645] . ALCOHOL AND WHITE RATS 305 cells liberated, while the somatie condition of the mother plays a part in determining whether or not a litter will be produced. The results from litter size agree strikingly, qualitatively and even quantitatively with those of Stock- ard and Papanicolaou from similar studies with guinea pigs; the results from the number of litters agree with Pearl’s on fowl in so far as they may be interpreted by assuming a selective action of the aleohol working upon existing genetie differences. In the fowl the aleohol ap- pears to select between germ cells; in the rats it appears to select between mothers of different physiological and genetic grades. WEIGHT The data on weight (see MacDowell, ’22b) form an ex- tensive series consisting of weekly weighings of practical- ly all the rats raised in the various generations herein de- seribed. Individual growth curves were plotted and from these the weights at six ages were taken for statistical study. This procedure was necessitated by the fact that all the rats were weighed on the same day each week, so that the rats were of different ages. The results are based primarily upon the males (see Table II), since the pregnancies of the females make their data less reliable. However, when the data from the females with arbitrary smoothing of the pregnancy peaks are summarized, the results so obtained support those given by the males. Each of the four strains shows that the treated rats grew more slowly than the controls. This is an influence shown . by the population as a whole, although there are some individual treated males that remained as heavy as the heaviest controls. The untreated offspring of the treated rats tended to grow more rapidly than their controls. This result is not so clear as the opposite result in the preceding generation; the absolute differences are not so large and the strains do not show this in equal measure. Treated rats from treated parents barely differ at all from their eontrols. Very little can be concluded from the weights of the untreated offspring from untreated THE AMERICAN NATURALIST [Vor. LVI 306 40110 o[qeqoid sj: £q popp oouoiogrp—q/q *4erveq S[ojuoo eu) oqvorpur sudis Sn[q v0 cc'erryvg9$ — | OL | Of 08s | IL | T9986 OST ET O0Cc'9L-60'8I— | OT | 00'802. | IT | 60986 OST eI 99'0rI-Fg&'02— | OL | OSTZT | IL | SO I6T OGI €'0 FOILS HOS — | OT] Ob Sal MIL] F9 ISI 06 -9'6 geg -F86'£1I— | OT | 06°26 TL {SE ET 09 6'0 809 CRAFT + FZ 22°08 OT | 06'S4 OF 1 2 2 sjuo1redpuuid pojuorj pus sjuored pojuorjun wory pojvorju[] et 104 FSTOI— | 46 | €8'886 | S6 | O0'6vG OST 9'T $29 EPIO | 26 88 6IS | 96 , Cr ORG OST . GS BIG FOLSI-— | Le TUGSI 96 9c'S$6T Oct 9S 7S FLP6 — | 66] 6LLIT | 96 | 9s LET 06 VY 99° FOV IT— | 66} SUITS 9c | 89°96 09 see Leg Fite — 6z Z8g'£9 9% £0'£4 Ob. bots coe iE Ke ros rs *'**** gque1ed pojvor) woy pejvoqu[) 9'T TTA FOSSI+ | ET | 00'893 | OT | OZSTG OST raat Sto FATS + | FPL | OTOES | SI | $8866 OST €T I0'9 Fer6 + | FT| OF L0G | FL | Z0'86I OZI 00 96°F Fero — | 9T | LE99T | FE | 09'991 06 OT IRE FISS F iori. IS Sk | SD] OO sal 09 : L0 ELS Foz ae: 9I IS'IS GT 98'£8 OF Meet a AO OR ree ee Tk ie ee ot es V RR ae S 6 ee ee gjuo1ud poyvom wo poqywory, gL ILS Forset i GE | FEPLE | T£ | SS'SEG O8T 8'8 ISP —S96v-- | 68 | 4686 | 8E | GESTS OST 66 A0 v -FI9Zg | OF | 12'683 | OF | OT CET [1/41 €'8 9re -007c-c | SS | $£9'02I | I9 | $9'6vI 06 6'9 ors F6o'zI+ | 99 | GE'GZE | TA | 99'0TTI 09 ST OFT FCSS +- }9 G6'64 04 e£Y'04 OF €" 9 4 4 4 Saw $4 » eo RE ORUM OR VUA Wo Oe: 88 8 a 99 9 *X wa Ne E E a squared peurou uoa} poqwery, . SOS'UI9AY SOd'€IOAV 'SON| Sure1r) /'SON| SUID | seq Tda 99uoJogr E quourvo1], TOYOTY 04 uorvpoqr asy S[o1ju02) 8389.T, ‘SIOULNOD SXZALLOSdSS?[ WANG HLIM GGWHYdWO,) SLVi ISAL MIVW JO SAJAL SQOTHVA AHL 4O SHADY WAISSHOOQnS LV SLHOTZAA THL II W'Id VL No. 645] ALCOHOL AND WHITE RATS 307 parents and treated grandparents. Two of the three strains represented in this generation show heavier aver- ages for the tests and the third shows heavier averages for the controls; when all the strains together are con- sidered (as in Table II), the test averages are higher at all ages. This shows a marked similarity to the results from the number of litters; just as the offspring of the treated rats appear to be genetically superior to the controls in the matter of litter production, so they are found to be superior in the matter of weight, with the result that when they themselves are treated, the immediate reducing effect of the aleohol makes them about equal somatieally to their controls, instead of growing markedly slower as did their parents. This likeness in results leads to a similar interpretation for the weight as for the number of litters: the aleohol has aeted as a selective agent, eliminating germinal material that included factors for slower growth. ; Discusston In view of the premature termination of these experi- ments no discussion or interpretation can be justified other than by its possible influence upon future work. The data on behavior and litter size taken alone may, if the controls are accepted as adequate, be considered to lead to the general interpretation of a direct and defi- nite modification of the germinal material brought about by the aleohol treatment. On the other hand, the data on the number of litters and weight, when taken alone, agree in inviting the interpretation that the alcohol has acted as a selective agent upon germinal differences that were present in the germinal material of the original animals. One tendeney pulls the race down, the other, by sacrificing the fullest reproductive expression of the treated individuals, tends to pull it up. The specific con- ditions found then are end-results that depend upon the interaction of different influences and do not measure di- rectly the amount of influence exerted by the chemical. 308 THE AMERICAN NATURALIST [Vor. LVI Obviously, the situation is complicated, and equally obvi- ous is the impossibility of proving the individual effects of two or more influences acting simultaneously. How- ever, in this case the evidence favoring one supposition (that of selective elimination of germinal material) is very mueh more convincing than that favoring the sup- position of germinal modifieation. So great, indeed, is this difference that the evidence of direct modification could easily be brushed aside and selective elimination be effectively championed as the effect of the alcohol, although even this involves two opposite results depend- ing upon the proximity of the alcohol. But if a true statement of the situation is desired, the conflicting evi- dence must not be brushed aside. If the germinal variability existing in the race is greater than the variability caused by the direct action of the alcohol upon the germinal material, the results actually obtained would be expected; that is, the effects of selec- tive elimination would appear more striking in the end results. Since the reductions in litter size and in beha- vior stand in spite of an apparently much stronger racial improvement, these reductions give stronger support to the supposition that germinal modification is a second activity of the alcohol than is indicated by their magni- tude. The fact that so many different conclusions have been reached by different investigators from experiments with alcohol would in itself suggest very strongly that the ac- tion of this chemical upon animals is not simple and di- rect like the action of an acid upon a base, yet the general attitude toward the problem seems to have been that there should be a single answer, in one direction or the other, and that as soon as an investigator devises the perfect method, this answer will be disclosed. As long as such an attitude persists the alcohol problem will flounder about in the morass of futile and inconclusive papers. The moment chemistry, and later, experimental breeding, turned away from end results to the phenomena No. 645] ALCOHOL AND WHITE RATS ^ 309 behind them (elements or factors), new epochs were started in these sciences. The problem should not be to judge how bad are the results of aleohol, but rather to find through what channels aleohol may work. The final results will differ in different eases according to differ- ently combined influences of various sorts, just as the same combination of chemieals will yield different results under different conditions, and the same combination of genetie factors will yield various somatie expressions; to know the modus operandi of aleohol is fundamen CONCLUSIONS 1. Beginning at the time of weaning, alcohol was ad- ministered to white rats every day, in sufficient quan- tities to cause complete anesthetization. This treatment appears to account for the following differences between the treated rats and their normal sibs: The treated rats—(a) took more time running the maze. (b) produced smaller litters. (c) produced fewer litters. (d) grew more slowly. 2. The treated offspring from the treated rats differed from their controls in the following ways: The treated offspring—(a) tended to take more time in running the maze, (b) produced smaller litters. .(e) produced somewhat fewer litters. (d) grew at a very slightly lower rate. 3. The untreated offspring from the treated rats dif- fered from their controls in the following ways: The untreated offspring—(a) took a very little longer in running the maze. (b) produced smaller litters. (c) produced more litters, (d) were heavier. 4. The untreated offspring in the second generation from alcohol treatment differed from their controls in the following ways: 310 THE AMERICAN NATURALIST [Vor. LVI The seeond genération of untreated offspring— (a) took more time in running the maze. (b) — smaller ey ced ers (c) prod (d) were kot baie 5. From these results it is concluded that the action of alcohol is complicated; that it works in two or more dif- ferent ways. The data on behavior and litter size suggest that the alcohol may modify germinal material directly. The data on the number of litters and growth indicate that the direct effect of aleohol upon these characters is in one direction and that its indirect effect is in the op- posite direction; this may be interpreted by the assump- tion of a selective róle played by the aleohol. It is urged that the aleohol problem can be settled biologically only when, instead of generalizing from the quality of specific end results, we deal with the channels through which al- cohol may work. COLD SPRING HARBOR, February, 1922. LITERATURE CITED Arlitt, A. H. 1919. T ara of Alcohol upon the Intelligent Behavior of the Fior and its Progeny. Psychol Monog., Vol. 26, No. bese eton Elderton, E. x Reg Pear 1910. g et Study on the Influence of Parental Alcoholism. on Physique and vig: of the Offspring. Eugenics Lab. m No. 10. London Hodge, C. F. 1903. T Influenee of Aleohol on Growth and Development. In ** Physiological nm of the Liquor Problem." New York, pp. 359-3 Laitinen, T. 1908. Ueber die Einwirkung der kleinsten Alkoholmengen auf die weiderstandfahigkeit des tierisehen Organismus mit besond- erer Berueksiehtung auf die Nani miea] Zeitschr. . Hygiene, Vol. 34, pp. 139—252. MaeDowell, E. C. 1922a. as escis of Aleohol upon the Fertility of White Rats. Yo 1922b. Pie ind the Growth of White Rats. Genetics, Vol. 7. No. 645] ALCOHOL AND WHITE RATS 311 MacDowell, E. C. and Vicari, E. M. 1921. SEEN and the Behavior of White Rats. I. The Influ- nce of Alcoholic Grandparents upon Maze-behavior. Journ. Bap. Zool., Vol. 33, pp. 209-291, Nice, L. B. 1912. Comparative Studies on the Effeets of Aleohol, Nieotine, To- acco Smoke and Caffeine on White Mice. I. Effects on Reproduction and Growth. Journ. Exp. Zool., Vol 12, pp. 133-152. 1913. Studies on the Effects of Alcohol, Nicotine and Caffeine on hite Mice. II. Effects on PETET Journ. Exp. Zool., Vol. 14, pp. 123-151 Pearl, R. 1917. The Experimental Modification of Germ Cells. Parts I, II and III. Journ. Exp. Zool., Vol. 22, pp. 125-186, and pp. 241- 310. Stockard, C. R 1912. An B NUN Study of Racial Degeneration in Mammals Treated with Alcohol. Arch. Internal Med., Vol. 10, pp. 369- 398. 1913. The Effect on the Offspring of Intoxieating the Male Parent and the Transmission of the Defeets to Subsequent Genera- AMER, xui dg 47, Stockard, C. E and Papanieo N, 1916. Peas pee id the Hereditary Transmission of De- gen gery and Deformities by the Descendants of Alcohol- a mals, Amer. Nar., Vol 50, Part I, pp. 65-88, AIL pp. 144-177. 1918. Arat prame on the Modification of the Germ Cells in Mam- mals: e Effect of Alcohol on Treated Guinea-pigs and their "isi ns Journ. Exp. Zool., Vol. 26, pp. 119-226. Watson, J, B. 1914. A Circular Maze with Camera Lucida Attachment. Journ. Animal Behav., Vol. 4, pp. 56-59 Yerkes, R. M, 1921. A New Method of Studying the Ideational Behavior of Men- tally Defective and Deranged as Compared with Norma] In- dividuals. Journ. Comp. Psychology, Vol. I, pp. 369-394. EXPERIMENTAL STUDIES ON THE DURATION OF LIFE. IV. DATA ON THE INFLUENCE OF DENSITY OF POPULATION ON DURATION OF LIFE IN DROSOPHILA !: PROFESSOR RAYMOND PEARL AND SYLVIA L. PARKER I Famy early in our experimental work on duration of life in Drosophila it became apparent to us that the num- ber of flies per bottle, or, since the bottles used are of uniform size, the density of population, had some influ- ence on the mean duration of life of the flies, when other environmental conditions are constant. Such a relation- ship might reasonably be expected a priori, from what is known of the influence of this factor on human death rates, commonly expressed as Farr’s Law (cf. Farr, W. (35), Brownlee, J. (36, 37)), and on other biological func- tions, such as growth (Semper, K. (38), Bilski, F. (39)), resistance to poisons (Drzwina and Bohn (40)), rate of reproduction (Pearl and Surface (41), Pearl and Parker (42)), ete. As soon as it was recognized that this vari- able, density of population, might influence our experi- mental results with Drosophila, care was taken in setting up experiments to make this a constant in each case. At the same time the records of the earlier work were care- fully re-examined to determine what part this variable may have played in the results. Happily it was found that in none of our work so far published upon the dura- tion of life in Drosophila had density of population varied enough to have any appreciable effect upon the results or conclusions. As was recently pointed out by Pearl and Parker (42), however, ‘‘there can be no question that this whole matter of influence of density of population, in all senses, upon biological phenomena, deserves a great deal more 1 Papers from the Department of Biometry and Vital Statisties, School of Hygiene and Publie Health, Johns Hopkins University, No. 63. 312 ' No. 645] THE DURATION OF LIFE 313 - investigation than it has had. The indications all are that it is the most important and significant element in the biological, as distinguished from the physical, en- vironment of organisms.’’ In pursuance of this idea we desire to present in this paper our accumulated statisti- cal data on the influence of density of population upon duration of life in Drosophila. This material is to be regarded as preliminary rather than final. For reasons which will appear as we proceed, we are inclined to with- hold final conclusions as to the exaet form of the regres- sion of duration of life upon density until we have com- pleted an extensive ad hoc experimental investigation of the problem. This experimental work is now in progress and we hope to be able to report upon it in full in the course of the next year. In the meantime we have an impressive body of statistieal data gathered from the eontrol groups of other experiments which it seems de- sirable to diseuss now in a preliminary way. II The data of this study are derived from the normal control groups of various experiments on duration of life which we have carried out with Drosophila, accord- ing to the technique described by Pearl and Parker (27). All of the determinations of duration of life recorded in the tables of this paper were made under constant condi- tions of temperature (25° C.), food, ete., as described in the paper referred to. We have divided the material for the purposes of the present study into three groups by stocks (cf. Pearl and Parker (27)), viz.: (a) wild type flies, including our Old Falmouth, New Falmouth, and Eagle Point stocks, (b) Sepia, and (c) Quintuple. Throughout this paper density of population is taken as the initial density (number of flies per bottle) in the small bottles used in testing duration of life. Thus a density of 22 means that 22 flies started in this particular bottle. As time went on the number was diminished by deaths until finally none was left. One of course might use as the variable mean density over the whole life of a 314 THE AMERICAN NATURALIST [Vor. LVI bottle, but a little thought will show that this would be an erroneous procedure when one is dealing with dura- tion of life as the second variable, because mean density bears a direet and implieit functional relation to mean duration of life of the flies in the bottle. We shall be on a clearer footing to take initial density as the variable. Since the cubical content of the bottles is constant throughout, there is no necessity of reckoning density per ee. The number of flies per bottle can be taken as the measure of density, and a good deal of useless com- putation saved. We are indebted to Dr. John Rice Miner for aid in the computations. IIT Table I presents the data for the correlation of dura- tion of life with density of population for the wild type flies. The material is in the usual form of a correlation table. An examination of this surface suggests at once that the regression is probably non-linear. Owing to the manner in which the material was obtained (by compila- tion of the control series of a number of different experi- ments) it results that the different arrays have rather highly different total frequencies. The number of flies per bottle was in no way artificially selected or prede- termined in this material. Instead it was determined solely by the aggregate fertility of the mating bottles furnishing the material for each particular experiment. As has been explained in the first of these Studies (Pearl and Parker (27)), the routine procedure in our experi- ments is to put into one bottle for duration of life test all the flies emerging as imagoes at the same time (i.e., usually on the same day). It therefore would result that if the hatch was particularly good on some day, there might be as many as 90 flies in the duration of life bottle initially. On the other hand, there might be only 2 flies, because only that number emerged on that particular day. Even in spite of the differences in the frequencies of THE DURATION OF LIFE 315 No. 645] 199'6 | 9ST |^ "| 691| ZB |08 | ZZZ 606 | ET | PZE | FES) 99 |86F | E9F | OE | 9ZZ | IGP | 9GP | C68 | CIS O90'T | ZFS | Z8 | EFPL | LTZ |" THIOL I LI * I LI LI s « ae * LL hs. (m Me LI LESEN TI LE ****—g0T G I . e I ove * . st ae a L Í ee eee kn T eee SiR LI LI *—LB sz e epe I zr |t I z |g |e e [jg e qtto v9 Y € 19 Uo ks qe Ik S 19 16 |8 8 8 9 16 T peti: SEI 9 9 Z IL OL S 19. Pe JOE £1 UE UC ||" || A 846 ; 3 uro F Sat HI 6 1G ST (SE TEC Sh i8 TOL O81 Fe BP fe |1|.]0 |^ | — ee IL¥ à g I I E JB X ITI | I8 JF 8L iLi Sh 10] V ie 98 TO? (98 (Se |] | 8 PLL JSI A I IPIS J|SE4JS8 8E. H 6h |99 | or | Se 168 109 (OZ | SS 126 96 710 (Sh 8 | O16 .|6 LT L4 I 9T 9 8g SI 4€ |98 |99 |8E | Ze IAF [GOL I4 ZOT [Z01198 |94 |OI | "^ —-99 OFT |Z SI 42 TT FS SB 150 Feast Pi OL ITA SF 108 | TL EF 1 ZIL | Pel OGL. | GOT | OLET | v8 1.6L | ^ OF 860'T | 6T Ue. WI IS TEE |OB |OP J8 Eo 249 199 |GE {OF |£9 166 108 |ES TƏT (98 | Igi EA | 15 | -gF 6F0'T | 8 446 JPE It 706 |} % |91 (02 | 88 |8. 109 198 112 16 (or |Z 108 | 86 Slt 98 [80110 191 | " E. 096 |f jpeg 19i [ET 109 186 TE 89 SF |9 t£ (TE | Lh 10L fP | LF 184 |06 |94 gS |88 |48 | 6I pui 409 |9 SECO ISE TE: Or A J DB 141 14 GE |0c |06 | So | TS |18 |164 | SF |04 Sr (68 194 01S. — MÀ 669 | Lh Bo [0I 1S oe as: 14 EE | LT | FE JOE (O08 | 19 | LT | he 1 SE |88 | Le m9 08 | Zo |89 | 6I eo N TSY | 06 6 I $5. 10:8 I GE [FE 165 | Fl IGI iE € |86 |06 | 68 |SE (69 SF |vF |99 |96 | — itis © | 62r |S$T ~ Fo {Si al HI I JEI P j|v6 18) 10. j9 iOS (OG. | OF (SE i68 HR 12h OE (Sh | — — 068. =| OT 6 GL | 1061 Ter |^ 06 [85 4.0] | tee 1B 48 (19) ge 66 TF | IF (01 | ^ I T830.L | -£6 | -68 | -98 | -T8 | -44| -E4 | -69 | -99 | -19 | -49 | -£9 | -67 | -F | -TF | -2£ | -£8| -62 | -SZ | -12 | -LT |-£1| -6| -9 | -1 eed s 8 ony onjog ur sə jo 1oquin N SXOOLg CIA ^NORNLVIndOQq AO e V'IIHdOSOW([ AO A&ISNSQ[ TVLLINT (q) ANV “SAVI AO NOLLvuaqp (V) SalavIuVA FHL AOA AOVaUNG NOLLY'IXNHO|) I W'ISViL 316 THE AMERICAN NATURALIST [Vor. LVI the several arrays, it still seems probable from mere in- spection of the general surface that the regression is non-linear. This idea is strengthened by examination of the regression line itself, shown in Fig. | “TITT | 3 LARA PEN IN / " : | V MEAN AGE AT DEATH IN DAYS I | x | | | s- 9 D H H 4 I9 339 AA 45 40 535 57.01 05 OD N T ai 8 89 98 NUMBER OF FLIES IN BOTTLE Mean eto of life of Drosophila for different initial densities of sd Wild s It is seen from this diagram that, neglecting the great dip of the line at density 55 which is consequent upon a very small array with large probable error, the general sweep of the curve indicates an optimum density (great- est mean duration of life) in the general region of 35 to 45 flies per bottle, with a decline on either side of that point, but falling lower on the side of high densities than on that of low. From this table we have the following constants: r — — .0511 + .0068, n= .2443 + .0064. There can be no question that the regression is non- linear. Blakeman’s (43) criterion has the following value: £— 0571 + .0031, It must therefore be concluded that the eee is significantly skew. The correlation between duration of life and density of population in the case of the Sepia stock is shown in Table II. THE DURATION OF LIFE 317 No. 645] (Stes; 18 | AL] eet | — — 09 Ig | 9F | TET | SST | 98 | 9T | T8 | HOS | OFT | C8G | £46 | 296 | 06 "ftt qol Sige amis WON ok Weg es Ga £m ee ola deeqesei t OON OU EU Dro Eu M ee oe ng . [A 1 5 RENEK I z c & AE ; -gL mo led S a. z I Cs ie di» o la T c ou m be vis : calce. bali li vis rih’ a akn kl a E r o P EPIT es |t | Jg f a a [M | ~ A sgi up ge g o I PAT IM p 9 Jra ip 4 u WEE yt || 45r o Z T [T |O8 |% |t jS [et ce | 9 ee |96 | oe | d a WES eda | 0 9 e |* | tt (St 1t |2 [Sr | Se 1 ot 1 Oe Fon 1 Bt ig uc m se |o o1 7] v £e1|* |æ | v6 |8 | et |6 | 0s |6 |o | OF |s | Et ees. | te |2 |g 18: 8I St} 08/6 19 | FI) te aioe |8 jog |g | |i Du £95 |Z | Zt | I6 9 |-jor]o s oe 16 me Aee Ta 18 14 | oo me |t 9g Bae Is Ta wm 0 ee e 10 0g o 5 gor |9 |6 £ 7 lg ls leche (oe rz n pe ae mis e dr WI 14109 1*1 P. | PT T. g | Tri p {oa et wo uc e ps T80L | -18 | -44 | -£4 |-69|-c9 |-19 | -49 | -89 | -6F | -97| -TF | -28 |-e8| -6g |-92| -iz | 41 | -et | -6 | -s |-1 yoq qe ony op3og ur Sot jo oquinyy MOOLg vIdug ‘NOMLVINdOG MO ALISNS([ TIVLLIN[ (Q) GNV "marg A0 NOnLvuaqp (P) swIgVIHVA AHL "O4 gOV4Nng NOLLVTXSHO[) II W'IdViL % 318 THE AMERICAN NATURALIST [Vor. LVI Here again there are a number of small arrays and gaps towards the right-hand side of the table, due as be- fore to the method by which the material was got. The regression of duration of life upon density is shown graphically in Fig. 2. 60 50 $ 40 2 AN E3 x LI ON Was E a wd. ies e M 30 * e i : \ ~ X, M 20 X Š E o 5$. 9 1 IT. 0 M 29 Jd 37 4| 458 4D 313 537 WW n 6D "3 7 à 85 UMBER OF FLIES BOTTLE Fic. 2. Mean pekea ls life of Drosophila for different initial densities of population. Sepia s It is apparent from T here as before that the regression is not clearly linear, but rather indicates an optimum density in the region of 35 to 45 flies per bottle, with a diminished expectation of life at both lower and. higher densities. The constants are — — 132 +.014, »- "283 -.013, f= .0629 + .0066. The criterion of linearity is nearly 10 times its prob- able error, and we may therefore conclude for the Sepia stock, as for the wild stocks, that statistically the regres- sion of duration of life upon density of population is significantly skew. The data for the short-lived Quintuple stock are given in Table III. Owing to the faet that the Quintuple stock is charaeter- ized by low fertility, as well as short duration of life, No. 645] THE DURATION OF LIFE : 319 TABLE III CORRELATION SURFACE FOR THE VARIABLES (a) DURATION OF LIFE, AND (b) INI STOCK NITIAL DENSITY OF POPULATION. QUINTUPLE Number of Flies in Bottle Age at Death | | | | | | | CIS |S |13- 17 21-|25—/29— 33-|37—|41- 45- 49-153-| Total Bless Do) 18) APP I 9I T4118] TEE SEU. 4| 108 pe li on 311.33; 14 31| OP 81 9111 Lebe 15| 146 SR d iS 28] 70] 50] 601.38 13:320] ak 3 LE. 17| 263 10 55.5. 22| 38| 47| 281|19|12|10|. D au v i Be s. sess 151 94! 28. 90] 8118119 |/8]. 1. 1. Si: 138 15 xl 4293 19 8 TT AT 41. Ll. 5 SS Lsb 89 IQ ie ee 11 14 | 11| 42| A| dl TL. d bad al 57 + vds 8:4 3031] 4165 8 ee ite es Sa Gite Bee i 44 Dp dul E Lo chbdgs|] 7| 8| B SL SU e ew. 35 ae nucon Oe Sere 2 SEE SY visos FoR a "NUM gs nsa 6L. Bl 1 B IL baL. he45 | 1 15 AL d 219. —.] 3 (EU NEUES T E 8 EL oui. ve Yl «i A3 dododd4 located bled das oped | 1 9 46. acta IONS WIALGLGI3] dh dq Lob 11 10 48526. x 8i 3i XI Ne IU Borloo 8] d oo oo s A bees [o uso pu ue Drei vii Va. dp e pompe e Aet vous 173415 SESS ALENS SISSE SALE LAS RES 3 Total...... 139 | 261 | 230 202 | 74 |70 | 79 |29]. ..... Lob mb] iam i " | i] this table is less extensive in either direction than the others. 60 50 Fig. 3. det" Quan MEAN AGE AT DEATH IN DAYS 8 (| 5 9 i3 M 2 £5 £9 33 37 Al 45 49 53 57 NUMBER OF FLIES IN BOTTLE SS a of life of Drosophila for different initial densities of stock. 320 THE AMERICAN NATURALIST [Vor. LVI The observed regression line is shown in Fig. 3. Here the regression appears at once to be substanti- ally linear, and is proved to be by the analytieal con- stants, which are as follows: r= — 057 + .020, y2 120 + .020, Die IE 004. The criterion ¢ is less than 3 times its probable error and eannot be regarded as significant. IV Putting all the data together, we have here indispu- table evidence that the density of population is a signifi- eant factor in influencing the duration of life (or death- rate) in Drosophila. The correlation ratio » is certainly significant in the case of all three stocks. Its lower value in the case of the Quintuple stock is almost certainly due to the fact that in the Quintuple experience there is not a sufficiently extensive representation of densities. If the other two tables were to be cut off at the density array where the Quintuple is, they also would show a much lower association between the two variables. So, then, the general portion of Farr’s Law which affirms that death-rate is some function of density of popula- tion receives experimental confirmation in a widely dif- ferent form of life. When one comes, however, to the precise form dis- covered by Farr (35) and confirmed by Brownlee (36, 37), the case is not so clear. We do not care to enter upon any detailed discussion of the point now, because we do not care to draw any conclusions as to the true form of the skew regressions observed till we have some additional experimental results in hand. Provisionally, however, it may be said that the indications are that in Drosophila something like the following relations hold: (a) the lowest density is not the optimum; (b) the mean duration of life tends to increase with increasing density up to a certain point which is optimum; (c) after the No. 645] THE DURATION OF LIFE 321 optimum region has been reached, increasing density is associated with diminished duration of life, which pres- ently falls below the lowest figure found with densities below the optimum. These conclusions must for the present be held as tentative. y In this paper data as to total duration of imaginal life of 13,117 individuals of Drosophila are presented in re- lation to the density of population. It is definitely shown in the case of Wild, Sepia and Quintuple stocks that there is a significant correlation between these variables. The regression of duration of life upon density appears to be significantly skew in the case of Wild and Sepia stocks. The precise form of the regression and theoretical ques- tions connected therewith are left for discussion in a later paper upon the basis of more extensive material. LITERATURE CITED (The plan of numbering citations followed is explained in the second of these Studies.) 35. Farr, W. Vital Statistics: A Memorial Volume of Selections from the Repor rts and Writings of Wiliam Farr, M.D., D.C. po Vul. F.R.S. Edit. by Noel A. Humphreys. London, 1885, xxiv + 5 36. Brownlee, J. Saag on the Biology of a Life-Table. na pei Stat. Soc., Vol pp. 34-65, 1919. Discussion, pp. 66-77. 37. Id. Density d Death- Rate: Farr’s Law. Ibid., Vol 83, pp. 280- 38. sempe, rs “The Natural Conditions of Existence as they Affeet Crimi- e. Fourth Edit., London, 1890 39. eria F. Über den Einfluss des Labani nsraumes auf das Wachstum der amag rw s Arch., Bd. 188, pp. 254—272, 1921 40. Drzwina, and Bohn, G. Action nocive de lean sur les Sinters, en Price PX la ENA de liquide. C. E. Soc. Biol. T. 84, pp. 917- 41. ~ 'R. ana Surface, F. M. A Biometrieal Study of Egg Production n the Domestic Fowl. I. Variation in Annual Egg Production. T. S. Dept. Agr. Bur. Anim, Ind. Bulletin 110, Part I, pp. 1-80, 1909. 42. Pearl, R. and Parker, S. L. On the Influence of Density of Population upon t he Rate of woes cq in Drosophila, Proc. Nat. Acad. Sci, Vol. 8, July, 1 43. Niles. J. On d. for Linearity of Regression in Frequency Distributions. Biometrika, Vol. IV, pp. 332-350, 1905. NOTES ON THE HYBRIDS BETWEEN THE CANARY AND TWO AMERICAN FINCHES O. E. PLATH MASSACHUSETTS INSTITUTE OF TECHNOLOGY, CAMBRIDGE, Mass. Peruars no animal has been so often crossed with other species, and even genera, as the domesticated canary (Serinus canarius). Darwin (1885, I, p. 311) speaks of “nine or ten’’ such crosses, but many more have un- doubtedly been made. The hybrids resulting from these crosses are usually, if not always, infertile, and hence are popularly known as ‘‘mules.’’ In almost all of these crosses the domesticated canary serves as the female and the wild finch as the male, but bird fanciers occasionally succeed in making the reverse cross. The wild species which is most commonly used for this ‘‘mule breeding" is the European goldfinch, Carduelis carduelis Linnzeus.! This fringillid is one of the handsomest finches in ex- istenee, the plumage of the adults of both sexes being made up of a beautiful combination of black, red, white, yellow, and brown patches. The hybrids which result when a yellow, or nearly yellow, eanary is erossed with this fineh are chiefly interesting for two reasons: (1) because they exhibit an apparently. endless chain of variability in eoloration, and (2) beeause their plumage, if dark, is conspicuously streaked, a character which is : lacking (as far as external appearance is concerned) in both the yellow canary and the European goldfinch. Concerning the first of these two points valuable data have been published by Bechstein (1795), Hünefeld (1864), Blakston (1880?), Klatt (1901), Davenport 1 According to Chapman (1916, p. 383), this finch was introduced into the United States at Hoboken, N. J. (in 1878), and Boston, and probably still is a resident near both of these places. . No. 645] HYBRIDS OF THE CANARY 323 (1908), and Galloway (1909). According to these au- thors, the hybrids between the yellow eanary and the European goldfineh may be: (a) completely dark, (b) mottled (spotted), exhibiting an apparently endless variation in color pattern, or (c) entirely white or yellow (very rarely).? The streaking in the dark plumage of canary-Euro- pean goldfineh hybrids has been variously explained as: (a) **derived from the original wild canary’’ (Darwin, 1885, IL, p. 15); (b) as reversion to the Serin finch, Serinus hortulanus Koch (Klatt, 1901, p. 508) ; and (c) as resulting from the latent streaking (visible in the *green'' variety of the domesticated canary) factor of the yellow canary, plus the color faetor of the European goldfinch (Davenport, 1908, p. 20). In 1914 the writer made several attempts to cross the domesticated canary with some of our native American finches, and some of the latter among themselves, since such crosses, if made, seem to have never been recorded. None of these experiments were successful. The work was again taken up in the fall of 1918, and this second attempt yielded several hybrids in 1919 and 1920. For these latter experiments the writer had at his disposal 22 wild finches belonging to the following species: Ar- kansas goldfinch (Astragalinus psaltria hesperophilus Oberholser), willow goldfinch (Astragalinus tristis sali- camans [Grinnell]), California linnet (Carpodacus mexicanus frontalis [Say]), and California purple finch (Carpodacus purpureus californicus Baird). Of these 22 wild finches, 5 were reared from eggs placed under 2 Galloway (1909, p. 4), who has probably reared more eanary-fineh hy- brids than any other breeder, reports the following proportions of self- eolored to variegated (mottled) individuals in the ease of canary-European gold-fineh hybrids: (1) dark plumage (with no white or elear feathers), 172; (2) slightly variegated (a few small white or elear spots in an otherwise dark plumage), 74; (3) variegated (1/4 to 1/2 elear), 75; (4) lightly bouton (1/2 clear to small ticks of dark in an otherwise clear plumage), 9; and (5) completely clear (total absence of dark feathers), 0. 3 A western sub-species of the American goldfinch doctis disini tristis tristis Linneus), popularly known as the **wild canary.’ 324 THE AMERICAN NATURALIST [Vor. LVI canary females and the remaining 17 were trapped shortly before the breeding season. It is chiefly due to this second fact that the number of hybrids obtained was not larger. All of the experiments were carried out in separate breeding cages. The matings which yielded results were the following: TABLE I i No. Cross No. Year 9 sh of Offspring T Dor 1919 Yellow oniy x California linnet 3 P EE a Ee 1920 Yellow canary* x Willow goldfinch 5 E, PUDE 1920 Willow goldfinch X Arkansas goldfinch 4 The four hybrids resulting from cross No. 3 (willow goldfinch 9 X Arkansas goldfinch d) died a few days after hatching, and the female could not be induced to breed for a second time. These hybrids differed from ordinary newly-hatched finches and from the eight hy- brids obtained from crosses No. 1 and No. 2 in having exceedingly large abdomens, a condition which was prob- ably due to the fact that a large quantity of yolk had not been assimilated. Cross No. 1 (yellow canary 2? X California linnet 4) yielded three hybrids, one of which was accidentally killed when nine days old. During the same summer (1919) Mrs. L. V. Irelan of Berkeley, California, like- wise sueceeded in rearing a brood (2 males and 2 fe- males) of canary-California linnet hybrids? which the writer was able to compare with his own. Before going into detail regarding the coloration of these canary-California linnet hybrids, it seems desir- able to refer briefly to the plumage color of the paternal species, the California linnet. Both sexes of this finch are grayish-brown in color, but, when about three months old, the male turns rose pink, orange red, or scarlet about * The same female which was used in eross No. 1. 5In this case the mother was also completely yellow. No. 645] HYBRIDS OF THE CANARY 325 the head, neck, breast and rump. These colors increase in extent and brillianey with each molt. Males reared and kept in captivity never develop anything but a yel- lowish-buff color in these regions, and if a mature wild male is confined, its red color, during the molt, likewise becomes yellowish-buff. Both adults and young are con- spieuously streaked, especially the latter. The six* eanary-California linnet hybrids were all completely dark (self-eolored) until the first molt (fall 1919), and closely resembled young California linnets, but their plumage was less intensely dark than that of thelatter. During the fall molt of 1919 all of the hybrids became slightly ‘‘washed’’ (tinged) with yellow where the California linnet d is red (or yellowish-buff). This yellow tinge was more eonspieuous in the males than in the females and became somewhat more pronounced during the fall molt of 1920. All six eanary-California linnet hybrids are streaked, like the paternal and the ‘‘green’’ variety of the mater- nal species. As regards size and shape, they differ very little from the parents, both of which are similar in these respects. Their notes are intermediate in timbre be- tween those of the two parental species, the males hav- ing a more powerful song than the canary. In the spring of 1920 the writer paired two of these eanary-California linnet hybrids. Both showed an ar- dent desire to breed and the female exhibited consider- able skill in nest building. The first egg was laid on May 6, and several days later a second (May 10). Both of these eggs were only about half the size of canary or California linnet eggs’ and were dark-blue in color, and not speckled, while those of both parental species are bluish-white and speckled. Both eggs were placed under canary females, but proved to be infertile. The male 9 The hybrid which was accidentally killed was identical in coloration with these six. 7 This corroborates similar observations by Bechstein (1795, IV, p. 469) and Blakston (18807, p. 265), both of whom compare the eggs of canary- fineh hybrids with peas. 326 THE AMERICAN NATURALIST [Vor. LVI used in this experiment was also mated with a yellow canary, but, despite much treading, all eggs were clear. From cross No. 2 (yellow canary 9 X willow goldfinch d) five? hybrids were obtained. A few years before, Dr. H. C. Bryant of the California Fish and Game Commis- sion also succeeded in rearing a canary-willow goldfineh hybrid, concerning which he has been kind enough to furnish the writer with complete information. Before considering the plumage color of these canary- willow goldfinch hybrids, it seems again desirable to sketch briefly that of the wild finch. Both young and adults of the willow goldfinch are chiefly olive-brown and black in color, but the sexually mature male turns eanary-yellow during the summer, with the exception of the wings, tail and a small patch on the head, which re- main blaek. Neither young nor adults show any streak- ing.? The three canary-willow goldfinch hybrids reared by the writer are (January 6th, 1921) colored as follows: No. 1, completely dark (self-colored); No. 2, likewise, except for a few yellow feathers near the left eye; No. 3, dark, with a yellow band, about 5 mm. in width, run- ning across the head; No. 4 (reared by Dr. Bryant),'? dark, with some white feathers on the tail. All of the hybrids reared by the writer are conspicuously streaked, which, according to Dr. Bryant, was also true of hybrid No. 4. As regards size and shape, the writer’s canary-willow goldfinch hybrids closely resemble the canary (this was also true of hybrid No. 4), especially in shape of beak . and length of tail, in which respects there is a consider- able difference between the two parental species. As in 8 Two of these died shortly after hatching and hence furnished no re- liable data as regards coloration, 9 This is also true of the remaining North American members of the genus Astragalinus, the Arkansas and the Lawrence goldfinch (Astragalinus „lawrencei Cassin), except that in the ease of the latter, the lower parts of the young are indistinetly streaked (ef. Bailey, 1912, pp. 322, 323). 19 The eanary mother of this hybrid was also completely yellow. No. 645] ` HYBRIDS OF THE CANARY 327 the ease of cross No. 1 (yellow canary 9 X California linnet $4), the notes of the hybrids are intermediate in timbre between those of the parents. We now come to the question as to how these hybrids compare with other canary-finch hybrids, and in how far they conform with Mendel's laws of inheritance. It will be noticed that in the case of the eanary-California linnet hybrids, as in many mammalian crosses, dark color is completely dominant over light color, but the number of offspring (7) is too small to warrant the con- clusion that this will always prove to be the case. On the other hand, as regards the canary-willow goldfinch hybrids, there is no complete dominance of one color, the hybrids in this case showing a similar variability to that of canary-European goldfinch hybrids. Davenport (1908, p. 23) believes that the variability in plumage color of canary-finch hybrids is entirely due to the ‘‘mottling factor’’ of the yellow canary. He says (p. 23): It [the yellow canary] carries a mottling factor. Consequently when the yellow canary is crossed with a pigmented canary or with a finch the hybrids are mottled. In support of this Spem he makes the following statement: That it is the yellow canary which contains the motie factor and is the source of the variability of the hybrids is shown by the fact that (1) hybrids with the green canary do not vary in this fashion, and (2) hybrids between any two species of finches—of which many are bred by fanciers—are “ cast in one mold.” As regards the first of these two points, it may be said that one should not expect canary-finch hybrids from a ‘‘oreen’’ (self-colored) canary to show yellow markings as frequently as when a yellow canary is used. In regard to the second point, Davenport (1908) seems to have overlooked the fact that Blakston (1880?), on whose authority this statement was probably based, states only (p. 274) that all bullfinch-goldfinch ‘‘mules’’ are ‘‘cast in one mould." In fact one of Blakston’s (1880?) re- 328 THE AMERICAN NATURALIST [Vor. LVI marks clearly indicates that this is not true of the hy- brids between all species of finches, for on the next page (275) he makes the following statement concerning the ‘‘much more common”’ greenfinch-goldfinch hybrid: It is not a very pretty bird, . . . partaking to a considerable extent of its [the greenfinch’s] dull colour, though occasionally a more bril- liant example than usual, having a good deal of the Goldfinch char- acter about it, appears on the stage. Davenport’s (1908) conclusion therefore does not seem to be very well founded. i Results published by Galloway (1909) since the ap- pearanee of Davenport's (1908) paper seem to throw some light on this question. As already stated, this author (Galloway) obtained 172 dark (self-eolored) to 168 variegated (mottled) offspring from his canary- European goldfineh (Carduelis carduelis) crosses. How- ever, when he used the siskin (Carduelis spinus), a closely related but darker species, he obtained nearly three times as many (36 to 13) self-eolored as mottled individuals, that is, almost a 3 to 1, instead of a 1 to 1 ratio. These results, supported by those set forth in this paper, suggest that the frequency of mottling in canary-finch hybrids is not solely due to the yellow canary," but probably also depends on the coloration of the wild finch. LITERATURE CITED Bailey, F. M 1902. Handbook of Birds of the Western United States. The River- side Press, Cambridge. Bechstein, J. M, 17 Gemeinniitzige Naturgeschichte Deutschlands nach allen drey Reiehen. Vol. 4. Siegfried Lebrecht Crusius, Leipzig. Blakston, W. A. 1880?. The Illustrated Book of Canaries and Cage-birds. Cassell, London. 11 A similar problem exists in regard to the mottled seed-coat of the F, of certain pigmented-white bean erosses. Shull (1907) suggested that it is the white, and not the pigmented bean to which the mottling is due. How- ever, Tschermak (1904, 1912) has shown that in some cases it is the pig- mented bean which is the source of the mottling, a view which was later aecepted by Shull (1908, pp. 437—439). No. 645] HYBRIDS OF THE CANARY 329 Chapman, F. M. 1916. Handbook of the Birds of Eastern North Ameriea. D. Apple- Co., New York and London Darwin, C. 1885. The Variation of Animals and Plants Under Domestication. 2 Vols. John Murray, London, cont DB 1908. Inheritance in Canaries. Carnegie Institute of Washington, Publieation No, 95. Galloway, A. R. 1909. prodi Breeding. A Partial Analysis of Records from 1891- 1909. Biometrika, Vol. 7, pp. 1-43, 5 figs., Hünefeld, H. V. 1864. Ueber Bastardzucht zwischen Stieglitz und Canarienweibchen. Der Zo Garten, Vol. 5, pp. 139—144 Klatt G. T... 1901. Uber den Bastard von Stieglitz und Kanarienvogel Arch. f. Entwicke'ungsmech. d. Organismen, Vol. 12, pp. 414—453 and 471—528, 1 pl Shull, G. H 1907. otio dn Characters of a White Bean. Science, Vol. 25, pp. 828—832. 1908. A New Mendelian Ratio and Several Types of Latency. AMER. NATURALIST, Vol. 42, pp. 433—451. Meer E. v Weitere WD elaine an Erbsen, Levkojen und Bohnen. Zeitschr. f. d. landw. Versuchswesen in Osterreich, Vol, 7, pp. 53 n (After Shull. 1912. ee e an Levkojen, Erbsen und Bohnen mit Riicksi f die Faktorenlehre. Zeitschr. f. indukt. Abstam- mungs- od Vererbungslehre, Vol. 7, pp. 81-234, 12 figs. COEFFICIENTS OF INBREEDING AND RELATIONSHIP DR. SEWALL WRIGHT Bureau or ANIMAL INDUSTRY, UNITED States DEPARTMENT OF AGRICULTURE Ix the breeding of domestie animals eonsanguineous matings are frequently made. Occasionally matings are made between very close relatives—sire and daughter, brother and sister, ete.— but as a rule such close inbreed- ing is avoided and there is instead an attempt to concen- trate the blood of some noteworthy individual by what is known as line breeding. No regular system of mating such as might be followed with laboratory animals is practicable as a rule. The importanee of having a coefficient by means of which the degree of inbreeding may be expressed has been brought out by Pearl! in a number of papers pub- lished between 1913 and 1917. His coefficient is based on the smaller number of ancestors in each generation back of an inbred individual, as compared with the maximum possible number. A separate coefficient is obtained for each generation by the formula an+ Zn = 100 (1— 577^) —100 ü— gen) where q»/2"* is the ratio of actual to maximum pos- sible ancestors in the n + 1st generation. By finding the ratio of a summation of these coefficients to a similar summation for the maximum possible inbreeding in higher animals, viz., brother-sister mating, he obtains a single coefficient for the whole pedigree. This coefficient has the defect, as Pearl himself pointed 1 AMERICAN NATURALIST, 1917, 51: 545—559; 51: 636-639. No. 645] COEFFICIENTS OF INBREEDING 331 out, that it may come out the same for systems of breed- ing which we know are radically different as far as the effects of inbreeding are concerned. For example, in the continuous mating of double first cousins, an indi- vidual has two parents, four grandparents, four great grandparents and four in every generation, back to the beginning of the system. Exactly the same is true of an individual produced by crossing different lines, in each of which brother-sister mating has been followed. Yet in the first the individual will be homozygous in all factors if the system has been in progress sufficiently long; in the second he wili be heterozygous in a maxi- mum number of respects. In order to overcome this objection Pearl has devised a partial inbreeding index which is intended to express the percentage of the inbreeding which is due to relation- ship between the sire and dam, inbreeding being meas- ured as above described. A coefficient of relationship is used in this connection. These coefficients have been discussed by Ellinger? who suggests certain alterations and extensions by means of which the total inbreeding coefficient, a total relationship coefficient and a total re- lationship-inbreeding index for a given pedigree can be compared on the same scale. An inbreeding coefficient to be of most value should measure as directly as possible the effects to be expected on the average from the system of mating in the given pedigree. There are two classes of effects which are ascribed to inbreeding: First, a decline in all elements of vigor, as weight, fertility, vitality, etc., and second, an increase in uniformity within the inbred stock, correlated with which is an increase in prepotency in outside crosses. Both of these kinds of effects have ample experimental support as average (not necessarily unavoidable) conse- quences of inbreeding. The best explanation of the de- erease in vigor is dependent on the view that Mendelian 2 AMERICAN NATURALIST, 1920, 54: 540-545. 332 THE AMERICAN NATURALIST [Vor. LVI factors unfavorable to vigor in any respect are more frequently recessive than dominant, a situation which is the logical consequence of the two propositions that mutations are more likely to injure than improve the complex adjustments within an organism and that injuri- ous dominant mutations will be relatively promptly weeded out, leaving the recessive ones to accumulate, especially if they happen to be linked with favorable dominant factors. On this view it may readily be shown that the decrease in vigor on starting inbreeding in a previously random-bred stock should be directly pro- portional to the increase in the percentage of homozygo- sis. Numerous experiments with plants and lower animals are in harmony with this view. Extensive ex- periments with guinea-pigs conducted by the Bureau of Animal Industry are in close quantitative agreement. As for the other effects of inbreeding, fixation of char- acters and increased prepotency, these are of course in direct proportion to the percentage of homozygosis. Thus, if we can calculate the percentage of homozygosis which would follow on the average from a given system of mating, we can at once form the most natural coeffi- cient of inbreeding. The writer? has recently pointed out a method of calculating this percentage of homozygosis which is applicable to the irregular systems of mating found in actual pedigrees as well as to regular systems. This method, it may be said, gives results widely different from Pearl’s coefficient, in many cases even as regards the relative degree of inbreeding of two animals. Taking the typical case in which there are an equal number of dominant and recessive genes (A and a) in the population, the random-bred stock will be composed of 25 per cent. 4A, 50 per cent. 4a and 25 per cent. aa. Close inbreeding will tend to eonvert the proportions to 50 per cent. AA, 50 per cent. aa, a change from 50 per cent. homozygosis to 100 per cent. homozygosis. Fora natural coefficient of elc we want a seale which 3 Genetics, 1921, 6: 111-178. No. 645] COEFFICIENTS OF INBREEDING 333 runs from 0 to 1, while the percentage of homozygosis is running from 50 per cent. to 100 per cent. The for- mula 2h—1, where h is the proportion of complete homo- zygosis, gives the required value. This can also be written 1—2p where p is the proportion of heterozygo- sis. In the above-mentioned paper it was shown that the coefficient of correlation between uniting egg and sperm is expressed by this same formula, f — 1— 2p. We can thus obtain the coefficient of inbreeding f» for a given individual B, by the use of the methods there out- lined. The symbol r», for the coefficient of the correlation between B and C, may be used as a coefficient of relation- ship. It has the value 0 in the case of two random indi- viduals, .50 for brothers in a random stock and ap- proaches 1.00 for individuals belonging to a closely in- bred subline of the general population. In the general case in which dominants and recessives are not equally numerous, the composition of the random- bred stock is of the form a? 4A, 2zy Aa, y? aa. The per- centage of homozygosis is here greater than 50 per cent. The rate of increase, however, under a given system of mating, is always exactly proportional to that in the case of equality. The coefficient is thus of general ap- plication. If an individual is inbred, his sire and dam are con- nected in the pedigree by lines of descent from a com- mon ancestor or ancestors. The coefficient of inbreeding is obtained by a summation of coefficients for every line by which the parents are connected, each line tracing back from the sire to a common ancestor and thence for- ward to the dam, and passing through no individual more than once. The same ancestor may of course be involved in more than one line. The path coefficient, for the path, sire (S) to offspring (O), is given by the formula pos — àv (1+ fs)/(1 + fo), where fs and f. are the coeficients of inbreeding for sire 394 THE AMERICAN NATURALIST [Vor. LVI and offspring, respectively. The coefficient for the path, dam to offspring, is similar. In the ease of sire's sire (G) and individual, we have Po.g = pos peo — 1 V (1 + fo)/ (1 + fo), and for any ances- tor (A) we have for the coefficient pertaining to a given line of descent p«« — (4)"V (1+ f.)/(1 + fo), where n is the number of generations between them in this line. The correlation between two individuals (rw) is ob- tained by a summation of the coefficients for all connect- ing paths. Thus 1 n+" 1 š T (3) N(1 4-5) + be) where n and w are the number of generations in the paths from A to B and from A to C, respectively. . The formula for the correlation between uniting gametes, which is also the required coefficient of inbreed- ing, is fo = rav (1 — fs) (1 -= fa), where rsa is the correlation between sire and dam and fs and fa are coefficients of inbreeding of sire and dam. Substituting the value of rsa we obtain fo — > (3)"""(1 + fa). If the ancestor (A) is not inbred, the component for the given path is simply (1)""" where n and w are the number of generations from sire and dam respectively to the ancestor in question. If the common ancestor is inbred himself, his coefficient of inbreeding (fa) must be worked out from his pedigree. This formula gives the departure from the amount of homozygosis under random mating toward complete homozygosis. The percentage of homozygosis (assum- ing 50 per cent. under random mating) is 4(1 + f.) x 100. 335 COEFFICIENTS OF INBREEDING No. 645] 1oddipo (St9T) png uuo vputrior) 22:139) } uv[sey pIo JUBPIO A. (90611) jouoZvujute[q (6882) \ ugoy suoon(Q) oup, 9uojsprey ) T uv[SUj por JOWO 10]8Uouw'rT 119 19380[r) JO sSoqon(q (FPSEL) uv[sey pIo j } jourop) I9JSLIUL'T €pISS04/) (19602) 19027) OWL 9nj4tA (£991) JOWO 19jsu2u'v'rT 90jopgstye (98ST) puvlsuy yo worduvyy 116 1945019 JO ssoqon(q (95941) puepgugp jo uordureq) | | | | Qut, ) (92941) puvlsugq jo uordureq;) j snury, (88292) 00661 19485019 jo exnq puriy- ~ [esoy ssoou1,] (9862) T0603 1945019 jo exuq re&oq J (P8398) 9489F jo[junur) uuoj[ THE AMERICAN NATURALIST [Vor. LVI 336 (61) \ mq s; YooTy (gg) Itmq \ S, I0XIUq[ “37 (moo) 93L10AU,T (e93) equru([o,p J Kiioqwvuijg Sunox (¢92) oqurel[o,7 (61) \ [nq s,x»oopy (ze) Iq ! s, 191g "M oxnq uoj[eq uojqsnuvq 1i (sg) tiq s, 104189 "N (402) 93LI0AV,T xpueuqq? (£97) oquiel [og XUD (zeg) 9]1110A9,T J (98) oxoaq Buro g (A402) OPLIOAR A xudud) (292) equrefro,q ALMIGMBIYQ Sunog (98) ayorqsurpog J (292) oque fog xiupqq Sunox ) (zoz) »- 9]HI0AU,T (ger) 19W07 No. 645] COEFFICIENTS OF INBREEDING 337 By this means the inbreeding in an actual pedigree, however irregular the system of mating, can be com- pared accurately with that under any regular system of mating. As an illustration, take the pedigree of Roan Gauntlet, a famous Shorthorn sire, bred by Amos Cruickshank. This bull traces back in every line to a mating of Cham- pion of England with a daughter or granddaughter of Lord Raglan. For the present purpose we will assume that these bulls were not at all inbred themselves and not related to each other. Since the sire traces twice to Champion of England and twice to Lord Raglan and the dam once to each bull, there are in all four lines by which the sire and dam are connected. Common Ancestors Individual of Sire $ n n! |)" and Dam X (1 + ba) Roan Gauntlet Champion of England 45,276 (35,284) CF OSB ea 2 1 .062500 2 .062500 Lord Raglan (13,244). . 0 3 3 -007812 3 .007813 .140625 The coefficient of inbreeding comes out 14.1 per cent., a rather low figure when compared to such systems as brother-sister mating (one generation 25 per cent., two generations 37.5 per cent., three generations 50 per cent., ten generations 88.6 per Se) or parent-offspring ma- ting, (one generation 25 per cent., two generations 37.5 per cent. three generations 43.8 Eum cent., approaching 50 per cent. as a limit). As an example of eloser inbreeding, take the pedigree of Charles Collings’ bull, Comet. The sire was the bull Favorite and the dam was from a mating of Favorite with his own dam. As Favorite was himself inbred to some extent, it is necessary to calculate first his own coefficient of inbreeding. 338 THE AMERICAN NATURALIST [Vor. LVL Common Ancestors : Individual of Sire 5 n n (iu and Dam x (1 + fa) Favorite (252) mem id (2853)... S 0 1 1 .1250 Favorite (cow)........ 0 2 1 .0625 1875 Comet (115) sis (252). eb Leo 0 1 2969 PCIE (22 e 0 1 1 1250 ee Un dde 0 2 2 0312 Peet (COW) Oia. ch 0 3 2 .0156 | 4687 In the ease of Comet, Foljambe and Favorite (cow) each appears twice in the pedigree of the sire and three times in the pedigree of the dam. However, only those pedigree paths which connect sire and dam and which do not pass through the same animal twice are counted. The , listing of Favorite (252) and Phoenix as common ances- tors eliminates all but one path in each case as regards Foljambe and Favorite cow. The remaining paths are those due to the common descent of Bolingbroke, the sire’s sire and Phenix as the dam’s dam from the above two animals. By tracing the pedigrees back to the beginning of the herd book, the coefficients cf inbreeding are slightly in- creased. This meant going back to the seventh genera- tion for one common ancestor of the sire and dam of Favorite. The coefficient in the case of Favorite be- comes .192 instead of .188 and that of Comet .471 instead of .469. Remote common ancestors in general have little effect on the coefficient. It will be noticed that Comet has a degree of inbreeding almost equal to three genera- tions of brother-sister mating or an indefinite amount of sire-daughter mating where the sire is not himself inbred. THE ASSORTMENT OF CHROMOSOMES IN TRIPLOID DATURAS JOHN BELLING AND ALBERT F. BLAKESLEE STATION FoR EXPERIMENTAL EvoLUTION, Corp SPRING ecc Lone Istanp, New YORK The present article is the one of a number of proposed papers which will deal with the behavior of the chromo- somes in the different classes of Datura mutants, the correlation of the chromosomal differences with changes in structural and other characters, and with the ratios in which Mendelian allelomorphs are found in the off- spring. The method mainly used in the microscopical examination, and the general principles involved, are given in two papers already in press for THE AMERICAN NATURALIST, Sizes of Chromosomes.—The diploid Datura Stra- [A TO CA Second metaphases of a normal Datura in a pollen-mother-cell. h half sir constricted, ration in priate pressure so that the chromosomes were in optical contact with the cover- 339 340 THE AMERICAN NATURALIST [Vor. LVI monium shows, in the metaphase of the second division in the pollen-mother-cells (Fig. 1), two groups, each con- sisting of 1 extra large chromosome, 4 large, 3 large me- dium, and 2 small medium chromosomes, 1 small and 1 ex- tra small chromosome. Thus the somatic formula is 2(L4- : " à ^ a 9 C ry f Y Ave TE ini i Fic. 2. One second metaphase plate of a tetraploid Datura, in a pollen- spi bgo Late prophase of a normal Datura in a pollen-mother-cell. The size Apea are especially distinct, for the smaller chromosomes have con- densed earlier. Fic. 4. Late prophase of a Visi Datura. The largest chromosome set natin agg was f latest to conden FiG rophase of a eae Datura. The largest chromosome set is hook- [xdi ed. we late prophase or early metaphase trivalents often have the form of a ring with a ha nale; whieh is indicated in only one trivalent in Fig 4, and is not shown in Fig. 5.) /.4]124- 3M 4-2m -- S 4- s). Tetraploid plants have arisen, in rare cases, from these diploid Daturas (2). They show (Fig. 2) twiee as many chromosomes in each of the size classes, and have the somatic formula 4(L + 41 + 3M --2m-FS--s). Ont of many crosses of tetraploid No. 645] ASSORTMENT OF CHROM ES 341 Daturas by pollen from normals, 4 triploid plants have resulted (3). Their somatic formula is shown to be 3(L+ 4]--3M --2m + S +s). Similar results have been ob- tained for triploid hyaeinths by de Mol (7). Attraction of Homologous Chromosomes.—In the nor- mal Daturas the late prophase or early metaphase of the first division in the pollen-mother-eells shows 12 sets with two united chromosomes (bivalents) in each (Fig. 3). These bivalents can readily be arranged in the six size classes. In the corresponding stage of the triploid Daturas there are 12 sets of three united chromosomes each, and these trivalents can be arranged according to the size formula (Fig. 4). Sometimes two of the three rod-shaped chromosomes are united together at both ends, and the third is joined on at one end only, or the three may form a hook (Fig. 5). Some trivalents were seen by Osawa in triploid mulberries (8), and a group of 9 trivalents was also found in a triploid Canna (1). (The complete group of 9 trivalents has also been seen in 4 other triploid Cannas.) i : 4 È 2 " We se - ; $9 s» ug * E / E d $e... $ 8 [ 3 E 49 a $3 m Ac P e IB. ^ 06 e, 8$ Fic Second metaphases in a pollen-mother-cell of a triploid Datura. The large and large medium chromosomes were not separable in this oT Separation (Disjunction) of Chromosomes.—So far as seen in Datura, two chromosomes usually pass to one pole, and one chromosome to the other, from each triva- lent, as is the case in triploid Cannas (1). Assortment of Chromosomes.—From one triploid 342 THE AMERICAN NATURALIST [Vor.LVI plant both groups of chromosomes were counted in each of 84 pollen-mother-cells, which were in the second meta- phases, and showed no detached chromosomes (Fig. 6). The assortments are given in Table TABLEI ASSORTMENT OF CHROMOSOMES IN 84 POLLEN-MOTHER CELLS OF TRIPLOID Datura, 19729(1) Metaphase of Second Division LD49 1.48 L BM il. 15 16 17 18 Assortment of Chromosomes | + | + + Bow aepo 24 | 23 | 22 1, |. 20 19 18 | 1 6 13 | 17 26 20 19.0 Nos. of double groups | Calculated on duces orients- | 0.04 | 0.5 tion of trivalents......... Early anaphase of the second divisio Fic. 7. pol Datura. es had apparently n in a pollen-mother-cell of a a s (The upper €——Ü plate was shifted upwa tly been detached at the first of the 3 extra large chrom MCN and divided at the siecle division. Probably a tetrad with 2 micro- d cytes would have result No. 645] ASSORTMENT OF CHROMOSOMES 343 It is evident that.the orientation of the trivalents in the first metaphase must be nearly or quite a random one, as was suggested in triploid CEnotheras (5, 6) and mulberries (8), and as is the ease in triploid Cannas (1). (Nearly similar results were also obtained from a total of 58 single-metaphase plates from this triploid Datura.) Detachment of Chromosomes.—' Three buds ‘yielded 62 pollen-mother-cells with both seeond-metaphase plates countable, and among these there were six cells showing that one chromosome had been detached at the first ana- phase (Fig. 7), one cell showing detachment of two’ chromosomes, and one cell showing both one and two de- tached chromosomes. Thus there were about 13 per cent. of cases of detachment. These detached chromosomes (8) form mieroeytes when the pollen-mother-cells con- strict to form tetrads (Fig. 8). Table II shows the num- NC 3 ¢ Fic. 8. Tetrads, etc., of a triploid Datura. Above: (1) a normal tetrad; (2) a tetrad with one microcyte; (3) a tetrad with 2 microcytes, Below: (1) a tetrad with 4 microcytes; (2) two giant cells; (3) rare form with 6 not very unequal cells. bers of microcytes seen in nearly 3,500 tetrads from 3 triploid plants. The average is 13 per cent. of cases of detachment, but the variation in different buds appears too great to be due to chance alone. In 100 pollen-grains there would be about 5 microcytes. 344 THE AMERICAN NATURALIST [Vor. LVI Non-reduction.—In_ belated pollen-mother-cells the chromosomes in the trivalents assume the four-lobed con- dition of those in the adjoining cells which are in the metaphase of the second division. The first nuelear divi- TABLE II DETACHMENT OF CHROMOSOMES. NON-REDUCTION Pollen Tetrads of Triploid Plants. (Percentages) Regular Double-sized Microspores Microspores | i | | Pa | ied = — | | | | Dunis 4 4 4/4 2 | 2 2 cent- | | age of | Nos. o | | | Ete. |Cases of. Tetads No < Micro- | | SETS 1 2 12L41—1Llt:11.4 tach- Plant and Bud | | | 19729 (1) a..| 67.0 | 20.0 | 9.0 | 0.7 | 0.5 | 2.0 | 0.5 | 0.2 30.9 403 19729 (1) b..|91.5| 3.0} -5.3 | 0.2 | 8.5 436 19729 (1) c..| 90.3] 3.3 | 5.9 | 0.2 | 0.2 | 9.6 425 19729 (1) d.. 96.1 16] 09| — prs 1.4 | 2.5 433 20345 (1) a../ 83.8, 7.9| 8.1] 0.2 | | 16.2 444 20345 (1) b..| 97.8 | 0.5 | 0.7 | — | — | 0.7 | 0.2 13 412 20345 (1) c.. 98.0 0.8 —|—]|10! 1.1 400 20380 (1) -.. 65.5 | 19.0 | 14.6 | — | — | 0.6 | — | 0.4 34.0 542 Average..... 863| 7.0| 5,66|02/|0.1|0.7/0.1/0.1 0.03 | 13.0 Total | | 3,495 Microcytes to 100 Percentage of double-sized pollen-grains = 4.9 pollen-grains = 0.4 sion is entirely omitted, there is no reduction (8), and two nuclei with 36 chromosomes each are formed at the second division. The two cells which result are twice the size of the average microspores, and can be seen in the pollen as giant grains. Non-reduction may be greatly increased by transient cold. It averaged 0.4 per cent. in the tetrads. A hundred full pollen-grains were meas- ured at random,from each of 8 flowers on 4 triploid plants. The average was 0.5 per cent. of giant grains. Chromosomes of Functional Egg-cells.—In one triploid Datura, from three (or fewer) capsules pollinated by a normal, there were produced 75 mature plants, 67 of which had their chromosomes counted. No. 645] ASSORTMENT OF CHROMOSOMES 345 TABLE III CHROMOSOMES OF PROGENY OF TRIPLOID DATURA POLLINATED BY DIPLOID Nos. and Assortment of Chro- olo Soe ieee cia vue MO 12 13 14 - .13 24 18 + + At se ts ate gas ES — 12 12 12 13 12 18 Nes? OF Plante: Vu 24 33 J0- er ee eee Calculated on random as- sortment for 4096 ovules. 1 12 66 oF pe Se HPS ee a sc The number of normal progeny shown in Table III is much beyond expectation (on the hypothesis that orienta- tion of trivalents in the first division of the megaspore mother-cell is random), even if we allow the excessive total of over 4,000 ovules to 3 capsules. Detachment of chromosomes in the megaspore-mother-cells to the max- imum extent found in the pollen-mother-cells will only partially account for this excess. Similar results were obtained by van Overeem with triploid @nothera bien- nis pollinated by the normal (9). Triploid Inheritance.—The 75 progeny showed triploid or trisomic (not disomic) inheritance (2) of two probably independent pairs of genes, those for purple and white flowers, and those for prickly and smooth capsules. Distribution of Extra Chromosomes.—Among the 33 plants with one extra chromosome, cases were found where this extra chromosome was extra large, large, medium, small, or extra small. These plants showed 11 bivalents and 1 trivalent at the late prophase and early first metaphase. Ten different forms were recognized by external features among 30 of the 33 forms with an extra chromosome. (Three plants have not yet been identified.) Among these ten forms, 1 form (Globe) occurred 5 times, 3 forms (Buckling, Ilex, and Reduced) occurred 4 times, 2 forms (Glossy and Elongate) occurred 3 times, 3 forms (Rolled, Cocklebur, and Poinsettia) occurred twice, and 1 form (Mieroearpic) occurred once. The ex- pectations for each of 12 possible forms are presumably equal, namely 2.5. The Datura plants with 2 extra 4 346 THE AMERICAN NATURALIST [Vor. LVI chromosomes so far examined showed 10 bivalents and 2 trivalents at the first prophase. Thus the random assortment of chromosomes in trip- loid Daturas parallels the conclusions as to the random assortment of genes in triploid (trisomic) inheritance, and adds to the evidence for the chromosomal theory o heredity given by the cytological and genetic work on Drosophila (4) and other insects. LITERATURE CITED 1. Belling, J. 1921. The Behavior of Homologous Chromosomes in a Trip- loi nna. Proc. Nat. Acad. of Science, 7: 197—201. 2. Lieu A. F. 1921. Types of Mutations and their Possible Signifi- ice in ToS AMER. NAT, pi 254-267 $. lenis, A. lling, and M. E. Yurnbán, 1920. Chromosomal sahil yw Mendelian Phenomena in Datura Mutants. Science, 52: 388-390 - 4. Bridges, C. B. 1921. Triploid Intersexes in Drosophila melanogaster. : 54 5. Gates, R. R. 1909. The Behavior of Chromosomes in (Enothera lata X gigas. Bot. Gaz., 48: 179—199. 6. oni de: Tes 1911. Cytologische Untersuchungen einiger Bastarde von a gigas. Ber. d. Deutsch. Bot. Ges., 29: 160—160. 7. de Ma oc i^ 1921. Over het voorkomen van heteroploide varieteiten van H: gown: orientalis L, in de Hollandshe kulturen. Genetika, 3: 97-192, 8. Osawa, I. 1920. Cytological and Experimental Studies in Morus, with S ference to Triploid Mutants. Bull. Imp. Sericult. Exp, Sta. Japan, 1: 317—369. Ð. van Overeem, C. 1921. Uber Formen mit abweichender Chromosomen- zahl bei Œnothera. Beih, z. Bot. Centralbl., B. 38, Abt. 1, Heft 1, S. 73-113. (ESTRUS AND FECUNDITY IN THE GUINEA PIG DONALD B. TRESIDDER DEPARTMENT OF ANATOMY, STANFORD MEDICAL SCHOOL Tui study was undertaken at the suggestion of Pro- fessor Meyer, primarily for determining the numerical relation between the corpora lutea of pregnancy and im- plantations in the guinea pig. -Most of the animals used in this experiment were pur- chased from dealers, for it was impossible, in the short time at my disposal, to obtain young animals of uniform age and with the exception of a few guinea pigs raised in our laboratory, only approximate ages were known. The guinea pigs were housed in a well-lighted, sunny, heated room. Lantz, '13, reported that the optimum temperature for the guinea pig is 65°. Draper, "20, stated that they thrive best at temperatures between 50? and 70? and found young animals extremely susceptible to small changes in temperature; some of them dying when the temperature was lowered permanently from 60? to 58? F. However, I did not notice any marked differ- ence in the behavior or condition of extremely young ani- mals kept at a temperature of 50°. They showed every sign of vigor and no animals were lost as a result of this exposure. Indeed, I learned of guinea pigs kept in the open in unheated pens, sheltered only from wind and rain. These animals were said to thrive and to multiply at the customary rate, but no records were kept. In my own work I found that a few degrees above or below 50° seemed to make no appreciable difference in the behavior of the animals, and I hence am somewhat sceptical about the marked susceptibility of the guinea pig to cold, so often reported. The animals were fed dry alfalfa and barley daily and green vegetables about twice a week. Many writers have reported that guinea pigs did not do well on dry feed, but it was my experience that, if fed an abundance of water, they throve on alfalfa and barley alone. Since they are subject to intestinal disturbances, it is of con- 347 * 348 THE AMERICAN NATURALIST [Vor. LVI siderable importance that they be fed with the greatest regularity. Several animals were lost during the course of the ex- periments and in each case a necropsy was performed. Illness, in several of the animals, extended over a period of weeks. They lost steadily in weight, and tended to assume a characteristically crouching attitude. The fur became rough and tousled. Some of them chewed in- eessantly, although some pain seemed to be associated with the process. The full significance of this behavior was made clear at the necropsy. Guinea pig No. 7, for example, which succumbed after an illness of three weeks, had an empty stomach, and the abdominal cavity was absolutely devoid of fat. There were no macroscopic signs of infection or disease. Examination of the teeth revealed that the upper incisors were worn down almost to the gums, with a more than corresponding increase in the length of the lower incisors, making occlusion of the molars impossible. The molars were loose and could easily be picked from the jaw with an ordinary labora- tory forceps. : The body of guinea pig No. 12, which died with prac- tically the same symptoms, showed extreme atrophy and emaciation. Ascaroid parasites were found in the rectum. The upper incisors were loose and worn and the short stumps remaining could be be removed with the fingers. The upper and lower incisors were separated by about 8 mm., due to the fact that the molars occluded first and pre- vented the short, probably fractured incisors from meet- ing. From the findings in these cases it would seem that guinea pigs may die of starvation because of the presence of worn or irregular teeth and consequent inability to masticate food. It may perhaps be that the changes in the teeth of these animals were due to senility, but fur- ther observations are necessary to confirm this before a definite answer can be given to the question. In order to study daily stages in the pregnancy of guin- ea pigs it became necessary to mate a large number of animals and to know the exact time of copulation. Stock- ard and Papanicolaou, 717, studied the estrous rhythm of No. 645] FECUNDITY IN THE GUINEA PIG 349 the guinea pig by making microscopic examination of the material found in the vagina. They found that ‘‘ Guinea pigs kept in a state of domestication and under steady environmental conditions possess a regular dicstrous cy- cle, repeating itself in non-pregnant females about every sixteen days throughout the entire year, with probably small and insignifieant variations during the different seasons. Each period of sexual activity lasts about 24 hours and is characterized by the presence of a definite vaginal fluid which is not sufficiently abundant to be read- ily detected on the vulva, but is easily observed by an examination of the interior of the vagina." They added that macroscopic signs of heat are unreliable. In my work it was found impractical to determine the existence of heat microscopically and the knowledge that heat should recur about every fifteen days furnished a starting point. . Each female was given a number and en- tered on an individual record sheet giving the following data: Date and hour of attempted mating. Result of attempted mating. Each time the animal eame into heat the record showed: Whether heat was recognizable by macroscopic examina- tion. Number of days since last heat. Number of hours since the first successful coitus. Number of hours that external signs of heat could be ob- served by examination. Matings were attempted daily, whether the animal was supposed to be pregnant or not. The males were intro- duced into the pens with the females regardless of whether or not the latter were thought to be in heat, and they were allowed to remain with the females from five to fifteen minutes. It was easy to follow the dicestrous cycle of any individual animal. A glance at the guinea pig’s record each day showed the number of days since the last heat, and, knowing that heat should return about the fifteenth day, it was practically impossible for it to come and go unnoticed unless it recurred altogether irregularly. We 350 THE AMERICAN NATURALIST [Vor. LVI found that after some practise heat could be determined rather accurately by inspection. A guinea pig in rut will often assume the position of copulation when stroked gently over the lumbar region. The vulva are swollen and moist, and often a cheesy secretion is seen. The latter is a positive sign of heat, but we found that some guinea pigs refused to mate during the entire period in which the secretion was present. In young animals we found heat recurring every fifteen or sixteen days with very little variation among indi- viduals of the same age. Three striking exceptions in which heat returned in twelve days will be reported later in this paper. Papanicolaou and Stockard found that in old multiparz the period may be lengthened to 18 days. I also found that as the animals grow older they seem to become more and more irregular in their rhythm. In three very old animals I was unable to find any signs of heat throughout an entire year, although I attempted to mate them twice daily. Three other animals maintained a cycle of 20 days, and in some cases we were unable to demonstrate any regular estrous rhythm at all, either by inspection or by the use of a male. Subsequently (1920) these workers have reported that ‘* underfeeding with a diet of 20 grams of carrots per day produces prolongation of the dicstrum, and at the same time a congestion in the ovary and uterus and a de- generation of developing Graafian follicles.” They con- eluded that ** the extent of prolongation of the dicestrum depends upon the stage at which an animal is underfed. . . . Large follicles seem to require better nutrition than a small primary follicle. . . . Thus a late underfeeding has a more injurious effect than an early one, and post- ponement of the next cestrus is correlated with a postpone- ment of new ripe follicles in the ovary." Stockard and Papanicolaou believe that the ovarian follicles are ex- tremely sensitive to environmental conditions. They be- lieve that extreme variations in the estrus cycle of cer- tain animals may be accounted for, partially at least, by differences in nutrition. In the course of these observations the intervals be- No. 645] FECUNDITY IN THE GUINEA PIG 351 tween attempted matings were shortened, with the idea that heat might be recurring unnoticed, but mating never oecurred at other intervals. It is doubtful whether any definite rhythm is maintained by old guinea pigs, for pig No. 9, which was observed to be in heat December 27, was not in heat again until 49 days later. Animal No. 20 was in heat October 17 and heat did not return until 91 days later. In another instance heat returned after 118 days. However, since the age of these animals is not known, it is impossible to be sure that these irregularities are due to senility. Bischoff, '44, stated that copulation in the guinea pig occurs within 3 hours after parturition. In four cases in which he prevented copulation heat returned after inter- vals of 40, 50,51, and 51 days. Hensen, ’76, and Rein, ’83, claimed that the most favorable time for copulation is within one hour after parturition. I observed copulation in 12 animals immediately after parturition. Matings were attempted at one-hour intervals for six hours after- ward. In four cases I was unable to mate the females at this time. They were found in heat again 34, 36, 81, and 120 days later. The first two animals were about six months old. The last two were very old, judging by their . teeth. Two females mated 1 hour after parturition, 2 after 2 hours, 1 after 3 hours, 1 after 5 hours, and 2 after 6 hours. In three cases no pregnancy resulted and heat returned in 31, 31 and 29 days. Many writers have reported that females refuse the male shortly after the first copulation. The inference is that some nervous mechanism automatically terminates heat soon after copulation. Instances have been reported in which the female refused the male 20 minutes after the first copulation. In observations extending over nearly a year, however, three cases were observed in which the female mated again eight hours after the first copulation. In the majority of cases the female permitted copulation three hours after the first mating. One animal mated 13 times in an interval of 8 hours. It seems that a female accepts the male at any time during the first stage of heat regardless of any previous intercourse, but apparently 302 THE AMERICAN NATURALIST [Vor. LVI she permits matings somewhat reluctantly after this.’ In- stead of assuming the position for copulation when ap- proached by the male she often runs around the cage and resists vigorously. Unless the male is very persistent and active copulation will not occur. One female resisted a second coitus for fifteen minutes by kicking, snapping, ete., only to stop suddenly and take the position for copu- lation. This behavior of the female may be due to pre- vious mating or it may simply mean that the period of heat is subsiding. I am inclined to the latter view, be- eause we have eneountered many females among animals which had not been previously mated, who resisted the males vigorously for a time, only to yield in the end. The time during which the females permitted copulation un- hesitatingly was a relatively short one, but after this phase had passed the animal might yet be mated if the male was persistent. Stockard and Papanicolaou, '17, are of the opinion that among domestieated guinea pigs only a slight seasonal variation exists in the occurrence of heat, but in the pres- ent series of guinea pigs the fall months were the most favorable for matings, as shown by the following table: Month Number of Resulting Percentage Matings Pregnancies alo ilie ee MANI AO 23 21 91.3% Octabum i P oro UY S 17 10 58.8 November... a a V LAS I 8 3 37.5 December oL. 11 4 36.3 JaDuAMY uL LL ee: 7 4 57.1 Febrtudz9 [105420228 eee coe 5 0 0.0 Mare; io eo LES 8 | 4 50.0 Apüho ui C d 12 | 5 41.6 MAS ess s xS a 9 5 55.5 PURO PA oU INL 6 5 83.3 The males seemed to be partly responsible for this wide variation. During the winter months they were lethargie and indifferent. When placed in a pen with a female known to be in heat, the male often ignored her, eating unconcernedly instead. In many instances several males had to be placed with such a pig, in succession, before a mating took place. This is in marked contrast to the customary behavior, for when placed in a pen with two No. 645] FECUNDITY IN THE GUINEA PIG 353 females, the male will often go directly to the female in rut. Sometimes, however, he will mistakenly pursue the one that is not in heat, although repelled by sharp bites and other negations, only to wheel suddenly and mount the receptive female. The pursuit of the wrong animal may only serve to stimulate him, but in some instances it was necessary to remove her before he would turn his at- tentions to the one in rut. Puzzling sexual idiosynerasies also were noted. Instances were observed, for example, in which a male would not under any cireumstances mate with a certain female which was in heat, although he was persistent in the ease of others. On the other hand, some females also were noticed to repulse a certain, male, al- though accepting others. It will be seen from Table I that 106 matings resulted in 61 pregnancies, or 57.5 per cent. Draper, '20, reported that only 40 per cent. of the animals bred by him became pregnant. Since Stockard and Papanicolaou found 95.4 per cent. out of 88 pigs pregnant, considerable variation would seem to exist. The large discrepancy between their results and ours may be due to the fact that the latter were working with uniformly young, selected animals or that the males were left to remain with the females, in- stead of being removed after several copulations. In the many matings not followed by pregnancies, the next estrous cycle was prolonged. This is shown by the accompanying chart. Guinea-pig Number “Heat returned after Do Car ek we UNA MAR LIE LER E 30 days. SE E E ve xe. Lh KA ERA E TA EY dex EE LrgNtog ee uuu cre ET 28 OU ie M S a et eee 15 DU ouis E E S E E dta VERS 46 Be es ese E E 29 eee ees 30 E ee Rb. 31 dU diokeeko ves a ere ess Eee oe eet 12 SÉ cicer creas Gas ciel 15 Be ea se ee eR Tru E 29 DIAC. Vine ee UERE VEGAS ER ee eet eey 30 Bi ei cadi ee ee ae 15 354 THE AMERICAN NATURALIST [Vou LVI As noticeable in the above chart, the lengthened dices- trous periods are nearly exact multiples of 15, the nor- mal period, thus showing that the cycle is definitely pe- riodie as reported by Stockard and Papanicolaou, "17b. Long, ’15, found that the cestrous cycle was prolonged by inserting a glass rod in the vagina of the rat. He held this prolongation to be due to a stimulation of the cervix of the uterus. Although I stimulated the uterus of guinea pigs by means of a warm glass rod in three cases only, heat returned in 15, 15 and 16 days, and I regret that I was not able to extend this series of experiments in order to obtain more data on this interesting phenomenon revealed by Long in the rat. However, from the above table, it is clear that copulation definitely prolongs the next œs- trous cycle in the majority of cases. This may be due to direct stimulation of the cervix of the uterus, as ex- plained by Long, or implantation may have occurred, fol- lowed by abortion or by absorption of the young concep- tuses, in cases in which the period was greatly prolonged. Guinea pig No. 39 (see Table II) was mated two hours after parturition, but no pregnancy followed. This ani- mal was remated 12 days later, with resulting pregnaney. This confirms a case reported by Rubasckhin, 705, in which heat returned 10 days after parturition. Stockard and Papanicolaou, in considering Rubasckhin’s report, re- garded 10 or 12 days as too short a period to indicate the return of heat. Nevertheless, in the ease reported here heat was unmistakable, and this animal which was mated 12 days after parturition became pregnant. I observed heat to return in 12 days also in two other pigs. Young animals constantly in association with males became pregnant at an earlier age than females isolated from males. Of a litter containing 3 females and 1 male, two females were placed in separate cages a few days after birth and the remaining male and female were allowed to run together. At the age of 5 months, the latter produced a litter. This indicates that the mating of this pair occurred before the animals were three months old. Yet no ill effects of this early mating or of the inbreeding could be detected in the offspring. No. 645] FECUNDITY IN THE GUINEA PIG 355 When the two sisters were two months old, males were introduced into the pens twice daily, but no signs of heat were observed, and no matings occurred until these fe- males were five months old. Similar results were obtained with two other litters. Since my work was done Mr. Warnock, a fellow student, has observed two females to — bear viable litters at the end of the third month. "This implies mating at the early age of one month. The pa- ternal male was several months older, however. Tur CORPORA LUTEA or PREGNANCY In order to study the correlation between corpora lutea and implantations during the various stages of preg- nancy, animals were mated and killed, from the seventh day of gestation on, for each day up to and including the fifteenth. From the fifteenth day to full term, animals were killed every other day. When the guinea pigs were killed, the ovaries and uteri were removed and placed in formalin for twenty- four hours and the number of embryos in each horn of the uterus recorded. The ovary corresponding to the horn of the uterus having the larger number of concep- tuses was arbitrarily chosen for use in determining what relation might exist between the number of conceptuses and the number of corpora lutea. Thus guinea pig No. 10 had two conceptuses in the right horn and one in the left. The right ovary was embedded and eut serially into thick sections. The left ovary was cut 7 micra thick for the study of changes in the corpora lutea during preg- nancy. In a study of 14 embryos, Draper, ’20, found 76 in the left horn and 69 in the right, a ratio of 1 to 0.9. Of 98 embryos from 35 guinea pigs, I found 55 in the right horn and 43 in the left, a ratio of 1 to 0.78. The average num- ber of feetuses per pregnancy was three. Table II shows that there is a marked agreement be- tween the number of eorpora lutea in an ovary and the number of implantations in the corresponding horn of the uterus. Out of 34 ovaries examined, the number of eorpora lutea was the same as the number of embryos 356 THE AMERICAN NATURALIST [Vor. LVI in the eorresponding horn of the uterus in all save six eases. In five of these six instances there was one em- bryo-less in the horn of the uterus than there were corpora lutea in the ovary. In the other case, the right TABLE II T : Embryos (Corpora Lutea Guinea Duration of Remarks Pig Pregnane Right; Left | Right, Left 25838 150 hes 2 se 7 1 1 3 1 Well-formed but p COO eM 8 0 3 1 3 no éxternal ine dd. iuis 9 2 1 2 f of implantations. nis NE MT 10 1 8 1 3 Bb ri 11 1 1 1 3 Well-marked evidence PAPOEA 12 8 0 3 0 of resorption. B vell 13 1 2 1 2 SN I IIO 14 g 1 3 1 2j lx ads 15 0 1 0 2 20:5... 1 x v 17 3 0 2 0 285. es 19 1 3 1 3 CA obi 21 2 1 2 1 Sd. cc rv 23 2 8 2 3 29 vuulou. 25 3 0 3 0 Zi 1 2T 2 0 2 0 NE ld i4 29 a 1 3 1 rp e ee qe 31 +2 1 3 1 Roo a 33 2 0 3 1 if 1 s 35 2 1 2 0 18.257. 2521 37 2 1 3 | 16250: 39 2 1 2 M Uae Se 41 1 2 1 2 PRO QE UM 43 i 2 1 2 : 12 311.5 5. 45 1 2 2 2 Conceptus on left side bi ces 47 I 1 2 1 almost completely re- wW ooa 49 2 1 2 1 sorbed. 9: 2: 0. 09 51 2 1 2 1 Son 53 1 2 1 2 Ti. 55 2 0 2 0 joa uo eS 57 1 1 2 1 be ae 59 1 2 2 2 L2 T QUIM 61 2 I 2 Siu SEVA 63 3 1 3 d lil eres Term 2 1 2 1 { AG 23 0 0 2 OIA OL of areni but Boi. 45 0 0 1 B | noimplantatio nd. . horn showed 3 embryos although only two corpora lutea of pregnancy were present in the ovary. Hence, in this ease, two embryos developed from a single ovum or a single follicle contained two ova. In the instances where there was one more corpus luteum than embryos it is possible that another conceptus was present and became No. 645] FECUNDITY IN THE GUINEA PIG 357 absorbed or that an ovum degenerated before implanta- tion, or that it failed of fertilization. As shown by Meyer, 717 and ’19, and Stockard and Pa- panicolaou, 718, absorption is not uncommon in the uteri in guinea pigs. In this series, three embryos which were clear-cut cases of absorption were found upon examina- tion of the uteri after their removal. In No. 12, which was killed forty-five days after copulation, two normal embryos were found in the left horn, but in the right horn there was nothing but a small mass which had undergone almost complete absorption. According to Stockard and Papanicolaou, ’18, embryos eight or ten days old may be detected by ‘‘ carefully feel- ing the uterus through the body wall of the mother.” They report a case as follows: A normally developed embryo 19 mm. crown rump length is shown in Fig. 6 and near it is seen an amorphous embryonic mass 2 mm. in longest diameter which represents the other member of the litter. . The entire mass of the smaller ovum in the uterus was about that of a ten-day specimen, while the normal individual was a typi- cal 20-day specimen. This case was detected by external examination and was merely opened in order to use the embryos for illustrating the phenomenon. Although I used the method of Papanicolaou and Stock- ard in palpating guinea pigs, in no instance was I able to determine the number of embryos with certainty under fifteen days. Because of this fact, I found it necessary to sacrifice the animals in order to determine the number of implantations before this period. Guinea pig No. 35 and guinea pig No. 34 were killed seven and eight days after conception, respectively, and the uteri removed. Careful palpation of the removed uteri failed to reveal the number of conceptuses. The uteri were then opened, but in order to determine the number of implantations present it was necessary to em- bed them and make serial sections. From this I am led to question the possibility of determining the number of em- bryos in the uterus by palpation through the abdominal wall on the eighth to tenth day of pregnancy. This skep- ticism seems warranted, further, by the measurements of 358 THE AMERICAN NATURALIST [Vor. LVI three ten-day conceptuses, 6.5 X 3 mm.; 6.8 X 4.5 mm.; '6.5 X 4.5mm. respectively. Draper gave the estimated length of an 11-day embryo measured under magnifica- tion as 2 mm. Stockard and Papanicolaou (1918) likewise reported that a ‘‘ slightly cystic ovary ”’ has frequently been diag- nosed by palpation through the abdominal wall of the guinea pig. In my observations 23 out of 75 ovaries were found to be cystic; but the largest cyst measured only 1.6 mm.X 1.68 mm. and not even this could by any chance have been palpated through the abdominal wall. Hence, it would seem that Stockard and Papanicolaou must have been dealing with markedly large and unusual, rather than with slightly, cystic ovaries. From a study of a large series of gestations in the domestic pig, Corner, ’21, concluded that internal migra- tion of ova is relatively common. This small series of pregnancies in the guinea pig furnishes very little evi- dence upon this question, for such a possibility is sug- ^ gested only by No. 17, a pregnancy of 35 days in which there were 2 corpora and 2 implantations on the right side and no corpora but one implantation on the left side. Since the total number of implantations in this case ex- ceeds that of corpora, one must assume that one ovum divided or that one follicle contained ova and that one of the ova arising from the right then migrated to the left cornu. However, since this pregnancy was so far advanced, this assumption implies that a corpus luteum of pregnancy in the guinea pig can not be wholly resorbed in 35 days and that it never fails to form. It is of special interest in this connection that a second case of this kind has been observed in this laboratory by Miss Clark. In this case there were two. corpora in the left ovary and none in the right, with ore implantation on each side. Since this pregnancy was only 17 days old, the question of early resorption of the corpus luteum probably can be excluded with considerable certainty but that of failure of the corpus luteum to form, remains. No. 645] FECUNDITY IN THE GUINEA PIG 359 REFERENCES Bisehoff, H, L. W. 1844. Beweis der von der Begattung unabhängigen periodischen Reifung und Loslósung der Eier der Süngethiere und des Entwicklungsgeschichte des Meerschweinchen Corner, George N. 2 Tate Migration of the Ovum. J. H, H. B., Vol, 32. Draper, 1920. The Prenatal pav of the Guinea Pig, Anat. Rec., Vol. 18, . 4, May, 1 Evans, H. M. id Long, J. a 1920. The Œstrus Cycle in the Rat (et seq.). Anat. Rec., Vol. 18. Lantz, David. ——. The Raising of the Guinea Pig. U. S. Dept. of Agriculture, armers’ Bulletin, Long, J. A. . 1919. The CEstrus Cycle in Rats. Proc. Am. Soc. Zool., Anat. Rec., Vol. 15, 1919. Meyer, A. W. 1917.. Intrauterine Absorption of Ova. Anat. Rec., Vol. 12, 1917. 1919. Uterine, Tubal and Ovarian Lysis and Resorption of Con ceptuses. Biol. Det, Vol. 36, April, 1919, Rein, . 1883. Beitrage zur Kenntnis der Reifungsercheinungen und Befruch- tungsvorgange am Saugetierei, Arch. f. Mikr. Anat., vol. 22. Rubaschkin, 1905. Ober die Reifungs- und Befruchtungs- processe des Meersch- weincheneies. Anat, Hfte., Abt. 1, Vol, 29, 1905 Stockard, C. R. and Papanicolaou, G N. 1917a. des sat re of a Typical CEstrus Cycle in the Guinea Pig, h a Study of im poss nr and Physiologieal Changes. Jour. Anat., , No. 2, Sept., 1917 (a). 19175. ^ "Bhythmical ro pe " MU ' in the Guita Pig. Science, N. S., Vol. 46, pp. 42, 44, 1917 (b). " Stockard, C. R. and Papani anieolao u, G. N. 1918. del BEE on the Modification of the enron in Mam e Effect of Alcohol on Treated G a Pigs their geri Tae Jour, Exp, Zool., Vol. XXVI p. 119, ay, 1919. The Vaginal Closure agii. Copulation and the Vaginal Plug in the Guinea Pig, Further Considerations of the Œstrus Rhythm, Biol. Dull, Ye XXXVII, Oct., 1919, Stockard, C. R. and Papanicolaou, G. N. 1920. Effect of oap on Ovulation and the CEstrous Rhythm Guinea Pigs. Proc. Soc. Exper. Biology and Medicine, XVII, iu. Hensen, iced " fBeobechtunges über die Befruchtung und Entwiekelung des Kanin eerschweinch Zeit. f. Anat. u. Ent- wick., Bd. I, 1876. VARIATIONS IN THE NUMBER OF VERTEBRA AND OTHER MERISTIC CHARACTERS OF FISHES CORRELATED WITH THE TEMPERA- TURE OF WATER DURING DEVELOPMENT CARL L. HUBBS Museum or ZooLocy, UwivERsITY OF MICHIGAN I Fon several years I have been studying the correla- tions between altered environmental conditions and the number of vertebre and other segmentally arranged struetures in fishes. Johannes Sehmidt, of the Carls- berg Laboratory in Copenhagen, has been carrying on a series of intensive investigations (see bibliography) which deal with the same problem, and which are for the greater part rather closely paralleled by my own studies. Both of us have obtained, independently, a rather large volume of experimental and observational evidence indicating that the meristie characters displayed by an individual fish are determined not alone by heredity, but in part also by the environmental conditions, particularly tempera- ture, which prevail during some sensitive developmental period. II The present study is one of those comprising the series just mentioned. It deals with variations in the number of vertebre, scale-rows and fin-rays within one year-class and between two successive year-classes of the lake ** shiner," Notropis atherinoides (Cyprinide), and in comparison between the corresponding year-classes of the ‘* blue-gill’’ sunfish, Lepomis incisor (Centrarchide). These variations appear to be correlated with differences in temperature prevailing eons the several develop- mental periods involved. The material of each species is probably a unit as re- 360 No. 645] VERTEBRJE OF FISHES E gards ‘‘ race." It was all obtained in a lagoon in Jack- son Park, Chieago, during the third week of December, 1919. At this time what seemed to be the entire fish population of the lagoon was eongregated in an opening, about five meters wide, in the ice-along shore. These fishes showed symptoms of asphyxiation. They were so abundant that at times, while they were gyrating about, the mass of fishes below would foree the almost solid upper layer a centimeter or two above the surface over an area of perhaps a square meter. A water bucket was filled with fishes, mostly Notropis atherinoides, by two or three sweeps of a small hand-net. More than one thou- bau rj 19196 o E 15 a t Length to &audal Rn, mon. Fic. 1. Frequency graph, indicating the year-classes of Notropis atherinoides. t sand of the young of that year (1919) of the Notropis were saved after random selection, and preserved for study with all older fish of the same species. All of the sunfishes (Lepomis incisor) obtained at the same time and place were preserved and studied. Of the two spe- cies, the sunfishes belonged to a population practically confined to the lagoon, while the minnows had moved into the lagoon, late in the preceding autumn, from the more open waters of Lake Michigan. The specimens thus obtained were grouped into year- classes. Age determinations were made by the usual methods of counting the annuli (winter lines) on the scales, and as a check the seasonal bands of the otoliths, and furthermore by the preparation of a frequency graph 362 THE AMERICAN NATURALIST [Vor. LVI from the length measurements of the entire material. The young of the year (obtained in 1919) are referred to as the 1919, class; those of the previous year as the 1919, class, and so forth. The 1919, year-class of the Notropis atherinoides is further divided into three subelasses, A, B and C, named in the direct order of hatching, hence in B- lndivtdùsis 2 4 4 T (0 V 1 * Fic. 2. Frequency graph illustrating the year-classes of Lepomis incisor. the indirect order of size. The year-classes for both spe- cies are indicated on the graphs forming Figs. 1 and 2. The symbols on the curve for the 1919, class of each spe- cies indicate the sex predominant among the representa- tives of each size. III A series of water temperatures appear unavailable, but in the case of such a shallow, nearly enclosed lagoon the air temperatures of the region may safely be substituted. Hence the Climatological Data (Illinois Division, 1918 and 1919) for Chicago were used in constructing Fig. 3; the temperatures given for each week were obtained by averaging the daily means. On the temperature chart there’ are indicated the periods of development for each of the two species as ob- No. 645] VERTEBRÆ OF FISHES 363 served at the same locality in 1919. The data for Lepomis incisor seem satisfactory (see Hubbs, 1919), but those for Notropis atherinoides are less complete and more cir- eumstantial. In the ease of the minnow, the develop- mental period is divided into three periods (A, B and C) corresponding with the three subclasses into which the 1919 year-class has been divided. Period A followed an inshore spawning migration of the mature individuals, coincident with the rapid rise in temperature during March; period C preceded the withdrawal of the breeding stock from the shore waters of the lake; the intervening period is termed B. The limited field observations on the spawning and developmental period for Notropis atherinoides during 1919 are, fortunately, strongly confirmed by a study of E Io18 Big e : Tempe rature Oo Jan. || Feb. | Mar. | Apr | May | Jupe | July | Aug Io 15 20 25 30 ‘WEEKS Fic. 3. Air temperature at Chicago, 1918—1919. the scales. The scales of the largest specimens taken in December, namely those comprising subclass 1919,A, and forming a distinct mode in the frequency graph (Fig. 1), show a well marked nuclear area of weak concentrated circuli indicative of retarded growth, followed by the coarser, more regular circuli indicative of normal summer growth. This initial period of retarded growth presum- 364 THE AMERICAN NATURALIST [Vor. LVI ably eorresponds with the cold period in April (see Fig. 3). The scales of the medium-sized specimens (subclass B) show on the average a narrower nuclear area suggest- ing slackened growth. It is presumed that these individ- uals passed through their early development toward the end of this cold period. The scales of the smallest spec- imens, those of subclass C, show no such nuclear area of weak concentrated circuli. These fishes supposedly developed during the warm weather of May. The data on the developmental period of these two species for the preceding breeding season (1918) are less complete than those for 1919, yet not wholly lacking. Lepomis incisor, at least, bred during the corresponding weeks in both years (but in less abundance in 1918 than in 1919). A comparison of the available observational data with the temperature chart (Fig. 3) indieates that, on the average, the developmental period for Notropis ather- inoides was colder in 1919 than in 1918, whereas these temperature relations were distinctly reversed in the case of Lepomis incisor, and furthermore, that the tempera- ture was distinctly higher at the beginning and toward the close of the 1919 breeding season for the Notropis, than during the middle of this period. IV These differences in the developmental temperature ap- pear to be correlated with variations in the number of segments in the case of both fishes. Comparisons will first be made between the two year groups of Notropis atherinoides, then between the same year groups of Lepomis incisor, and finally between the three subclasses into which the 1919 brood of the Notropis has been di- vided. The vertebrz in the 1919, class of Notropis atherinoides are sufficiently more numerous on the average than those of the 1919, class to shift the modal number from 41 to 42, the average from 41.41 (+ 0.04)* to 41.74 (+ 0.015). 1 The probable error of the average. No. 645] VERTEBRZ OF FISHES 365 SPECIMENS t ES 9/9/92 3 E i ME 38 39. m T i 43 D? VERTEBRAE! FI Comparison of number of vertebre in successive year-classes of Notropis atherinoides. The portion of the vertebral column affeeted is the eaudal, not the precaudal M plea division: the averages for the precaudal vertebre are 22.82 (+ 0.02) for 1919, and 22.85 (+ 0.01) for 1919, for the caudal vertebrz, 18.60 (+ 0.035) for 1919,, and 18.87 (+ 0.01) for 1919,. Simi- larly, the number of scales in the lateral line averages higher in the 1919, lot : the modal numberis 40 rather than 39 as it is in 1919, class; the average number is 40.05 (+ 0.04) rather than 39.65 (+ 0.04). The modal number wv S SPECIMENS . es S YS Fic. 5. Comparison of variati n number of branched and anal rays in different year-classes of Notropis E ME 366 THE AMERICAN NATURALIST [Vor. LVI of branched anal rays? is 10 in the 1919, series, 9 in the 1919, class; the averages are 9.52 (+ 0.05) for 1919, males, FREQUENCY TABLE I COMPARISON OF THE MERISTIC FEATURES OF THE 1918 AND 1919 BROODS OF Notropis atherinoides Character Year-class Total Number of Vertebrze : | : | 39 | ao | 4 | 42 | 4 | 44 i | — | a7 | s s | 13 | 1901905. 1. 1 4 356 539 | 137 | 12 ! Number of Precaudal Vertebre 21 | 22 23 | 24 | 25 | IBI. ae des 3 48 | 191: n — 101905... s 2 240 | 766 | 78 1 | Wither ad 6 | 74 L-38 os VG aie d 6 | 269 | 661 | 142 | 10 | | | Number of Seales in Lateral Line a | x | » | o | a | 2 | oo | ow | os 389.0 2 26 | 95 92 | 34 11 2 —|-— 1519... a) — | 22 | 90 | 104 | 66 | 21 9 2 1 | Number of Branched Anal Rays EX 9 1 | 11 12 1002: asc, ka | foo k 233 | 99 |. 3 Mad.. FÉ 51.187 | 10 | — 19199 ....... 3 534 | 991 $6 | — 193109... 025. 4 72-1 H8 1.12 1 Í 2 The last ray as usual was counted as double, i.e., as divided to the base. esa the posterior half of this divided ray is again divided well the base. In fact a complete transition can be traced between fins pest a given number of rays with those having one more ray. It is highly improbable, however, that this transition is suffieiently frequent as to permit a serious modification of the average number of rays, through a personal error in counting. No. 645] VERTEBRZ OF FISHES 367 9.53 (+ 0.06) for 1919, females, and 9.69 (+ 0.03) for both sexes of the 1919, class; in material collected in 1902 in the same lagoon the average is still higher, 9.83 (+ 0.02). The data on which these figures are based is given in Fre- quency Table I. In all three characters, namely the num- ber of vertebrz, of scales along the lateral line, and of branched anal rays, the year-class developed in the cooler season displays a significantly higher average. A highly similar yet exactly reverse condition is dis- played in the analysis of the counts on the Lepomis in- cisor material. In this ease the total number of vertebrae, and the number of caudal, but not precaudal, vertebre; the number of dorsal spines, dorsal soft-rays, anal soft- rays, and hence the total number of vertical fin-rays, all E e O Opec Emens Oy + eni 35 Se .5b 256 ST. ae 9 Total dorsa/ and ana/ rays Fie Comparison of number of dorsal and anal fin rays in successive year- classes ef Lepomis incisor. average higher in the class born in 1918 than in that of 1919. But we noted above that the temperature relations during the developmental periods of the two years were likewise reversed. In both Notropis atherinoides and 368 THE AMERICAN NATURALIST —. [Vor.LVI FREQUENCY TABLE II COMPARISON OF THE NUMBER OF VERTEBRZ IN THE 1918 AND 1919 BRoops or Lepomis incisor Year-class | Character Average oe Total Number of Vertebrze 28 | 29 | 30 INC el 95 | 9 29.10 0.02 2i. DR Nl Seen PUE >g 219 | 8 29.00 0.00 Number of Precaudal Vertebræ 11 | 12 | 13 gt UR ET aaa S 2 | 100 2 12.00 0.01 10106. 5s vs 2 | 230 2 12.00 0.00 Number of Caudal Vertebree 16 | 17 | 18 19106... 2002: 2 224 — 95 9 17.09 0.02 T0104... is ee Sud zi 219 8 17.00 0.01 4g0| 4 - s 7 E E E E 26V E E x 5 $ : a an ee i #0} og a Bo 3b HO | 6 Fic. 7. Illustrating seasonal variation in number of vertebrs in Notropis atherinoides. No. 645] VERTEBRJE OF FISHES 369 Lepomis incisor, therefore, a higher number of segments was developed in the year class developed at the lower temperature. The detailed data are given in Frequency "Tables II and IIT. Evidenee has already. been given indieating that the 1919 year-elass of Notropis atherimoides is divisible into three subclasses, of which the middle (B) developed dur- ing colder weather than either the first (A) or the last (C). The data given in Frequency Table IV and in figure 6 demonstrate that this subelass B possesses a decidedly higher number of vertebre than either of the other two. The averages are as follows: for the 146 specimens of sub- class A, 41.38 (+ 0.04); for the 845 comprising subclass B, 41.84 (+0.02); for the 97 individuals of subclass C, 41.42 (4- 0.05). FREQUENCY TABLE III COMPARISON OF THE NUMBER OF FIN-RAYS IN THE 1918 AND 1919 Broops or Lepomis incisor Year-class Character | Average Li sos | Number of Dorsal Spines | IX | x | XI XII | JEDE ss i 79 22 v 10.21 0.03 IM s 2 74 11 1 10.125 0.03 Number of Dorsal Soft-rays 10 11 12 | 13 1915... els 1 37 63 — 11.61 0.03 JM. uu sss 3 51 32 2 11.375 0.04 Number of Anal Soft-rays 9 | 10 | 11 12 I9 a. — 2 80 19 Tia 0.03 In. ak: 1 12 70 5 10.90 0.03 Total Rays in Dorsal and Anal Fins 33 34 35 | 36 | 37 | 38 | 39 1915b....... — 2] 20] 68 | 18 1 — 35.96 0.05 1981087... ls i 12 | 34 | 33 6 | — 1 35.40 0.07 370 THE AMERICAN NATURALIST [Vor. LVI FREQUENCY TABLE IV VARIATION IN NUMBER OF VERTEBR2Z WITHIN ONE-YEAR CLASS OF Notropis atherinoides - Number of Vertebree Sub- Size Average | Probable class Group Error ; 39 40 41 42 43 44 1919s C.... 27 — — 1 =- — -— (41.00) — 28 — -= 1 1 1 — (42.00) 0.32 29 — — 5 2 — — (41.50) 0.29 30 — — 9 4 — — 1.31 0.09 31 — 1 5 3 — — 41.22 0.16 32 — 2 7 4 — 1 41.36 0.175 33 — 1 4 T 1 — 41.62 0.145 34 — 1 7 10 — — 41.50 0.095 35 — 1 13 7 1 —— 41.36 0.09 1919) B.... 36 — — 11 13 4 — 41.75 0.09 37 — 1 16 1 6 — 41.71 * 0.07 38 — 5 10 36 5 — 41.73 0.05 39 1 4 24 39 8 1 41.68 0. 40 — 3 44 Ww — 41.82 0.0. Al — 2 24 46 16 2 41.91 0.055 42 — 1 21 67 18 3 42.01 0.04 43 — 3 24 41 1 41.87 0.06 44 — 1 18 43 14 1 41.95 0. 45 — 2 21 31 6 2 41.76 0.07 46 — — 13 23 10 — 41.93 0.07 47 — — 14 23 2 1 41.75 0. 48 — 1 16 16 7 — 41.72 0.08 19199 A.... 49 — 4 6 14 — — 41.42 0.10 50 — 3 14 |. 16 1 — 41.44 0.1 51 — 2 14 9 2 — 41.41 0.095 52 — 2 16 9 — — 41.26 0.075 53 ei 2 6 6 1 — 41.40 0.1 54 — — 7 2 — — 41.22 0. 55 — E 2 2 — — 41.20 0.225 56 — — |.1 2 — — (41.67) 0.18 57 — — — 1 — — (42.00) ee "d 58 Aude ae MISES dé d ene MEER HSH 59 — — — — — — — — 60 = — 1 — — — (41.00) — V It has generally been taken for granted, as a basic as- sumption, that such differenees as those here shown to hold between two successive year-classes, and between successive groups within a single year-class, are indie- ative of racial distinetion. Obviously this assumption ean not be maintained as wholly true. Moenkhaus (1895, 1898) indeed long ago demonstrated the occurrence of a significant annual variation within one race of fishes (in the ease of the darters Percina caprodes and Boleosoma No. 645] JA VERTEBRAE OF FISHES 311 nigrum). Sehmidt (1921) has lately studied such annuat fluetuations in great detail in Zoarces, and has induced like changes by experimental control of temperature in Lebistes (1919a, 19195) and Salmo (1921). I have ob- tained similar experimental results for coregonine fishes and for Esox lucius (data yet unpublished). On the other hand it has been clearly demonstrated in a number of cases that fine ‘‘racial’’ differences are in- herited. Thus Schmidt (1917a, 19175, 1918, 1920, 1921? has determined by his ‘‘offspring analyses" that a high degree of positive correlation holds between the number of segments and other features of the maternal parent and the unborn embryos of Zoarces. Similar results were obtained by Punnett (1904) for the viviparous shark, Etmopterus [Spinax] niger. In Salmo, Schmidt (1919c) has lately demonstrated that the finer differences in the number of vertebre of both parents are inherited, and in the viviparous teleost Lebistes reticulatus, the same author has found (1919a, 19195) that minor varia- tions in the parental number of dorsal fin-rays are in- herited. In somewhat similar fashion Summer (1918, ete.) has demonstrated that subspeeifie differences in color and size in the mouse genus Peromyscus are in- herited, even under changed enviromental conditions. A considerable body of indirect observational evidence might be brought forward, if needed, in confirmation of the assumption that these fine racial differences are in- herited. Clearly the same sort of variations as are induced by altered environmental conditions do characterize geneti- cally distinct local races of fishes. Furthermore, these two sets of correlations display certain striking simi- larities or analogies, the significance of which the writer is attempting to determine in the series of studies of which the one here reported is a part. LITERATURE CITED ice Carl L 1919. The Nesting Habits of -Certain Sunfishes as Observed in a Park Lagoon in Chicago. Aquatic Life, Vol. 4, pp. 143-144. Sia THE AMERICAN NATURALIST [Vor. LVI Moenkhaus, W. J. 1895. Variation of North American Fishes. n The Variation of Etheostoma d Rafinesque in Turkey Lake and Tippe- noe Lake. c. Indiana Acad. Kci., Yol. for 3805; pp. 278- 296. 1898. spectu for the Study of the Variation of Etheostoma ah Bex e and Etheostoma nigrum Rafinesque in Turkey e aas Lake. Ibid., Vol. for 1897, pp. 207-228. Punnett, R. C. 1904. Merism and Sex in Spinax niger. Biometrika, Vol. 3, pp. 313- 362 puer Johs. 1917a. Racial Investigations. I. Zoarces viviparus L. and local races of the same. Comptes-Rendus Trav. Lab. Carlsberg, Vol. 13, pp. 279—397. 1917b. Racial Investigations. II. Constancy Investigations Continued, Ibid., Vol. 14, No. 1, 19 pp. 1918. Racial Studies in Fishes. I. Statistical Investigations with Zoarces viviparus L. Jour. Gen., Vol. 7, pp. 105-11 1919a. Racial Investigations. III. Investigations with Lebüstos reticu latus (Peters) Regan. Comptes-Rendus Trav, Lab. codibiry, Vol. 14, No. 5, 8 pp. 19195. Racial Stu diii in Fishes. II. Experimental Investigations with Lebistes reticulatus (Peters) Regan. Jour. Gen., Vol. 8, pp. 147—153, 1919c. Racial Studies in Fishes. III. Diallel crossings with trout arogi trutta L). Jour. Gen., Vol 9, pp. 61-67. 1920. Racial vestigations. V. Experim rimental Investigations with Zour atk s L. Comptes-Rendus Trav. Lab. hatibele, ol, 14, No. 9, 14 pp. 1921. iii Tovesligationg, VII. Annual Fluctuations of Racial Characters in Zoarces viviparus L. Ibid., Vol. 14, No. 15, 4 a Smith, Kirstin 1921. uid Investigations on Inheritanee in Zoarces viviparus L. Ibid., Vol, 14, No. 11, 64 pp. Sumner, F. B. 1918. Continuous and Discontinuous Variations and their Inheritance in Peromyscus, IV. Heredity of the Racial Differences. Amer. Nar., Vol. 52, pp. 290-301. STUDIES ON FISH MIGRATION II. THE INFLU- ENCE OF SALINITY ON THE DISPERSAL OF FISHES* DR. F. E. CHIDESTER, West VinaGiNIA University, MonGaANTOWN, W. Va. In connection with an extensive study of the factors influencing fish migration, certain experiments were per- formed during the summers of 1919 and 1920 to deter- mine the effects of different salinities on the reactions of fish under laboratory conditions. Besides testing the animals with the salts of sea water, preliminary experi- ments were made with changed temperature and stream flow. MATERIAL AND METHODS The apparatus consisted of a two-tributary unit of a river system so arranged that different solutions could be introduced, affording the fish an opportunity to select the more favorable one. Two almost parallel troughs were so directed as to let the solutions flow down into a long receiving trough that had adjustable outlets in the middle. . There was also an intake at the extreme end of the large receiving trough so that if desired three intakes could be used. When only the two converging troughs were supplied with currents, a partition was placed across the middle of the receiving trough so that the water could flow laterally and eventually escape from the pool by the regular outlet. The two tributary troughs were each 10 feet long, 4 inches deep and 44 inches wide and the receiving trough was 10 feet long, 8 inches deep and 82 inches wide. The twin troughs were marked off in feet and conspicuously 1 Contribution from the Biological Laboratory of the U. S. Bureau of Fisheries at Woods Hole, Mass. 373 374 THE AMERICAN NATURALIST [Vou. LVI labeled at the proper points so that from a single ob- servation post, record could be taken of the distances traveled by fishes responding to the streams flowing down the incline. Streams were introduced after temporary storage in two barrels located above the ends of the experimental troughs. In some experiments the inflowing currents came directly from the circulation pipes of the laboratory. Experiments were performed with sea water, fresh water and combinations of the two, followed by tests with the individual salts of sea water in m/10 solutions. Tem- perature and stream flow were varied and proved most important adjuncts to the salts in affecting behavior. In order to be quite certain that habit formation as a factor was eliminated, it was customary to select a trough used during the night for sea water inflow and introduce a substance less attractive, for the first few experiments with a group. As conditions of illumination were uni- form and the troughs were so near each other, this proce- dure probably reduced the error due to a habit factor. The fish were males, selected for apparent vigor and averaged about 12 centimeters in length. They were used for a complete series of experiments in lots of ten, then replaced by another ten of similar size. In the majority of the experiments, the species used was Fundulus hete- roclitus. Its habits throughout the year were already known to the writer (1916, 1920). Loeb, Thomas and others had already studied its susceptibility to toxie sub- - stances. It is anadromous, highly resistant, yet furnishes quick reactions. | Fundulus majalis was used less frequently as it is not so resistant to laboratory conditions and behaves differ- ently with reference to tides. The observations of Mast (1915) made it especially desirable to study the reactions to currents and accordingly a series of eer Y was made. Clupea harengus dies quickly in ete Its re- sponses are extremely delicate and it has been used quite No. 645] FISH MIGRATION 915 successfully by Shelford, Powers and others in experi- ments on temperature, acidity, alkalinity and salinity. It proved too excitable for the experiments with which the present work was concerned. EXPERIMENTS Fresh Water and Sea Water. (Temperature 20? C.) With apertures # in. in diameter in two glass tubes directing horizontal streams of fresh water and sea water to a point six inches from the ends of the experimental troughs, it was found that 10 fish responded during 25 trials in such a manner that 11.8 was the value for re- sponses to fresh water and 44.6 was the value for the sea water. These figures were obtained by multiplying the number of fish responding by the feet traveled up the trough towards the current, adding the total of 25 trials and securing averages for control and experiment. The fish responded readily to the flow of water and since there was an admixture of fresh and salt water in the lower ends of the troughs, they did not at first dis- criminate the sea water before reaching a point 6 or 7 feet from the pool, that is 3 or 4 feet from the intake. As their reactions to the currents became established, however, they came in smaller numbers and finally be- came aligned along the sea water current at a distance of not more than a foot from the intake. On changing the flow of fresh water to salt and vice versa, it was noted that at first the fish came into the trough formerly salt, and proceeded beyond the point where they usually traveled in fresh water. This was in part due to the habitual response and partly to the pres- ence of some salts in the trough. On reaching the intake, they rapidly returned to the pool, one or two pioneered in each trough, then the whole group explored the salt trough and finally came to a point near the salt water intake. 376 THE AMERICAN NATURALIST [Vor. LVI Reactions to Salts in Solution A preliminary series of experiments was run with fish immersed in m/10 solutions of the salts of sea water, made up in fresh water. Results were obtained similar to those recorded by Loeb, Thomas and others with fish and corresponding ones known to the writer from experi- ments with the larve of mosquitoes (1916). By using the barrels above the experimental troughs solutions of the salts individually and in combination were introduced into the apparatus, with fresh water or sea water run as the control current. At first tempera- ture and stream pressure were kept constant. The tem- perature averaged 20.5? C. and the pressure was suffi- cient to send the eurrents horizontally to a distance of six inches from the 2-in. glass tubes. The reactions to individual salts as compared with fresh water are shown in the table below, only the aver- ages at the end of 25 trials with 10 fish being recorded. RESPONSES OF FISH TO SALTS MONO (dui as ee a ees 46 Control, 0 MUE seicesduecprssi MINE A E T 22.6 Control, 1 QUU cic ain os est E eevee vies 6 Control, 20 NUR I a ene career herp Ty 5.7 Control, 21.5 EA 8) GEERT eta E Opps FBLP Eis Va 2 Control, 15 It is quite evident that with temperature and stream pressure the same, Fundulus heteroclitus will react quite definitely to salts. It is attracted to the less toxic ones, MgSO,, and NaCl, and is repelled by those that are most toxic to it. | Similar experiments with sea-water solutions and sea water as the control eurrent brought out quite clearly that for the species used, m/10 solutions of the more toxie individual salts were not strong enough to repel the fish. For example in the case of the most toxic, KCl, the score for 25 trials with 10 fish was 43 for the control sea water and 34 for the experimental current with KCl in m/10 solution. Likewise, combinations of the salts showed only too No. 645] FISH MIGRATION 377 well the attractiveness of the mixed solutions. With an m/10 solution of MgCl, plus MgSO, and fresh water as control, the record was 11.2 for the control and 34.2 for the mixture. Again, in the case of KCl plus NaCl in m/10 solution, the score was 31 for the control and 17 for the mixture. With double sea water (specific gravity 1.050) and ordinary sea water at 20° C., it was found that the fish were attracted at the ordinary pressure and temperature, reacting to the stronger solution an average of 19.3 and to the control sea water 17.8 times. Further experiments should be run to determine the influence of antagonistic action of the salts in pairs. Whether or not the results will coincide with the results of permea- bility experiments will probably depend somewhat on the factor of temperature (Loeb and Wasteneys, 1912). Influences other than Salts The foregoing experiments indicate clearly that the behavior of the fish under consideration is materially af- fected by the salts with which they come in contact in fresh water. However, the factors involved in the migra- tion of fish are by no means thus explained. It is worthy of note that the reactions of Fundulus heteroclitus to toxic salts or even sewage are dependent on Papan and stream pressure. Temperature Numerous experiments were tried with varying tem- perature and it was found that a temperature greater than 23? C. repelled the fish and eaused them to align themselves along the current of fresh water at 20? C. in preference to the slightly warmer sea water. With a reduced temperature, even one degree less than the control (19° C.), the fish were markedly attracted. In fact it was possible to lure them into double sea water, KCl or fresh water if these were presented at the proper temperature. Further experiments and observations are necessary for these and other species in order to deter- mine the relation between gonad development, bodily condition and the responses to temperature change. 378 THE AMERICAN NATURALIST [Vor. LVI As pointed out by Gurley (1902), the minnows migrate to warming water for the purpose of spawning, while the cod and the salmon migrate to cooling water for the same purpose. Chamberlain believes that the salmon come into water warmer than the sea water (1906). Field records for Fundulus heteroclitus secured by the writer in eonneetion with another investigation (1916) indicate the importance of temperature. The fish began coming inland in the spring when the water was about 15° C. and continued to run in and out until the inland pools had reached a temperature in August of about 24° C. Then for a period of over two weeks, they ceased running. About September 1, when the temperature had again lowered, they appeared again and continued to run until the temperature ran down to 10° C. Stream Pressure When sea water was introduced through the 2-in. glass tube with a force sending it horizontally to a distance of 6 inches, while fresh water was introduced through the experimental tube into the adjoining trough with a force sending it 12 inches from the end of the tube, there was no difficulty in luring the fish away into the fresh water and keeping them directed towards it. Many experiments were made, toxic substances such as KCl and double sea water also being introduced, but the increased pressure always proved the powerful factor. Chamberlain (1906), Prince (1920) and others have pre- viously shown that in the case of the salmon, migration into fresh water is delayed until the floods come down into the bays and small streams. The arrival of a vol- ume of rushing water furnishes the needed stimulus and the fish proceed forthwith to obey their instinct to swim against the current. That fish can determine the presence of toxie sub- stances in sea water or in fresh water is unquestionably demonstrable. But we have much evidence that those fish lying offshore and habitually migrating up a certain No. 645] FISH MIGRATION 379 stream, will journey into polluted water, spawn in places where the eggs can not develop and in many cases, die in such water themselves. Salmon are reputed to return to the lake-fed streams where they were spawned and there is considerable evi- dence that they are guided by temperature difference, probably also by the current pressure, number of water- falls, oxygen content and even by food. There is no ques- tion (Meek, 1916), however, that salmon ascend streams where no salmon could hitherto have spawn The destruction of protecting forests, spoliation of natural waterways and the utilization of streams by manufacturers wishing to dispose of wastes are the fac- tors which not only cause the death of fish embryos and adults, but prevent the natural control of insect pests by their destruction in the larval state. SUMMARY 1. Fundulus heteroclitus is able to discriminate toxic from non-toxie salts at a temperature and stream flow the same as the control. 2. Variations in temperature or in stream flow pro- foundly influence the reactions and are more powerful factors in the behavior of the fish than presence or ab- sence of salinity. 3. In the apparatus used, errors due to the notable re- actions of fish to currents of water have been reduced by presenting the control and experimental flows parallel to each other. BIBLIOGRAPHY Calderwood, W. L. 1907. K Life of the Salmon. London, 1907. Chamberlain, F. M. 1907. Some Observations on Salmon and Trout in Alaska. Report and special papers for 1908. U. S. Bureau of Fisheries, Document Chidester, F. 1916. A Biological aoed of the More Important of the Fish Enemies rsh Mosquitoes. Bull. N. J. Ag. Exp. Sta. No. 300, pp. 134 380 THE AMERICAN NATURALIST [Vor. LVI Chidester, F. E. 920. The Behavior of Fundulus heteroctitus on the "us Marshes of New Jersey. Am. Nat. Vol. 54, pp. 551-55 Chidester, F. E. 1921. A Simple a EE E Studying the Factors Influencing Fish oe Pro c. for Exp. Biol. and Med., Vol, 18, pp- 7. wit A 1902. Tue Habits of Fishes. Am, Jour. Psy., Vol. 13, pp. 408-425. — di uo Wasten On the vice of Fish (Fundulus) to Higher DEN tures. Jour. Exp. Zool. Vol. 12, No. 4, pp. 543—557 Loeb, J. 1912. erp dues Aetion of Eleetrolytes and Permeability of the Cell Membrane. Science, N. S., Vol, 36, No. 932, pp. 637- 639. Lyon, E. P. 1904. On Rheotropism. I. Rheotropism in Fishes. Am. Jour. Phys- 4 9. Lyon, E. 1909. p ithaca II. Rheotropism of Fish Blind in One Eye. m. Jour. Physiol., Vol, 24, pp. 244—251. Mast, S. O. 1915. The Behavior of Fundulus with Especial Reference to Over- land Escape from Tide Pools and Locomotion on Land. Jour. An. Beh., Vol. 5, pp. 341-350. MeDonald, M. 1885. Report on the Pollution of the Potomae River by the Dis- charge of Waste Produets from ied Manufacture, Bull. 1884, Vol. 5, U. S. Bureau of Fisher Meek, A. 1916. The Migrations of Fish. London, 1916, Prince, E. E. 1920. Why Do Salmon Ascend from the Sea? Trans. Am. Fish. Soc., Vol. 49, pp. 125-140. Ringer, S. 1883. Concerning the Influenee of Saline Media on Fish. Jour. ysiol., ba 5, p. Shelford, V. E. and Powers, E. B. 1915. An Paia Study of the Movements of Herring and Other Marine Fishes. Biol. Bull., Vol. 28, No. 5, pp. 315- 334. Thomas, A. 1915. ciate of Certain Metallie Salts upon Fishes. Trans. Am. Fish. Soe., March, 1915, pp. 120-124, SHORTER ARTICLES AND DISCUSSION NOTE ON ASSORTATIVE MATING IN MAN WITH RESPECT TO HEAD SIZE AND HEAD FORM ASSORTATIVE mating in man has been much discussed! but has been little investigated by scientific methods. For eharaeters sueh as age of husband and age of wife where there is an obvious preferential mating we may have coef- ficients of assortative mating as large as r= + .15. For stat- ure, span and forearm Pearson has determined eoeffieients of about + .20 for span and span, + .20 for forearm and forearm, and + .28 for stature and stature in husbands and wives in his English series. The cross correlations for these various charac- ters are in general smaller. For bodily characters other than stature the data are very few and are in general unsatisfaetory. With characteristic caution Pearson long ago suggested? that eoeffieients of assortative mating might be due to the husbands and wives being drawn from the same local races. The im- portance of this factor seems to be very small in his own ma- terials. This question must eontinually reeur whenever assortative mating for physieal eharaeters is diseussed. It seems very de- sirable, therefore, to obtain some measure of the correlation be- tween husband and wife with respect to cephalic index, a char- aeter whieh has been eonsidered of great importanee by an- thropologists in differentiating the raees of Europe. For head size and head form we have had, as far as we are aware, until reeently only the data for forty-eight families of Eastern Euro- pean (Russian) Jews living in New York City, for whieh Boas* found assortative matings for cephalic index measured by r—.15 + .10. Recently Frets in a series of papers‘ has given data for head 1 The literature of the field has been reviewed up to 1912 by one of us: Harris, J. Arthur, ‘‘ Assortative i of Man,’’ Popular Science Monthly, 80: 476-492, 19 5 Peation, E, “Data for the Problem of Evolution in Man. III. the Magnitude of Certain Coefficients of Correlation of Man,’’ ete., Proc. Roy. Soc., Vol. 66: 23-32, ; 3 Boas, F., ‘‘Heredity ir Head Form,’’ Amer. Anthrop. N. S., 5: 532. 1903, 4 Frets, G. P., ‘‘Heredity of Head Form in Man,’’ Genetica, 3: 193-400. 1921. This paper contains the original measurements. These have been to some extent checked against his other papers. 381 382 THE AMERICAN NATURALIST [Vor. LVI length, head breadth and cephalic index in a series of Dutch families. He has himself calculated a coefficient of correlation of .039 + .034 for the cephalic index of husband and wife in 389 families? We have felt it desirable to determine the cor- relation for length and width of head, as well as that for index. Because of a suggestion by Pearson (loc. cit.) that the cor- relation apparently indieating assortative mating may be really due in some eases to an association of fertility with homogamy, we have thought it desirable to calculate all the coefficients of correlation in two ways: (1) by using the actual number of parents, and (2) by weighting the parents with the number of offspring indicated in Frets’ tables.* The correlation coefficients are as follows: Length of husband’s head and length of wife’s head: Parents only, r — 4- .0487 + .0377. r/Er — 1.29. Parents weighted with their children, r=+ .0616 + .0376. r/r == 1.63.. Breadth of husband’s head and wife’s head Parents only, r =:+ .1197 + .0372. et 22. Parents weighted with their MESE = + .1184 + .0372. r/Er=3.18. Index of husband’s head and index of ite? s head: Parents only, r==-+ .0231 + .0377. r/Er=0.61. Parents weighted with their US — .0546 + .0376. r/Er = 1.44. The constants are with one siens positive in sign. That for the breadth of husband and breadth of wife may perhaps be considered statistically significant in comparison with its probable error. The others, particularly that for the cephalic index, can not be so considered The coefficients may, therefore, indicate a slight assortative mating for the dimensions of the head. The coefficients, in common with those for physical characters other than stature, are relatively low. That the correlation for the cephalic index is so low is a point of particular interest. If cephalic index be a character of great importance in distinguishing races, and if correlations which have been demonstrated between the physi- 5 Frets, G. P., ‘‘Erfelijkheid, correlatie en regressie,’’ Genetica, 3: 1-27. 1921. 9 We are able to abstract from Frets’ tables 319 pairs of parents in whieh there were no indieations of typographieal errors when different tables were checked against each other. These had a total of 1328 recorded children. In ealeulating the probable errors of the coefficients we have used the unweighted number of parents as N No. 645] SHORTER ARTICLES AND DISCUSSION 383 eal characteristics of husband and wife be due primarily to the tendency to marry within the same racial group, one might ex- pect a large correlation for cephalic index. Instead we find the lowest correlation of the three determined. J. ARTHUR HARRIS, ALBERT GOVAERTS STATION FOR EXPERIMENTAL EVOLUTION, COLD SPRING HARBOR, L, A GYNANDROMORPH IN DROSOPHILA MELANOGASTER + In 1916 Hyde and Powell described a mosaic female with one eye eosin and the other blood. They interpreted this case in the light of Morgan’s suggestion of 1914 that ‘‘Gynandro- morphs and mosaics may arise through a mitotie dislocation of the sex ehromosomes." In other words they believed one X chromosome earrying the gene for eosin went into the cells of one eye and the other X chromosome earrying the gene for blood went into the other eye. In 1919 Morgan and Bri deseribed a large number of gynandromorphs. The hypothesis of chromosomal elimination explains most of them, but a num- ber of speeial eases are explained in other ways. One of their special cases was a male with one eye eosin and the other eosin vermilion. They explained this ease by assuming that the egg had two nuclei, one of which after maturation had an eosin vermilion X ehromosome and the other an eosin X ehromosome. Further, they assumed each nucleus to have been fertilized by a Y sperm. These hypotheses would explain the facts that the individual was male throughout and that one eye was eosin vermilion and the other eosin. In our experiments a somewhat similar mosaie appeared. The individual was made throughout, with one eye garnet and one white. - The parentage was as follows: a garnet male was mated to a yellow white female. An F, wild-type daughter was mated to an F, yellow whife male. From this pair of parents the mosaic arose. It was fertile and was bred to a garnet female. In F, all males arid females were garnet. The F, garnet males and females were inbred. In F, the females were garnet but the males were garnet and white in approxi- mately equal numbers (1,089 garnet to 1,026 white). This demonstrates very clearly that the mosaie was genetieally a 1Zoologieal Laboratory Contribution No. 191. Indiana University. 384 THE AMERICAN NATURALIST [Vor. LVI garnet white. Professor Morgan writes us that he would also interpret this ease on the binucleated egg hypothesis. We see clearly how the hypothesis may be applied and that the binu- eleated eggs described by Doncaster may give indirect evidence in its favor. Perhaps it is the best interpretation. We wish to point out, however, that there are other possibilities although they may have no direet or indireet morphologieal or experi- mental evidence in their favor. Let us assume the individual started as a rud male, the single X chromosome carrying the genes for garnet and white. Since the mosaic did not carry the gene for yellow, the garnet white genes must have been brought together by a double cross- ing over in the mother. The only assumption we need to make is that during somatogenesis, the white end of one of the daughter X chromosomes became in some way inactive or lost. This would leave in one cell a whole X chromosome carrying white and garnet; in the other an imperfect X chromosome carrying garnet only. We know by test that white and garnet in the same chromosome give an eye practically indistinguishable from white. If one eye arose from the descendants of one of these two cells and the other eye from the second cell, we could ae- count for the difference in color. The only assumption we need to make then is the loss or inaetivation of the white gene in one of the early cleavage cells. On the binucleated egg hypoth- esis we must assume, first, the presence of two nuclei within the egg; secondly, that each nucleus is fertilized by a Y sperm; and thirdly, that the sex cells of the male arose from the de- seendants of only one of these nuclei, as all sperm were alike, earrying garnet and white. A second possibility is that of somatic mutation. If the white gene in one of the cells should mutate to red, we would have a cell whose X ehromosome earried the gene for garnet. If the deseendants of this eell gave rise to one eye and the descendants of the other cells to the seeond, we would have one eye garnet and one garnet white, whieh is white. It is true that white eye has never reverted to red in all the thousands which have been bred. This fact renders this Suggestion improbable but not im- possible. F. PAYNE, MARTHA DENNY* ZOOLOGICAL LABORATORY, NA UNIVERSITY THE AMERICAN NATURALIST Vor. LVI. September— October, 1922 No. 646 EXPERIMENTAL STUDIES ON THE DURATION F LIFE V. Ow THE INFLUENCE or CERTAIN ENVIRONMENTAL Factors on Duration or Lire IN DROSOPHILA !' PROFESSOR RAYMOND PEARL AND SYLVIA L. PARKER A. INFLUENCE OF VENTILATION ON DURATION OF LIFE Tue standard method of handling Drosophila cultures, as described by Pearl and Parker (27), includes the plugging of the mouth of the bottle with absorbent cotton to prevent the escape of the flies. The theory of this practice, which is the custom in Morgan’s labora- tory, presumably is that air will pass in and out through the plug while the flies can not. No physicist or ventila- tion engineer would, we believe, accept this theory. Many years ago the senior author had occasion to make some observations on the ventilation of curtain-front poultry houses, and soon came to the conclusion that curtains of one thickness only, of the very porous jute bagging which is used for bran sacks, are practically nearly as effective in preventing the natural unforced circulation of air as a half-inch pine board would be. We may be sure that the plug of cotton used in Droso- phila bottles will be an even more certain preventative of the natural unforced circulation of air. Theoretically one may perhaps hold that there is more circulation of air with a cotton plug than there would be with a cork stopper, but the difference must be infinitesimal. 1 Papers from the Department of Biometry and Vital Statisties, School of Hygiene and Public Health, J ohns Hopkins University, No, 67. 385 386 THE AMERICAN NATURALIST [Vor. LVI In the systematie survey which we are making, at the beginning of our experimental study of duration of life, it seemed desirable to test the influence upon this char- | acter of degree of natural ventilation of the bottles. It is the purpose of this first section of this paper to pre- sent the results of some experiments on this point. Material and Methods The experiments were carried out in two series. For the first, wild type flies of our Old Falmouth stock (Pearl and Parker (27)) belonging to Line 107, the dura- tion of life constants of which have been given by Pearl and Parker (32), were used. These flies were of the 25th pedigreed generation. Eighteen mass matings of the flies of this line were started for the present experiments on March 13, 1922, and the flies to be used emerged March 23-30, 1922. For the second series short-lived flies of Quintuple stock (Pearl and Parker (27)) were used, in the 27th pedigreed generation. Twenty-five mass matings of Quintuple line 405 were started April 10, 1922, and six mass matings of mixed Quintuple stock were started April 11, 1922. The flies for use in the experiment emerged April 22-27, 1922. The procedure in making up the experiments was as follows: The flies were counted out each morning upon emergence into our standard one ounce screw top shell vials used in the determination of duration of life (Pearl and Parker (27)). Fifty flies were put in each bottle. The wild type flies were counted in through the counting tube described in these Studies, III (Pearl and Parker (44)). The Quintuple flies move through the tube so slowly, however, that there is time for moisture to con- dense and accumulate on the walls of the tube, killing some of the flies by drowning, and injuring others. Con- sequently the flies of this type were etherized and counted into the bottles. It has been shown in these Studies, IIT, that such etherization has no measurable No. 646] THE DURATION OF LIFE 387 influenee on duration of life. Each day's flies were di- vided equally between control and experimental groups. Fertility in the Quintuple flies is so low that even with the large number of matings the hatehes on some days did not equal 100 flies. It thus resulted that there were a few bottles of the Quintuples with fewer than 50 flies to the bottle: 1 ease with 40 flies in control and 40 in the experimental bottle, 1 with 37 flies in each bottle, and 1 with 23 flies in each bottle. The control bottles were plugged with cotton in the ordinary way. The experimental, ventilated bottles were covered with one layer of silk bolting cloth of No. 48 mesh (48 meshes to the linear inch), this being the largest mesh which could be used without any possibility of a fly squeezing through the openings. The cloth was held firmly and evenly in plaee by an aluminum serew cap in the top of which a central hole a little more than 94 inch in diameter had been punched out. This is practically the internal diameter of the shell vials which we use. Both control and ventilated bottles were carried in 25? incubators, and all other procedure was that which has been described by the authors (27) as standard in the work of this laboratory on duration of life in Drosophila. Results The observed l- distributions (survivors out of 1,000 starting together) for the wild type Old Falmouth Line 107 are given in Table I, together with the absolute num- bers of flies on which the distributions are based. These distributions of Table I are shown graphically in Fig. 1. It is evident at once that the flies in the well-ventilated bottles outlive those in the ill-ventilated. Their expecta- tion of life is greater at every age. The magnitude and significance of this difference can best be appreciated from the constants of duration of life set forth in Table IL 388 THE AMERICAN NATURALIST [Vor. LVI TABLE I SURVIVORSHIP DISTRIBUTIONS (Iz) OF VENTILATED AND CONTROL FLIES Old Falmouth, Line 107 Number of survivors up to indicated age. Cc l Age in days Ventila ontro Be pes eee Pe he T C Cre 1,000 1,000 T CaL HIM AE D d siu nals s 986 984 AM ee Ces Cee ee ek ie qa 970 969 34 Oed as P eFra tau rcs uae 946 929 BB DP IIO EELVPDÜLE Cae EE CR 900 846 Blol.claideneWaesr5 cia Apis 804 730 CYESPOQs rast SEA MADE MEM C ED 697 615 di ehh es eked ers RÀ er rn 589 485 AM ioo LEUR aL soe ds d RD 480 397 DD Liao poe etc RR Sie AD CA 371 292 OL. Clu ULCUS S d E WEE TU. S Ee 278 195 OF ios wees ri Nod ee LEG 185 119 Y opu MUCH RU UE unge nca ie wes 103 44 vi ANEA EE se cee hs 12 2 BD E T ee ae ee T 0 0 Absolute number of flies ......... 946 931 TABLE II FREQUENCY CONSTANTS OF dz DISTRIBUTIONS Wild Type, Line 107 | Standard Coefficient of i Poup | Meah Deviation Variation Con as ere te s | 43.66 + .39 17.63 + .28 40.38 + .73 Ventilated. 1.22 20 9.4 | 47.92 + .40 18.22 + .28 38.01 + .67 Differenoé. ... 2x a is | + 4.26 + .56 + .59 + .39 — 2.37 + .99 There is clearly a significant increase, amounting roughly to 10 per cent., in the mean duration of life, or expectation of life at emergence, in the flies in the venti- lated as compared with the unventilated bottles, all other conditions both genetic and environmental having been the same in the two series. That the increased amourt of fresh air is the cause of the difference is evidenced by the behavior of the flies. In the ventilated bottles the flies tended at all times to congregate on the under side of the bolting cloth going down to the bottom for food occasionally, but otherwise exhibiting a strong preference for the region about the No. 646] THE DURATION OF LIFE 389 mouth of the bottle, where the diffusion of air between inside and outside was going on most rapidly. This be- havior is in no way characteristic, in our experience, of flies in the bottles stoppered with cotton. In those there 4000 S SURVIVORS S `~ ES jeep E NUN gy a NO ee DA Fic. 1. Survivorship (Z,) lines for ventilated (solid line) and control (broken line) flies, is generally a fairly even distribution of flies throughout the bottle, with such tendency towards concentration as there is, in the direction of the bottom near the food rather than the top. In Table III are presented the survivorship distribu- tions for the Quintuple flies. Because of their much shorter life span, as shown in the life tables of the first 390 THE AMERICAN NATURALIST [Vor. LVI one of the Studies (27), a shorter abscissal interval has been used in the grouping. The figures for Quintuple stock flies, and for an inbred Quintuple line (No. 405) - are given separately. TABLE III SURVIVORSHIP DISTRIBUTIONS (Iz) OF VENTILATED AND CONTROL FLIES uintuple Stock and Line 405 Number of Survivors up to Indicated Age in Age in Days Stock Stock Line 405 Line 405 Ventilated Control Ventilated Control To eee AUTE. 1,000 1,000 1,000 1,000 Seed orn Cue AV 993 992 783 878 f Tenet ange oats aia 816 812 422 443 JO ace ens redi 218 368 62 IS Qv EE ERDSVONDES 81 158 18 36 18: 17 Oe RR 44 143 9 14 40 o ee DTE 29 113 9 5 Pee STR EA eis Pas 22 83 4 0 Di E ERE T ey a re week 22 60 4 — DELI RA UE CR 22 30 0 — SEEN Gi ec eno 22 23 — — OR, iu E EE 15 23 — — OF cm Ed 0 15 — = FT RGR aaa uU ec S E — 15 — — di. oS E S E E ek — 8 — — AB uL odo s RE -— 8 — — a I LIT — 8 -— — DESI uta ERA aA mo 0 — — Absolute number of files 136 133 209 221 The calculated constants from the de distributions are given in Table IV. TABLE IV FREQUENCY CONSTANTS OF dz DISTRIBUTIONS Quintuple Stock and Quintuple Line 405 Standard Coefficient of Sort Group Mean Deviation Variation Quintuple stock. .| Control....... 11.07 +.41 7.04 2-.29 63.58 +3.54 Ventilated ..... 9.34 +.26 4.57 +.19 48.92 +2.43 Difference EEE DARET A EET TE — 1734.49 | —2.47 +.35 | —14.66 +4.29 | eee GS |Control.......| 6.90+.18 | 200400! —42.09::1.57 Ventilated. 5... " 784.14 | 2.99210 | 44064171 Be RE LIAE ARS TA — 12+.19 | + 094.13 | + 1.97 +2.32 No. 646] THE DURATION OF LIFE 391 The situation here is evidently quite different from what obtained with the wild type flies. The Quintuples lived somewhat longer in the control bottles than in the ventilated. In the ease of flies from stock, the difference in the means amounts to 1.73 days, and is 3.5 times its probable error. The numbers are, however, small, and as an examination of Table III shows, the long survival of 2 individuals in the control series after age 34 ac- counts for a considerable part of the difference in the means. With a larger experimental sample much of the difference in the means would, we feel sure, disappear. The influence of these same two individual flies is clearly seen in the greatly inereased variability of the control series over the ventilated in the stock groups. In general we are of the opinion that in the case of Quintuple flies the difference in ventilation represented by a bolting cloth screen versus a cotton stopper has no significant influence upon duration of life. The results with extremely short-lived line 405, we regard as typical of what one should expect with Quintuple flies in this sort of an experiment. The reason for the difference between wild type and Quintuple flies in their response to ventilation is founded, in our opinion, upon the normal differences in behavior between the two types. In Quintuples the wings do not function (the wing mutation in this stock is Vestigial). The consequence is that these flies are much less active, and generally appear to live on a lower metabolic plane, than wild type flies. Their oxygen needs are presumably smaller, and it would therefore be reasonable to expect that they would not show the difference in duration of life with increased ventilation that the wild type flies do. In this connection, it should be noted that their actual behavior in this experiment was in accordance with the view here suggested. They showed no such definite tend- ency to congregate at the top of the bottle under the bolting cloth as the wild type flies did. Their distribu- tion was about the same in ventilated as in the control 392 THE AMERICAN NATURALIST [Vor. LVI bottles. Another consideration is that genetically Quin- tuple carries factors for very short life. These genes appear, in our experience with these flies, to be the over- whelmingly important factors in determining their length of life. No environmental factor, however favor- able, makes much difference in their duration of life. Summary In experiments involving the determination of the duration of life in 2,576 individual flies, it has been found that in the case of Drosophila of wild type (i.e., carrying no mutations so far as known), an increase of roughly 10 per cent. in the mean duration of life is brought about by inereasing the ventilation of the cul- ture bottles, by covering the mouth with one layer of No. 48 mesh bolting cloth, as compared with the use of cotton plug stoppers as is the usual practice in the cul- ture of Drosophila in the laboratory. Owing, in our opinion, to fundamental differences in behavior, no such differenee appears in the ease of Quintuple flies. B. Can THE Duration or LIFE BE INCREASED BY EMBRYONIC JUICE? If the theory of senescence and natural death which the senior author has developed in his ‘‘Biology of Death’’ (1-7) is true, one consequence of it should be that it might be possible to increase the duration of life, if by appropriate means one could restore the normal func- tional balance of the parts of the body after changes had set in with advancing age. Might it not be possible, by the use of X-rays for example, at the right stage of the life curve, and in proper dosage, to destroy cells, or per- haps even parts of tissues, which have got out of proper functional balance, and thus pave the way to their re- placement by regeneration with fresh, ‘‘young”’ cells or tissues? In this way the life of the whole organism might be prolonged. The work of Frisch and Starlinger (45) with blood suggests that such a result might at least No. 646] THE DURATION OF LIFE 393 be hoped for. In view of the known facts as to the po- tential immortality of tissue cells in cultures in vitro, and the apparent reason for the difference in the behavior of the same cells in respect of duration of life when they are in the multicellular body, all of which has been rather fully discussed by Pearl in the ‘‘ Biology of Death’’ (loc. cit.), it would seem that this is a line of experimenta- tion well worth following. We have a number of experi- ments along this line now in progress, particularly with X-rays, which we expect later to report upon. Some of the purely preliminary work has already been finished, and we wish in this paper to report one piece of it. The brilliant researches of Carrel and his coworker Ebeling (cf. 46, 47) on the duration of life of cells in eultures in vitro have brought to light the extraordi- narily interesting and presumably important faet that for the continued life of such cultures it is apparently essential to have in the culture medium a small amount of embryonic juice. In just what manner this functions is not yet clear, but the necessity of its presence seems well established. It occurred to us in our preliminary work on prolonga- tion of life in Drosophila, or as it is perhaps better to put it, on changing the form of the l- line of the Droso- phila life table, to see whether embryonie juice, applied at a point on the ls curve after senescence had definitely set in, would have any effect upon the subsequent course of the curve, or in other words, upon the duration of life of the organism as a whole, comparable to its effect upon the life of cells in culture. The ideal way, of course, in such an experiment would be to get the embryonic juice to the tissues of the fly by a par-enteral route, but as no practical method of doing this occurred to us, we decided to feed it, and see if any results followed. Material and Methods The flies used in the experiment were all wild type, of Old Falmouth stock, and belonged to Line 107, pedigree 394 THE AMERICAN NATURALIST [Vor. LVI bred for 21 generations.. The individuals for the experi- ment came from 20 mass matings of 3 pairs of parents each from this line. The bottles were started December 16, 1921, and the flies used in the experiment emerged December 28, 1921, to January 9, 1922. The flies were eounted through the counting tube into our standard shell-vials in groups of 50 each. Each day's bottles were divided into three groups at the be- ginning of the experiment, but all had the same regular treatment until the flies in them were 30 days old. This is a point where the le line is beginning distinctly to turn downward. From that time on until the end of their life one series of flies was given chicken juice in their food, and one series the juiee and pulp of erushed Drosophila larve. The chick embryos used were 14 days old. The juiee was extraeted in a beef-juice extraetor, and added to the regular food at the rate of approximately 2 c.c. to 100 c.c. of food. With the Drosophila larve, the whole pulp was used, and that too was added to the regular food at the rate of approximately 2 c.c. to each 100 e.c. of food. All the flies, experimental and control, were transferred every day to fresh food, made up that day, except on Sundays. On Feb. 8 an accident happened to the incubator at the source of our chicken supply, so that for 12 days no chickens were obtainable. In all particulars except those specified above, the procedure in these experiments was the standard tech- nique of this laboratory in duration of life work de- seribed in (27). Results The survivorship distributions are given in Table V, on the basis, A, of 1,000 starting at emergence, and, B, of 1,000 starting at age 31 days, that is at the time when the experimental feeding began. 2 We are greatly indebted to our colleagues, Dr. and Mrs. Warren H. Lewis, for furnishing us with chicken material for this work. No. 646] THE DURATION OF LIFE TABLE V SURVIVORSHIP DISTRIBUTIONS (lz) OF GROUPS OF DROSOPHILA FED In D Old Falmouth Stock, Line 107 395 Number of Survivors up to Indicated Age in Age in Days Controls Chicken Juice Larval Pulp A B A B A B Lo 0 eae 1,000 — 1,000 prm ,000 pes Nu Ili I 975 — 970 — 970 — tO epee OU 945 — 935 -— 930 — 18. 7 AA 884 — 873 — 858 — ZR AR IYV RAV 797 -— 784 — 719 — BEC. SI A An 616 1,000 648 1,000 591 1,000 ol QUSE na CE 491 796 456 703 464 785 4i) 1 1s 362 588 313 484 283 479 842 0n VL 286 465 246 380 210 356 DU CDU aaa 212 345 198 306 156 264 BL ere 156 253 152 234 116 196 eT S nus 120 197 110 170 82 140 Facti y 59 96 58 89 46 78 TE ull 21 34 24 38 28 48 Sh ovi ud ví ES 9 14 3 5 Bi. lloc vut 1 2 0 0 1 2 Wi Sd 1 2 — — 0 0 108 iis os 0 0 — — mx e pepe number EET 1,013 — 983 -— 994 —- The biometric constants of duration of life calculated from the dz distributions are given in Table VI. TABLE VI BIOMETRIC bere FOR DURATION OF LIFE IN DROSOPHILA ire DIFFER- T CONDITIONS OF Foop. A. From EMERGEN B. From Acs 31 Days on G Stand. Coefficient Class Group Deviation Gn days) | (in days) | Variation Pes ees T E N i siio ccs 9.60 + 18.76 +.28 | 47.384 .85 Chiekon Jalo: ana: 38.66 +.40 | 18.61 +.28 | 48.12+ .89 Va E Eoi Y 36.74 +.39 18.01 2.27 03+ Bv. Ph ci os II .72 2-.40 14.66 +.28 aq 35 +1. ux Chicken juice. ....... i. 512-.40 | 15.05+.27 96 +2. a hare or E. 17.12+.39 14.03 +.28 * 982-1. e 396 THE AMERICAN NATURALIST [Vor. LVI It is evident that there are no large differences in mean duration of life between any of the groups. The l- distributions and the constants are closely similar throughout. This is true whether the whole life is taken, or the expectation after age 31. "The exact nature of the differences is shown in Table VII. TABLE VII DIFFERENCES IN MEANS OF TABLE VI Class Difference Taken Value of | Diff./P.E. Diff Difference or mue X uc. Control — chicken juice. ............ .94 27.56 | 1.66 Control —larval pulp. .............. 2.86 +.55 5.15 Chicken juice —larval pulp. ......... 1.92 +.56 3.46 Bill Control —chicken juice............. 2.21 +.56 3.93 Control —larval pulp. .............. 2.60 +.56 4.67 Chicken juice —larval pulp. ......... .39+.56 .69 The control group had slightly the greatest duration of life, both as a whole, and from the time of the begin- ning of the special feeding on. The flies fed larval pulp had the worst expectation of life, with those fed chicken juiee in an intermediate position. None of the differ- ences, however, is large. That some of them are signifi- cant statistieally probably means no more than that the changed food is not quite so favorable for the flies as the normal, standard food. The numbers involved are large relatively, and the probable errors eonsequently small. We must then conclude that the administration of em- bryonic juice in the manner, amount and time in the life cycle, which defined its administration in these experi- ments, does not bring about any prolongation of the life of the whole organism, comparable to its effect in tissue cultures in vitro. This does not necessarily mean that under other eonditions of administration or dosage an effect in this sense might not be produced. We believe, however, that it is not probable that any prolongation of life can be brought about by this method, for the rea- son that in the first place the results of the present ex- No. 646] THE DURATION OF LIFE 397 periment give no suggestion that with larger dosage any such result would appear, and in the second place, be- eause the experiments of Baeot and Harden (48) indi- eate that as slight (or slighter) alterations of the food of Drosophila as those of the present experiments may produce marked effects in respect of viability.? For some reason which we are unable to explain, the -flies of Line 107 had, in all the series of this experiment, a lower mean duration of life than this line has ever shown before (cf. 32, 44, and section A of the present paper). The values are extremely even and consistent in this feeding experiment, but are about 10 days lower than what previous work has indieated as the normal duration of life in this line. "There has been no other change in the line, in fertility or other characters. We are inclined to believe that the low values in the present experiments represent merely a temporary secular change (? seasonal) in the duration of life characteristic of the line. Summary In experiments involving the determination of the duration of life in 2,990 individual flies, it was found that there was no prolongation of the life of Drosophila pro- duced by adding embryonic juice (either from the chick, or from the larve of Drosophila itself) daily to the food, to the amount of 2 per cent. of the total food material, beginning with the 31st day of the flies’ life. LITERATURE CITED (The plan of numbering citations is explained in the second of these Studies, AMER. NAT., Vol. 56, p. 174.) 44. Pearl, R. and Parker, S. L. Experimental Studies on the Duration of Life, III. The Effect of Successive Etherizations on the Duration of Life of Drosophila. AMER. NaT., Vol. 56, pp. 273-280, 1922. 45. Frisch, A. and Starlinger, W. Zur Frage der Protoplasma-aetivierung. Zeitschr. f. d. ges. ezp. Med., Bd. 24, pp. 142-158, 1921. 3 It ought, however, to be pointed out that the experiments of Bacot and Harden are extremely faulty from a technical standpoint. They evidently know little of the practical husbandry of Drosophila. Their cultures were incubated at 30? C. At this temperature one does not get anything remo otely resembling normal physiological processes or duration of life, except after many months of acclimatization. 308 THE AMERICAN NATURALIST [Vor. LVI 46. Carrel, A. and Ebeling, A. H. The Multiplication of Fibroblasts in vitro. Jour. Exp. Med., Vol. 34, pp. 317-337, 1921. 47. Id. Heterogenie Serum, Age, and Multiplication of Fibroblasts. Ibid., Vol. 35, pp. 17-38, 1922. 48. Baeot, A. W. and Harden, A. Vitamin Requirements of Drosophila. L Vitamines B and C. Biochem. Jour., Vol. 16, pp. 148-152, 1922. VI. A Comparison or THE Laws or MORTALITY IN DROSOPHILA AND IN Man PROFESSOR RAYMOND PEARL I In the first of these Studies (27) there were presented for the first time, so far as I am aware, complete life tables for any other organism than man. Up to the pres- ent time there have been presented in the published re- sults of the work of this laboratory on Drosophila (27, 32, 44, 49, 50) exact determinations of the duration of life in 24,329 individual flies. This is a statistically re- spectable mass of material, and warrants some general discussion. In the first study a rough, purely graphical compari- son of thé l- lines of the Drosophila and certain human life tables was instituted. This comparison, rough as is was, made apparent at once the fact that there was a fundamental similarity in laws of mortality in these two organisms. It is my purpose in the present paper to make a more exact comparison of the values of the life table functions in the two eases. It will be seen that the similarity is even closer than was supposed from the rough compari- son, and that in faet we are dealing here with qualita- tively identieal expressions of an obviously fundamental biological law. II Upon what basis shall any life table function, say le, of the Drosophila life table be compared with that of man? The life span of one of these organisms is best measured in days, while that of the other is measured in No. 646] . THE DURATION OF LIFE 399 years. This fact, however, offers no insuperable diffi- culty to the comparison. What is needed is to superim- pose the two curves so that at least two biologically equivalent points coincide. The best two points would be the beginning and the end of the life span. But in the ease of Drosophila our life tables start with the be- ginning of imaginal life only. The larval and pupal durations are omitted. In our preliminary comparison (27) we took human age 15, as the point corresponding biologically to the beginning of imaginal life in Droso- phila. I think we can get at this starting point more exactly by putting the human and Drosophila l- curves together as a starting point at the age for each organism where the instantaneous death rate qz is a minimum. In the ‘case of Drosophila, I think we are safe in concluding, on the basis of the work of Loeb and Northrop (14-17) as well as frqm our own observations, that this point is at or very near the beginning of imaginal life. We shall accordingly take Drosophila age 1 day as this point. Our life tables show that certainly after this time qz never again has so low a value. Indeed the fundamental law of mortality in Drosophila imagoes was stated in (27) in this way (p. 492): *'the instantaneous death rate inereases with age as a modified logarithmie fune- tion of x." The latest edition of Glover's (51) United States Life Tables gives (p. 68) for white males in the original registration states the following values for qz: for age 11-12 2.28, and for age 12-13, 2.29. We may, therefore, with sufficient accuracy take exactly 12 years as the minimum point, particularly as the l- figures we shall have to use are tabled as of the beginning of the age interval: ` For the other end of the life span we may conveniently take the age at which there is left but one survivor out of 1,000 starting at age 1 day for Drosophila and age 12 years for white males. This age for wild type Droso- 400 THE AMERICAN NATURALIST [Vor. LVI phila is, to the nearest whole figure, 97 days. To de- termine it for white males we have Table I, calculated from Glover's Table 9 | TABLE I SURVIVORSHIP OF WHITE MALES IN ORIGINAL REGISTRATION STATES ON THB Basis oF 1,000 AT AGE 12 Number Number Number Alive at Alive at Alive at Age Beginning Age Beginning Age Beginning of Age of Age of Age Interval l, Interval /, Interval l; 12-18... 1,000 45-46 803 78-79 194 13—14..... 998 46-47 792 79-80 Iri 14—15..... 995 47—48 782 80-81 150 15-16..... 993 48-49 771 81-82 130 1 rg 990 49-50 760 82-83 110 IT-IB..- 987 50-51 749 83-84 93 18-19..... 83 51-52 737 84-8 TF 19-20..... 979 52-53 725 85-86 63 20-21..... 975 53-54 713 86-87 51 po Reo 5: 970 54-5 699 87-88 41 22-23.... 965 55-56 686 32 28—24..... 960 56-57 671 89-90 25 24-25..... 955 57-58 655 90-91 1 25-26..... 950 58-59 639 91-92 14 W- 944 59-60 622 92-93 10 27-28..... 939 1 604 — 7 28-29..... 934 61-62 585 94—95 5 29-30..... 928 62-63 566 95-96 4 30-31..... 922 63-64 54 96-97 2 81-32..... | 916 64-65 525 97-98 1.59 32-33..... | 910 65-66 504 98-99 1.01 BUREN 903 66-67 482 99-100 .63 34-35..... 896 67-68 459 35-306..... 889 8-6 436 36-37..... 881 69-70 412 37-38..... 873 70-71 38 38-39..... 865 71-72 364 39-40..... 857 72-73 340 40-41..... 849 73- 315 41-42..... 840 74-75 291 42-43.... 831 75-76 66 43-44..... 822 76-77 241 44-45..... 812 71-78 217 From this it appears that there is almost exactly one survivor at 98 years. So then we have as biologically equivalent life spans 97 days of Drosophila life as imago — 86 years of human life. . Whence it follows that No. 646] THE DURATION OF LIFE 401 1 day of Drosophila life = .8866 year of human life and 1 year of human life — 1.1279 days of Drosophila life. it We are now in position to make an exact comparison between the life tables of the two organisms. This may be done perhaps most instructively by setting up lz lines for the two forms on the basis of age in centiles of the life span, rather than days or years. That is to say, the whole comparable life spans (as defined in this paper) of 97 days in Drosophila and of 86 years for white males will each be divided into 100 equal parts, and the survi- vors at the attainment of the beginning of each centile interval will then be computed. This is done for wild-type (long-winged) Drosophila males (Pearl and Parker (27) Life Table II) and male whites in original Registration states in 1910 (Glover's Table 9), in Table II. : The two life curves of Table II are shown graphically 1n Fig. 1, plotted on an arithlog grid. We have, in Table II and Fig. 1, for the first time, so far as I am aware, a precise quantitative comparison of the life spans and one of the mathematieal funetions of the mortality of two different organisms. It will be noted that: 1. The form of the l- distributions is fundamentally the same in both of these organisms over the equivalent life spans. Considering the extreme differences in habits of life, structure, physiology, and environmental stresses and strains in the two cases, this is a truly re- markable result. It seems to me to mean that the fae- tors which determine individual longevity, and differ- ences in this character, are biologically deeply rooted, at least as fundamental, apparently, as the faetors which determine the specifieity in the morphogenesis of organ- isms, and perhaps even more so. We are accustomed loosely to think that the prime faetors in determining 402 THE AMERICAN NATURALIST [Vor. LVI TABLE II SURVIVORSHIP DISTRIBUTIONS (Iz) FOR EACH CENTILE OF THE COMPARABLE SPANS OF (a) WILD TYPE DROSOPHILA MALES AND (b) W EN IN THE ORIGINAL REGISTRATION STATES IN 19 Centile |Numbers Alive at Beginning| Numbers Alive at Beginning of Com- of Centile Age Interval Centile of of Centile Age Interval parable Comparable Life Life Span Span Drosophila Man Drosophila Man On il. 1,000 1,000 51-52 360 673 Te:M cen 991 998 52-53 344 659 2-8... 981 996 53-54 328 645 ge du. 972 994 54-55 312 631 4— 5..... 963 991 55-56 296 616 5- 0..... 54 989 56-57 280 601 6— 7. 945 986 57-58 265 585 7—- 8..... 935 983 58-59 250 568 8- 9..... 926 980 59-6 235 551 9-10..... 917 976 60-61 221 533 10-11. .;.. 972 61-62 207 515 H-a 898 968 62-63 193 496 2-13... 888 963 63-64 180 477 13-14..... 879 959 64-65 167 458 14-15.... 869 955 65-66 155 438 15-10... 859 950 6-67 143 418 16-17..... 946 67-68 132 397 17-18..... 941 68-69 121 377 18-19..... 828 936 69-70 111 356 19-20..... 818 932 70-71 102 335 20-21.... 807 927 71-72 314 2122..... 796 922 72-73 292 py ea, Saas 785 916 73-74 76 271 23-24..... 773 911 74-75 24-25..... 761 905 75-76 61 229 25-20..... 749 899 76-77 55 208 20-21;. 737 893 77-78 49 189 27-28..... 725 887 78-79 43 169 28-29..... 712 880 79-80 38 151 29-30..... 699 S74 80-81 34 133 0-31..... 867 81-82 30 117 31-82..... 672 ` 860 82-83 26 101 82-33.... 659 852 83 22 33-34.... 645 845 84-85 19 74 34-35..... 630 838 17 62 35-36..... 616 830 86-87 14 52 36-37..... 601 822 87-88 12 43 87-38..... 586 814 88-89 10 35 38-39..... 571 89-90 9 28 NUES 555 797 91 T 22 40-41..... 540 7 91-92 6 17 -42.... 524 719 -93 5 13 42-43..... 770 93-94 4 10 Vea 492 760 94-95 3.43 8 ES 475 750 95-96 2.80 6 45-40..... 459 740 7 2.97 4 SPA 443 730 97-98 1.84 3 47-48..... 426 720 98-99 1.47 2.11 48-49..... 410 709 99-100 1.18 1.37 49—50..... 697 100 .94 .87 50-51..... 377 685 SURVIVORS No. 646] THE DURATION OF LIFE 403 human longevity are such things as the infectious dis- eases, exposure to unfavorable environment, ete. But Drosophila, which so far as is known has no infectious diseases, and in general meets a set of environmental L000, m — ~~, ~ he “W Q Z D 2 a y NL N MALAS N \ 100 M V N X \ XII Ad ; \ PT X / O 4 8 2 6 20 24 28 32 36 40 44 48 52 56 60 64 68 72 76 80 84 68 92 96 00 AGE CENTILES 1. Comparing the survivorship Ss aoe. of Drosophila and man Prec in both cases) over the equivalent life s conditions wholly different, both qualitatively and quan- titatively, from those which operate on man, shows fundamentally the same form of distribution of degrees of longevity 404 . THE AMERICAN NATURALIST [Vor. LVI 2. When compared exactly, on the basis of comparable life spans, the human being has at every equivalent age a higher relative expectation of life than does Droso- phila, measured in terms of its own life span in each ease. That this was the case for all but old age was concluded from the rough graphical comparisons of the first Study in this series. It is now seen that the same is true over the whole of life. From this faet the con- elusion appears warranted that while the laws of mor- tality are fundamentally the same in kind for Drosophila and for man, they differ somewhat quantitatively. There is a temptation to conclude further that the quantitative difference finds its cause,in man's own control and amelioration of his environment though sanitation and hygiene. Such a conclusion, however, seems to me not to be strietly warranted, in the light of our present knowl- edge. There is some suggestion that it is true, as was pointed out in the first of these Studies, from the fact that the progressive change of the human ls curve in form during historical times has been in the direction of moving from the form typical of Drosophila to that now found for progressive, highly civilized groups of men. But definitive conclusions on the point must await further research. 3. The details of the quantitative differences in the two curves are interesting. When the first 25 per cent. of the equivalent life spans has been passed Drosophila has lost almost exactly 25 per cent. of the individuals starting life together, while man has lost but 10 per cent. When 50 per cent. of the life spans has been completed Drosophila has lost 72 per cent. of the individuals start- ing together, while man has lost but 31.5 per cent. At 75 per cent. of the life span, Drosophila has lost 94 per cent. of the individuals and man 77 per cent. From the 93d centile of the equivalent life spans on practically to the end, man has more than twice as many survivors out of a thousand starting together as does Drosophila. Exactly similar results to those here presented are ob- No. 646] THE DURATION OF LIFE 405 tained if one compares human and Drosophila life curves for females. Since nothing new in principle is brought out, it is not thought necessary to present the female curves here. IV In this paper it is shown that if we take as equivalent life spans in Drosophila and man the period between (a) the point in the life history of each organism where the specific death-rate (qz) is a minimum, and (b) the point where there is one survivor out of 1,000 starting at the beginning as defined in (a), and then divide these equiva- lent life spans into 100 portions (thus measuring age not in absolute units but in centiles of the life span), the laws of mortality are fundamentally the same in kind in the two organisms. There is a quantitative difference ex- pressible in the statement that at each centile age throughout the life span the number of survivors, out of the same original number starting together, is higher in man than in Drosophila. In a subsequent paper, I hope to take up in detail the funetional relations (in a mathematieal sense) be- tween the human and Drosophila equivalent le curves | here presented. LITERATURE CITED (The plan of numbering citations is explained in the second of these Studies, Amer, NaT., Vol. 56, 4.) 49. Pearl, R. and Parkir, S. L. Experimental Studies on the Duration of IV. Data on the Influence of Density of Population on Dura- tion of Life in Drosophila. AmER. NAT., Vol. 56, pp. 312-322, 1922. 50. Id. _Bapermental Studies on the Duration of Life. v. On the Influence Certain Environmental Faetors on Duration of Life in Drosophila. 922. Tii, Vol. 56, pp. 51. Glover, 3; W. United States Life Tables 1890, 1901, 1910, and 1901- 10. Washington (Bureau of the Census), 1921. Pp. 496, 4to. THE SYSTEMATIC LOCATION OF GENES BY MEANS OF CROSSOVER OBSERVATIONS R. A. FISHER RorHAMSTED EXPERIMENTAL STATION INTRODUCTORY . Ix the construction of a chromosome map, the dis- tances between neighboring genes are equated to the per- centage of crossovers which have been observed between them. Owing to errors of random sampling, and some- times to other disturbing causes, inconsistencies always arise between the distances so determined. For example, in the important data given by Lancefield and Metz for the sex chromosome of Drosophila willistoni [1, p. 241] we have the following values: TABLE I | Crossover Number of | Number of stat Percentage Observations Crossovers Staite to Beaded. 606s. o Lee 279 | 4 Beaded to Hough... v y ea A 2.42 455 | 11 Pante to Roh. c oe | 7.09 6388 | 453 Within such a small range, double erossing over may be ignored; yet it would be wrong to use such inconsist- eneies as an argument against the linear arrangement of the genes. For although the true erossover values may be accurately additive, errors of random sampling will certainly disturb the observed percentages. The practi- eal problem is to assign to the distances between the genes values which shall be as far as possible in accord with the whole of the observations available. In other words, we have to make use of as much as practicable, ideally the whole, of the information supplied by the data; giving due weight (i) to the greater accuracy of the values obtained from the larger number of observa- tions, (ii) to the greater accuracy of values obtained from 406 ‘ No. 646] THE SYSTEMATIC LOCATION OF GENES 407 closer pairs. In general, too, we shall have to consider not three genes only, but a large number, lying sufficiently close together for double crossing over to be ignored, the percentage observed between each pair of which gives indirect information as to the position of all the others. AE E Pines In its general character the prob- lem resembles those problems in- M. DEFORMED volving errors of observation, where 8 L a smaller number of unknowns are determined from a larger number f ee of inconsistent equations, and which 7 + are usually solved by the method m mu of least squares. The practical so- " lution depends on the construction of a number of ** normal equations ”’ for the unknowns, in which the in- consistencies of the data are prop- erly weighted and made to balance. To make the sum of the squares of 4 L the errors of the crossover percent- ages a minimum would, however, be wrong, and the method of least 3 Me Beapepd squares is not directly applicable. It has been shown that the whole of the information supplied by the 2 F ~ data (2) is made use of by the E Peach method of maximum likelihood, and by a first approximation the required normal equations may be . construeted. o LB Scute 2. MATHEMATICAL THEORY : —l—- REDUCED In the above example, if we71 f write p, and p: for the two adjacent erossover ratios, the probability of the aetual series of observations 408 THE AMERICAN NATURALIST [Vor. LVI will be proportional to pil — p)?5p4(1 — p: i(pi + p2)5(1 — pi — p2)995 and the likelihood of any given pair of values for p, and p. will be proportional to the same quantity. In order: to make this quantity a maximum for variations of pi and p, we have the equations 4 275 453 i 5935 n p 1—mp 2t» 1—m—7 453 5935 11 444 pts Lom "95 ^ im These equations are exact, but for practical purposes we need equations linear in p; and p;, and a first approxi- mation is sufficient; if p differs little from z/(« + y) = a/n, then : 1 4 (GF Er lue a) te LP So that we may rewrite Sees (1) in the practical and approximate form 279: 6388? ax2:5P7 t i x 2 (m t») = ae T 5935 ' 6388 6388? | 455? 453 x 5935 P! + P9 ta x qi nt = 5035 | 444 For each percentage observation, therefore, we have merely to calculate the two quantities »?/zy and n?/y; then normal equations may be constructed in the form aP F458, E iR. ah +%.P,+:--=), where a2 is the sum of the quantities »*/zy for which both p; and p; are involved, a,, the corresponding sum for all in which p, is involved, and b, the sum of the quanti- ties n?’/y for which p; is involved. 3. Practica, EXAMPLE In order to illustrate the practical application of this - method to a complex ease, we will consider the location of the 8 genes, from Reduced to Rimmed, in the middle No. 646] THE SYSTEMATIC LOCATION OF GENES 409 of the sex chromosome of Drosophila willistoni. We have here 7 intervals to determine, and fifteen crossover per- centages are given [1]. Table II shows the data, and the series of weighting quantities derived from them. TABLE II Per- cent-| x n n?/y ni[zy Unknowns Involved Reduced-Seute..... .95 | 27| 2,848 |2,875.26 gr 287 pı Reduced-Rough....| 6.24 | 37| 593 | 632.46 0,136 pi, pz, ps, De Seute-Peach....... 81 8|. 442| 450.15 aL 1 Scute-Beaded...... 1.4 279 ,742 Seute-Rough...... 7.09 | 453| 6,388 |6,875.58 | 96,956 p» Ps, Ds Scute—Deformed 7.24 50 6 2 ,205 ps, ps, P4, Ds, Seute-Rimmed.....| 9.91 | 189| 1,908 2,117.78 | 21,379| P2, ps, pa, Ds, De, D7 Peach-Beaded..... 1.70 3 6| 179.05 | 10,504 i each-Rough...... 5.05 | 33| 654| 688.75 | 13,650) - ps, 74 Beaded-Rough..... 2.42| 11 455 | 466.27 | 19,287 pi Rough-Triple...... 49 4| 809| 813.02 | 164,433 Ps Rough-Deformed ..| 2.39 | 12 503 | 515.29 | 21,599 ps, pe Rough-Rimmed....| 2.26 | 62| 2,742 |2,805.43 | 124,072 ps, pe, PT Triple-Rimmed. 1.00 6| 601| 607.06 | 60,807 pe, P7 Def formed-Rimmed. 417| 2 48 50.09 1,202 From this table we write down the normal mice 313,423p + 10,136(p, + P, +P, 7.72 10 1136p, + 183,380p, “4 158 509p, -+ 138,766p, + 31,674p, + 31,6 lip, 21, 379p, = — 11,103.93 10,136p, + 158,509p, + 182,663p, + 152,416p, .- 31 enam, + 31,674p, 1,379p, = — 11,521.58 10,136, + 138,766p, + 152,416p, + 171,703p, + 31 srin, + 31,674p, 21,379p, = — 11,525.74 31,074 (p, + P, + p,) + 341,778p, + 177 345p, + 145,451p, = 6,996.42 31,674 (p, + p; +r, J+ 177,345p, + 238,152p, + 206,258p_ = 6,790.46 21,379 (p, he p, +4 p, ) +145 451p, 4 206,258p, + 217,460p — 5,580.36 Using a calculating machine, the work so far is rapid and mechanical; the solution of the normal equations may in this case be much simplified by observing the uni- formity of some of the sets of coefficients, a type of uni- formity which is probably characteristic of crossover data. Thus by considering (p: +p: + pı) as a single quantity, p, is immediately expressible in terms of it, and by solving the last three equations we may do the same for ps, p, and pz; substituting finally in equations (2, 3, 410 THE AMERICAN NATURALIST | Vou. LVÍ 4) we solve them for p», p, and p,, and obtain the values shown in Table III. The seven values obtained give mutually consistent values for the crossover percentages between the fifteen pairs tested, and are therefore suitable for the construc- tion of chromosome map. If the conditions of Maximum Likelihood had been exactly fulfilled they would agree better than any other consistent series of values with the percentages observed. As it is, it is only in the ab- errant value of p; that the assumption that the observed values are approximately correct breaks down, and it is probable that such cases will ves occur when the data are TORENT insufficient. TABLE III is : | Standard da CM Observed oe d i LC air Boso Preece. .90 pı .95 + .05 18 0 ced-Rough.......... 7.66 6.24 —1.4 1.09 1.70 Seats Paes wie Gree Gee) 1.67 ps 1.81 + .14 05 S WI. acci. 2.98 1.43 —1.53 1.02 2.31 Seute-Rough............. 6.76 7.09 + .33 1.18 te-Deformed.......... 8.40 7.24 —1.16 1.06 1.20 Scut SE S VIVIDE A. 8.97 9.91 + .94 65 2.09 Peach-Beaded............ 1.31 ps 1.70 — .39 86 2 Ph ROUGH: i sous 5.05 c .86 s Beaded-Rough........... 3.78 pa 2.42 —1.36 .89 2.34 Rough-Triple... irere: ps 4 — .29 A8 Rough-Deformed......... 1.64 2.39 + .75 .57 133 Rough rs BARAT 2.21 2.26 —..05 .28 *- X Triple-Rimmed.......... 1 1.00 -— .50 1.08 Deformed-Rimmed....... 57 pr 4.17 +3.60 1.09 10.91 i | | | | x! = 25.34 Table III is arranged to compare the differences be- tween the calculated and the observed percentages with the standard errors due to sampling; except for p: all the differences are less than twice their standard errors; thus showing the general agreement between the data and the theory of linear arrangement of the genes. The fit, however, is not a close one, even if we omit pz; in the present state of our knowledge this will not throw any No.646] THE SYSTEMATIC LOCATION OF GENES 411 doubt on the scheme of linear arrangement, but will sug- gest that the erossover ratios in this part of the chromo- some were not constant in all the strains used to compile the data. In estimating the Goodness of Fit of data of this kind, x? may be calculated by summing the values of d?/c?, as in Table III Attention should, however, be called to the faet that it has been recently shown (3) that in enter- ing Elderton's Table we must put n’ equal to one more than the number of degrees of freedom, remaining after we have fitted our unknowns to the data. In the present case we have found 7 unknowns from 15 equations, leav- ing 8 degrees of freedom, so that w' should be 9, and not 16. In eonclusion it should be noted that to be available for the use of this process the erossover data should be stated in the form in which it is given by Lancefield and Metz, in whieh the erossovers tabled between any two genes do not include those experiments in whieh an inter- mediate gene was under observation. The practice of throwing together all the crossovers between two genes, in order to improve the ratios between the more distant points, causes the same crossover to appear repeatedly in different entries. The data are no longer the product of independent experiments, and must be re-summarized before reduction. REFERENCES 1. R. C, Lancefield and C. W. Metz. The Sex-linked Group of Mutant aeters in Drosophila wilistoni. AMERICAN NaTURALIST, LVI, pp. 211-241. 2. R. A. Fisher. On the Mathematieal Foundations of Theoretieal Statis- ties. Phil. Trans, A, COXXII, pp. 309-368. ; 3. R. A. Fisher. On the Significance of x’ from Contingency Tables and on the Caleulation of P. Journal of Royal Statistical Society, LXXXV, pp. 87-94. LINKAGE IN PEROMYSCUS DR. F. B. SUMNER Tue Scripps INSTITUTION For Bronoorcan RESEARCH, La JOLLA, CALIF. Srupents of Mendelism are beginning to display the same interest in possible homologies between the genetic factors or ‘‘genes’’ of different species of animals or plants which the morphologists of thirty years or more ago did in homologies between organs. In considering a given case of suspected homology between genes, two criteria are, so far as I know, employed: (1) Resem- blance between the developed characters which are at- tributed to the action of supposedly homologous genes. Mere similarity of appearance, however, is recognized as an extremely fallible criterion of homology here as in the case of comparative anatomy. (2) Agreement be- tween the ‘‘cross-over’’ value shown by a pair of linked factors in one species, as compared with the correspond- ing value shown by supposedly homologous factors in another species. If both of the two linked genes under consideration are found to have much the same somatic effects in the two species, and if, furthermore, the de- gree of linkage is approximately the same in the two cases, the argument is strong for a twofold homology. Metz! and Sturtevant? have been investigating the parallel mutations of several species of Drosophila, and it is not unlikely that this genus will furnish the best material for the study of genic homologies, just as it has shown incomparable superiority for certain other lines of genetic research. For rodents, what appear to be parallel mutations have been shown to occur among numerous species, even ones 1 Genetics, March, 1918. 2 Genetics, January, 1921. 412 No. 646] LINKAGE IN PEROMYSCUS 413 belonging to widely different families. In one case, that of the mutation known as ‘‘ pink-eye,"" not only is the visible modification elosely:similar in rats and mice, but the linkage relations between this factor and that for albinism are known to be of the same order of mag- nitude in the two animals.* Some years ago, Castle? described two similar muta- tions in the Norway rat, which he termed ‘‘pink-eyed yellow’’ and ‘‘red-eyed yellow,’’ respectively. These, according to the published descriptions, differ chiefly in the color of the eyes, the latter variety having darker eyes than the former. These two mutations, and like- wise true albinism, were all found to result from the modification of distinct genetic factors. Any two of them, when crossed, gave rise to the wild type in the first hybrid generation. On the other hand, further breeding tests led Castle to conclude that all three of these factors were linked. When red-eyed and pink-eyed rats were interbred, the cross-over percentage proved to be about 18. When pink-eyed rats were crossed with albinos, this value proved to be about 21. On the other hand, the linkage between red-eye and albino proved to be almost absolute. -One hundred and sixty F, albinos and 57 F, red-eyed yellows, when mated with pure red- eyes and albinos, respectively, yielded but a single off- spring which was not of the wild type. More recently Dunn? has tested the linkage between this same red-eyed condition and albinism in the rat. From his own data he computes a cross-over value of 1.8 per cent., but when his data are combined with those . of Castle, this value falls to less than one per cent. Castle and Dunn have likewise tested the degree of linkage between ‘‘pink-eye’’ and albinism in the mouse 3 Dunn has eompiled these cases in a useful article in the Journal of Mammalogy, August, 1921. * According to Castle, the percentage of cross-overs is 21 for rats and 14 for mice. This may or may not be construed as evidence of ‘‘homology.’’ 5 AMERICAN NATURALIST, February, 1914; Science, August 6, 1916 (with Wright); Carnegie Institution Publications 241 and 288. 6 Genetics, May, 1920, 414 THE AMERICAN NATURALIST [Vor. LVI (Mus musculus). The proportion of cross-overs was found to be about 14 per cent. Some five years ago I described a pale, red-eyed mutant of Peromyscus,’ which originated among the off- spring of three sibs in the F, generation of a cross be- tween P. maniculatus rubidus and P. m. sonoriensis. Since I have already described this ‘‘mutant’’ race rather fully, and since it will again be discussed shortly in a paper by Mr. H. H. Collins and myself, I need not enter into a detailed account of it here. I have not seen specimens of the ‘‘red-eyed yellow’’ rats described by Castle, but I find little in the description of that race which is at all at variance with my own ‘‘pallid’’ race of Peromyscus. The latter has undergone a great re- duction of the black pigment, while the yellow pigment has been little if any affected. The eyes are commonly dark red, rather than pink, though they present a con- siderable degree of variability, ranging from a condi- tion not much darker than the true pink of albinos to a condition not much paler than the normal. There are, however, no real intergrades between the pallid mice and the wild type, and the behavior of this complex of char- acters in crosses is that of a simple monohybrid reces- sive. Furthermore, it is not an allelomorph of albinism, since the wild type alone results from matings between albinos and pallids. I have recently carried out tests of the linkage rela- tions between this factor and that for albinism. Thus far, it has not been found practicable to devote any con- siderable proportion of my time to this phase of the sub- ject, and the numbers are accordingly inadequate for any exact measurement of cross-over values. They are, none the less, sufficient to show the existence of a high degree of linkage between these factors. The number T Genetics, May, 1917; AMERICAN NATURALIST, August-September, 1918. This mutant was at first referred to as a ‘‘partial albino’’; later the non- committal term ‘‘pallid’’ was applied to it. 8 The albinos used were all derived from a single brood belonging to the subspecies Peromyscus maniculatus gambeli. No. 646] LINKAGE IN PEROMYSCUS 415 of F, individuals derived from simple F, x F, matings is too small to give a representative dihybrid ratio. The really important tests have been made with ‘‘extracted”’ albinos and pallids of the F, generation. Matings have been made (1) between ‘‘ extracted ’’ al- binos and ‘‘pure’’ pallids (1.e., those known to be free from the factor for albinism), (2) between extracted pallids and pure albinos, and (3) between extracted pal- lids and extracted albinos. There were likewise a num- ber of matings in which the pedigrees were less simple.? On the assumption of a wholly independent segregation of these factors, our F, pallids (of simple pedigree) should have a 2/3 chance of being heterozygous for al- binism, while our F, albinos should have a 3/4 chance of being either homozygous or heterozygous for pallid.?° Eighteen F, mice were involved in these tests. The total number of offspring derived from these was 135, the number per parent ranging from 3 to 26. By no means all of these parents, taken singly, have thus far given birth to a sufficient number of young to prove their genetic composition with any certainty. But the cumula- tive testimony of all of these matings is overwhelming. Not a single pallid mouse and only two albinos have ap- peared among the 135 young which have thus far been born. Had there been a normal proportion of ‘‘carriers’’: among the parents, these matings should have yielded 37 albinos and 18 pallids, as the most probable ‘‘expected’’ numbers. That all of the offspring with two exceptions (these being sibs) were of the wild type is evidence of a high degree of linkage (in this case ‘‘repulsion’’) be- tween the albino and the pallid factors.” ® Baek-erosses and heterozygous albinos figured in some of these pedi- grees. In these cases the odds are different from those which hold for indi- viduals derived from the simpler types of mating. They have, however, been computed for every animal used. In about half of the ‘‘extracted’’ albinos, for example, there was only a 5/8 chance that the individual carried the pallid factor. 10 It is a safe assumption that the double recessive form would be albino. 11 It might be supposed that the testimony of 18 parent mice, even if all of these were shown conclusively to be lacking in ‘‘cross-over’’ gametes, 416 THE AMERICAN NATURALIST [Vor. LVI From these considerations we may regard it as not unlikely that my ‘‘pallid’’ race of Peromyscus has re- sulted from the mutation of a genetic factor homologous with that which has mutated in the case of Castle’s ‘‘red- eyed yellow”’ rats. This decisive result, as regards the existence of link- age between the pallid and albino faetors in Peromyscus, stands in contrast with the apparent absence of such linkage in another eross between mutant strains of these mice. Albinos were mated with mice belonging to a strain which I have elsewhere referred to rather inap- propriately as 'yellows.''!? The latter vary from clay color to a distinctly reddish hue, according to the strain, and are characterized primarily by a marked increase in the length of the ‘‘agouti’’ cross-band and by a decrease in the proportionate number of all-black (unbanded) hairs in the pelage. Where present, however, the black pigment is of full intensity. This applies to the basal zone of the body hairs, both dorsal and ventral, to the black hairs of the dorsal tail stripe, as well as to the eyes, ears and soles of the feet. Matings between albinos and ‘‘yellows’’ have resulted exclusively in F, mice of the wild type (dark). An F, generation of 83 was obtained, consisting of 52 dark individuals, 13 yellows and 18 albinos. On the assump- tion of purely random assortment of gametes, the ‘‘ex- pected” numbers are 44, 15 and 20, respectively. ‘The observed numbers are doubtless within the range of ‘‘ac- eidental" variability. In any case they give no evidence would not be sufficient to prove the existence of linkage. It should be repeated, however, that we are not here dealing with cases in which there would be merely an equal chance of combining the two mutant factors in the same individual, The odds in favor of this (linkage aside) may stated above, be as high as 2 to 1, or even 3 to 1. Thus, the likelihood of obtaining, pi opr alone, 17 non-cross-over cases out of 18 becomes vanishingly 12 Genetics, ae 1917; AMERICAN NATURALIST, Augast -Hepteuber, 1918. A more complete account of these mice, dealing with two subvarieties dif- fering somewhat in color, is included in a forthcoming paper by Mr. H. H. Collins and myself, No. 646] LINKAGE IN PEROMYSCUS 417 of linkage, the occurrence of which would have reduced the proportionate number of dark individuals, instead of increasing it. The number of F, albinos and yellows which have been thus far tested is very small, but it is of interest that the proportion of recombinations is even greater than would be expected from random assortment. In- clusion of these meager results in the present report seems justified by the probability that we shall not soon rear any considerable number of hybrids between the yellow and albino varieties. Seven extracted albinos have been mated with pure yellows. Three of these have given only yellow offspring, the numbers being 9, 13 and 21, respectively. Thus, three of these seven albinos are, in all probability, double re- eessives (ccyy). (One in four should be double reces- Sives, according to chance.) Three other albinos have given mixed offspring. They are evidently of the for- mula ccY y. The remaining one appears to have the for- mula ccY Y, as judged by the produetion of 15 dark young. Two extracted yellow females mated with a (sup- posedly) pure albino male gave birth to 4 albinos and 4 dark.'* No albinos would be expected here if linkage were complete, while only one third should be albinos in the total absence of linkage. Thus the number of re- combinations is again too high, even on the assumption of no linkage. These numbers are, of course, very small. But even here such proportions would have been quite improbable had any marked degree of linkage existed—such, for ex- ample, as has been found to exist between the pallid and albino factors. 13It is only fair to add that 4 yellows likewise resulted from these ma- tings. This was doubtless due to the fact, unsuspected at the time, that the albino male carried the ‘‘yellow’’ factor, one of his two great-grand- parents having been heterozygous for yellow. THE SOUND-TRANSMITTING APPARATUS OF SALAMANDERS AND THE PHYLOGENY OF THE CAUDATA E. R. DUNN SMITH COLLEGE ResEARCHES by Kingsbury and Reed, extending through a number of years, have shown that the sound-trans- mitting apparatus of salamanders consists of two ele- ments. These are the columella and the operculum. In the most recent paper on this subject, Reed (1920) gives a résumé of all the previous work, an extensive ac- count of the state of affairs in the Plethodontide, a brief account of the conditions in other forms, and the findings are presented in the form of a family tree. The purpose of the present article is to add an aecount of the condition of the apparatus in two forms not seen by Reed, to question the condition described by Kings- bury and Reed for Dicamptodom ensatus (Ambystoma tenebrosum Auct.), to suggest a somewhat different inter- pretation of the faets observed by them, and to propose a somewhat different phylogeny, which seems to agree quite as well with the otic apparatus and far better with other anatomical features. Kingsbury and Reed (1909) were unable to examine any of the Asiatic forms related to Hynobius. These forms, as Cope pointed out long ago, are rather different from the Ambystomide, with which they have usually been associated, and should in fact form a family Hy- nobide. I have recently been able to examine large series of Hynobius leechii from Korea. This animal shows a con- dition of the otic apparatus different from any seen by Kingsbury and Reed, and a condition which I am com- pelled to consider primitive. Both columella and oper- 418 No. 646] THE PHYLOGENY OF THE CAUDATA 419 culum are present as free and distinct elements. Both are readily movable. There is a m. opercularis. I have not been able to examine skulls of Onychodac- tylus, or of Ranodon. Okajima’s (1908) figures of Ony- chodactylus show only one element which is in appear- ance much like that of Cryptobranchus. This is very different from the appearance of the apparatus of Hyno- bius. It is evident that either fusion of operculum and columella has taken place or that the operculum has not developed. Onychodactylus is partly aquatic, a moun- tain brook animal. Cryptobranchus, which, as I shall show later, is a derivative of the Hynobiide, has failed to develop the operculum. Probably the same is true of Onychodactylus and of Ranodon as well, although for the latter Wiedersheim’s (1877) figure is all we have. Still, as Kingsbury and Reed (1909) say, his Fig. 67 ‘‘suggests a condition such as is found in Cryptobran- chus.”’ i: Rhyacotriton olympicus was not examined by Kings- bury or by Reed. This animal (Dunn, 1920) possesses both columella and operculum. The columella is free from the periotic and is readily movable. The operculum is little developed. The animal is in part aquatic, a moun- tain brook species. Dicamptodon ensatus was examined by Kingsbury and Reed (1909), and while my dissection of an adult showed the state of affairs which they describe, I can not follow them in calling it ‘‘much like that in the adult Amby- stoma.” In adult Ambystoma the columella is solidly fused to the periotic. A bony operculum nearly fills the opening of the fenestra, and is attached by a membrane around its circumference. In Dicamptodon, on the other hand, about half of the fenestra is filled by the plate of the columella, and the remainder by cartilage. The car- tilage extends around the plate of the columella. There is nothing that could be called a definite operculum. If the cartilage is called the operculum, then the columella . and operculum are fused and the operculum is fused to 420 THE AMERICAN NATURALIST [Vor. LVI the ear capsule by nearly its whole border. It seems to me that in this ease the columella is at least more free than in Ambystoma, and the operculum less developed. This would be in line with what is known of the habits of Dicamptodon. It is a much more aquatic animal than is Ambystoma. In the Caudate sound-transmitting apparatus, taking Reed (1920) as a basis, there are the following sets of conditions: I. Both columella and operculum present. Both fre Hynobius, coa ue II. Opereulum not developed. Columella free. ptobranchus. Sil See dl Ranodon Onychodactylus goes t III. Operculum developed, free. Columella fused to periotis, stylus present. alamandra, ystoma. IV. Opereulum developed, free. Columella fused to periotie. Stylus absent. Triturus, Pachytriton, P:eurode'es 1, Ty ototriton 1. V. Operculum developed, free. Columella ?, Siren, Batrachoseps. VI. Both columella and opereulum present, fused together, free from periotie. Necturus. — columella and operculum vitu Fused together. Opereulum ed by narrow fusion to per par e Plethodontide (exe. Batrachoseps). Inasmuch as II is a condition found also in larvæ, there is no reason to suppose that the animals in which this condition occurs form a natural group. Condition V has been commented upon by Reed (1920), and I am fully in accord with his ideas in this connection. Siren and Batrachoseps are certainly not related. Both are extremely specialized. Batrachoseps has certainly passed through stage VII. Siren has certainly passed through an ancestral period of terrestrial life, yet its other peculiarities are such that it is dangerous to state that its relationships are with the forms in stage IV. he forms which show condition VI and condition VII No. 646] THE PHYLOGENY OF THE CAUDATA 421 form what Reed (1920) calls Legion II, as distinct from the forms which show conditions I-V (exe. Batracho- seps), which Reed calls Legion I. But the sound-transmitting apparatus of Necturus agrees with that of Amphiuma and the Plethodontide only in having the columella and operculum fused. There is no reason to suppose that such a fusion may not have occurred twice, especially as the details of the fusion in Necturus differ somewhat from the manner in which the fusion occurs in Amphiuma and the Pletho- dontide. In Necturus the columella forms a goodly part of the plate-like portion of the apparatus. In the forms of condition VII, the plate-like portion is almost entirely composed of the operculum, and the columella is repre- sented by the stylus. In this case the evidence of the ear bones is non-eommittal. Considered apart from all other features of the anatomy condition VII might equally well be derived from condition VI or both inde- pendently from condition I. But, as we shall see, evi- dence from other features of the anatomy precludes our regarding Necturus as intervening between the Pletho- dontide and the other Mutabilian forms. It is extremely interesting to note that Reed has found almost exactly the same state of affairs in Amphiuma and in the Plethodontide. The exact relationships of Am- phiuma have long been in dispute, and while I prefer to be eonservative about the position of the animal, I think it extremely likely that further evidence will show that it is closer to the Plethodontide than it was placed in the older classifications. Any classification should be based upon all available characters, so that possible parallelisms will not lead to wrong conclusions. In the present instance we are deal- ing with a stock neither absolutely terrestrial nor abso- lutely aquatic. From this stock there have been several branches which have become more aquatic and several which have become more terrestrial. Excellent examples of this are the numerous incursions into a mountain 422 THE AMERICAN NATURALIST [Vor. LVI brook habitat, with the penalty of loss or reduction of lungs. The list is extensive, Onychodactylus, Rhyaco- triton, four species of Triturus, Salamandrina, Chio- glossa, all the stock of the Plethodontide, an assemblage representing four families. The sound-transmitting ap- paratus is admittedly correlated with the mode of life. Therefore as a character in determining relationships it must be used with extreme caution. The following outline classification of salamanders does not counter any of the facts concerning the otic apparatus, and is based on many characters. As regards the Plethodontide and the Hynobiide, re- visions of both are nearly completed, based on the exam- ination of some 8,000 specimens of the first family and 1,000 of the second. The Sirenide are the most isolated group. Searcely a character can be found to ally them with one or an- other of the main stocks. The pelvis is gone, the skull is that of a very specialized larva, the hyoids are those of almost any larva, the tail vertebre are very different from those of any other salamander, inasmuch as there is no hemal arch. There are flat plates on each side which do not meet in the mid-ventral line. There is no prearticular. The Proteide are only slightly less isolated. The pelvis differs in having an anterior median projection and no ypsiloid apparatus. The skull is larval. The bran- chial arches are reduced from the primitive larval quota. The prearticular is absent. The Amphiumide also have modified larval branchial arches, and the pelvic girdle lacks the ypsiloid appa- ratus. But Amphiuma has an adult skull which resembles remotely that of the Salamandride. The otic apparatus is that of the Plethodontide. There is no prearticular bone. It is quite possible that this genus is descended from primitive Salamandrids. The others have directly comparable skulls, branchial arches, and pelves, and in dealing with their relationships we are on much firmer ground. No. 646] THE PHYLOGENY OF THE CAUDATA 423 Several characters divide them into two series, which should, I think, rank as superfamilies. 1. Prearticular bone. Present in Cryptobranchide and Hynobide, and absent in Ambystomide, Salaman- dride and Plethodontide. Second epibranchial. Present in Cryptobranchide and Hynobiide, and absent in Ambystomide, Sal- amandride, and Plethodontide. First ceratobranchial and first epibranchial fused into a single cartilaginous rod in Cryptobranchide and in Hynobide. Separate elements in Ambysto- ' mide, Salamandride (exe. Salamandra, where all parts fuse), and Plethodontide. Nasals meeting in median line and premaxille without nasal process in Cryptobranchide and Hynobide. Nasals separated by nasal spines of premaxille in Ambystomide, Salamandride, and Plethodontide (exe. Pseudotriton, where nasals overlap premaxil- lary spines). Pubotibialis muscle fused with puboischiotibialis in Cryptobranchide. The two muscles are separate in all other salamanders (Noble, 1922). I have ascertained that the two are fused in Hynobius and in Onychodactylus. Larve of Ambystomide, Salamandride, and Pletho- dontide have the first ceratobranchials fused with the second basibranchial (Smith, 1920). This fu- sion does not occur in larve of he ant or of Hynobide. ho d e les g Within the superfamily Salamandroidea the Ambysto- mide and the Salamandride are about parallel. The long posterior process of the prevomer distinguishes the Salamandride, and as the parasphenoid tooth patches of Plethodontide are the morphological equivalent of this process (Wilder, 1920) it is probable that some primi- tive Salamandrid (having the two otie elements free) gave rise to the much degenerate Plethodontide. The 424 THE AMERICAN NATURALIST [Vor. LVI mountain brook habitat of the ancestral Plethodontid (Wilder and Dunn, 1920) accounts perfectly for the re- tention of the columella through adult life as a working part of the sound-transmitting apparatus. "The Cryptobranchoidea contains two families. Of these the Hynobiide is the more primitive. The Crypto- branchide differ in lacking the lachrymal bone, in the larval position of the vomerine teeth, and in the much depressed form of the body and head, the last two evi- dently adaptations for aquatie and bottom-living habits. Besides the characters mentioned in the list as aligning the Cryptobranchide with the Hynobiide, several minor points also show this relationship. Both Ranodon and Hynobius frequently have a lateral fold between the in- sertions of the legs. This is very prominent in both Cryptobranchus and in Megalobatrachus, and is not found elsewhere. Onychodactylus larve have a marked fold on the posterior side of the limbs. This is seen else- where only in Cryptobranchus and in Megalobatrachus. Inasmuch as the characters differentiating the two gen- era of Cryptobranchide have not been clearly understood in the past they are here stated. Megalobatrachus, Two persistent branchial arches: Frontal not entering naris: Branchial clefts closed in adult. Cryptobranchus, Three persistent branchial arches: rontal entering naris: Branchial clefts open in adult. In all three of these characters the American genus shows greater adaptation to aquatic life. The European fossils of this family appeal to Megalobatrachus in the one skull character which separates the two genera. Neither in Andrias schuchzeri nor in A. tschudii does the frontal enter the naris. It is also interesting to note that Megalobatrachus shows no ‘‘Derotreme’’ characters whatever, although in the older classifications it was included in the Dero- iremata. No. 646] THE PHYLOGENY OF THE CAUDATA 425 The extreme antiquity of the Caudata can be readily seen when an end form, a river adaptation, is found in Oligocene times. This of course puts the origin of the main stocks back at least to the end of the Mesozoic, a conclusion to which the distribution also forces us. The primitive characters appear in widely scattered and rather unrelated forms. "The free prearticular has already been mentioned. A free lachrymal is found in Hynobiide and in an Ambystomid, Rhyacotriton. A postfronto-squamosal arch is found in one group of the Salamandride. A T-shaped parasphenoid 1 is found in an Ambystomid (Dicamptodon) and in a Salamandrid (Tylototriton). Long maxille are found in the two forms just mentioned and in another Salamandrid, Pachytriton. Posteriorly projecting prevomers are found in Amphiuma, in all Salamandride, in some Hy- nobiide (Hynobius, Pachypalaminus), and to a less ex- tent in Dicamptodon. All these are theoretically primitive skull characters of amphibians. Their appearance separately in diverse forms is sufficient indication that the three families Hynobiide, Salamandride, and Ambystomide, while con- taining all the more primitive forms of the order, stand in no direct genetic relationship to each other, but must be derived from a more or less remote common stock which combined the otic apparatus, lachrymal, and pre- articular of Hynobius with the long maxilla, T-shaped parasphenoid, and postfronto-squamosal arch of Tylo- totriton. The evidence of Paleontology, as far as it goes, sup- ports this view. I intend in a later paper to assemble the meager facts regarding fossil salamanders. These facts, it may be here stated, lend no support to the prevalent view that the Proteida are an old, a primitive, or an an- cestral group. The following outline classification indicates the size and position of the modern groups. The genera and 426 THE AMERICAN NATURALIST [Vor. LVI species of the Salamandride are probably not wholly ac- eurate. Future work will perhaps indieate the affinities of Amphiuma, the Proteide, and the Sirenide. Of the larger families, the Hynobiide are entirely Asiatic, the Salamandride are Eurasiatie with four American species, the Plethodontide are American with two species in Europe and four in South America, and the Ambystomide are American with one Asiatic species. As the Northern land masses have been connected with each other during Tertiary times this distribution is not extraordinary, although close resemblance between widely separated species is eloquent testimony as to the antiquity of some of the ‘‘modern’’ forms. Twenty-two of the recognized genera and 105 of the species are restricted to North America, 13 genera and 96 species are Eurasiatie, while three genera are found both in North Ameriea and in some parts of the Old World. Mutabilia Salamandroidea l. Ambystomide .......... 3 genera, 16 species Dicamptodon 2, Rhyacotriton 1, Ambystoma 13 2. Salamandride .......... 7 genera, 37 species. Salamandra 5, Chiog!ossa 1, Ty ototriton 2, Pachytriton $ P eurodeles 3, Triturus 24, ; ; , Salamandrina 1. 9. Plethodontide ber ds +16 genera, 83 species Desmognathus 7, Leurognathus 1, on 11, E aa dipus Bai ina Amphiumoidea (Relationships uncertain, possibly should stand as @ amily under Salamandroidea) 4. Amphiumide n ie cks .... 1 genus, 2 species, Amphiuma 2. No. 646] THE PHYLOGENY OF THE CAUDATA 427 Cryptobranchoidea B. Hye iecore 5 genera, 20 speei Hynobius 15, Po kes duces $. ee 2, Ranodon 1, Batrachuperu s1. 6. Cryptobranchidæ ........ 2 gen , 2 species. RSR A " Cryptobranchus 1. Proteida d uncertain) ee eat are Be ...2 genera, 3 species. Necturus 2, Proteus 1. Lissa Lis EC E 2 kaei 2 speci nl, PESEE 1. Total number of genera 38, ty species 165. LITERATURE CITED Dunn, E. R. 1920. Notes on Two Pacific Coast Ambystomide. Proc. New Eng and Zool. Club, VII, pp. 55-59. Kingsbury, B. F., and Reed, H. D. ea The Columella Auris in Am- phibia. dien Ri. XX, pp. 549 Noble, G. K. 1922. The Phylogeny of e ‘Balientia, Part I. The Oste ology and the de Museulature, their Amie on Classifieation and Phylogeny. Bull. . Mus. Nat. Hist., XLVI, pp. 1-87. Okajima, K. 1908. wen Osteologie ei Onychodacty'us japonicus. Zeitschr, wiss. Zool., XCI, 3, pp. 351- ed, H. D. 1920, The Morphology of the Sound -transmitting Apparatus in Caudate Amphibia and its Phylogenetic Significance. Jour, Morph., 5-375. Smith, L. 1920. The Hyobranchial Apparatus of Spelerpes bislineatus. Jour. Morph., XXXIII, pp. 527-5 Wiedersheim, R. 1877. Der Kopfskelet der Urodelen. Morph. Jahrb., » pp. 35 . Wilder, I. W. 1920. The Urodele Vomer. Anat. Record, XVII, p. 349. Wilder, I. W., and Dunn, E. R. 1920. The Correlation of Lunglessness in Salamanders with a Mountain Brook Habitat. Copeia, 84, pp. 63-68 AGENCIES WHICH GOVERN THE DISTRIBU- TION OF LIFE A. BRAZIER HOWELL PASADENA, CALIF. Tue problems presented by the distribution of plants and animals is a fertile field for investigation. "These problems are essentially ecological in character, for often, perhaps always, the range of a species or genus is dependent upon a number of diverse environmental faetors, some of which are readily apparent, while others are obscure; but always they merit careful study. In investigating and mapping the ranges of living organisms and in following the evolutional tendencies of species in so far as we are able, environment and its in- fluences are of the greatest moment, especially from an ecological standpoint. Botanical subjects may usually be allocated in relation to their surroundings with con- siderably greater ease than can active forms of life, for the former are acted upon only by the agencies to be found in one spot, while the latter may experience not only all the influences operative over several square miles, more or less, of diversified territory, but, in the case of a migratory bird or mammal, will be subject dur- ing a part of the year to environmental factors of which we may know nothing. Whether a species is common or rare in a certain area depends upon its rate of reproduc- tion, which is usually entirely adequate unless new and disturbing influences have been introduced; upon the number of favorable or unfavorable conditions which it encounters, the amount of competition with which it has to contend, and its phylogenetic characters, as to whether it be of a plastic type or one which is senescent and over- specialized: all of which may be summed up in the phrase ‘adaptability to its habitat.’’ 428 No. 646] THE DISTRIBUTION OF LIFE 429 In any one realm, or larger region of the earth’s sur- face, there are various climatic divisions, the chief of which have been named zones, and these stretch across the continent following isotherms, or mean temperature bands, usually, for our purpose, based upon the average amount of heat present during the three chief reproduc- tive months. Zones are divisible into faunal districts, whose bounds are limited by conditions of humidity, pre- cipitation and a few other causes that may be operative over considerable areas. These are further divisible into associations, an almost limitless number of which may be recognized. Thus, we have littoral, riparian or stream bank, palustral or marshy associations, the latter being capable of still narrower subdivision into tule, arrow-head or salt grass associations, and so on with- out end. Associations are sometimes but little considered in parts of the country where climatic conditions are uni- form over a wide extent of territory; but in the moun- tainous parts of the west, where practically every pos- sible local environment from the hottest, most arid des- erts, to arctic-alpine conditions may be encountered within a few miles, the importance of their recognition can hardly be overestimated. In considering the agencies governing the range of a form, the question of temperature is undoubtedly of chief importance as a usual thing, but in some cases physical barriers should be given greater weight, for it need hardly be indicated that it is such directly—and temperature only indirectly if at all—that keep many forms of life from greatly increasing their ranges. In studying such barriers, manner of dispersal may be of much importance. In the case of plants more than of vertebrates (with few exceptions), human agency must now be taken into account, for the activities of man, both intentional and unintentional, are responsible in greater degree for the widespread dissemination of seeds and insects over vast stretches of the earth’s surface than 430 THE AMERICAN NATURALIST [Vor. LVI any other cause. Natural manner of dispersal must also be earefully serutinized as a preliminary step, for what will prove a barrier to the extension of the range of a plant with what I may term unadorned seeds may be inoperative in the ease. of seeds adapted to dispersal by the wind, and again, those whose covering is fitted for adhesion to the coats of mammals will often be still more widely seattered. In the ease of an animal, the first thing to be considered is the life-type to which it belongs, the chief divisions of which are aquatic, fossorial, terrestrial, arboreal and volant types, which are limited in varying degrees by physical barriers. To an aquatic form, land masses are insuperable obstacles, while to many terrestrial species, especially such as live in very arid regions and are totally independent of water, even a large river may prove a delimiting agent. A strip of rocky country or . an extent of arid plain will prevent the spread of such a fossorial mammal as the mole. Arboreal forms are checked by large, treeless areas, and animals which are adapted to a life on the plains will usually shun the for- ests. Volant types are the most independent of physical barriers of all, and to some even wide stretches of ocean are no obstacle, as in the ease of the Pacifice Golden Plover (Charadrius dominicus fulvus), in its annual migrations between Alaska and the Hawaiian Islands. While eoneeding that temperature is the most impor- tant faetor in the distribution of life, the writer is of the opinion that not enough importance has been credited to other agencies. Dr. C. H. Merriam was, I believe, the first to formulate the theory that the northward range of a species is governed by the mean amount of heat present during the season of reproduetion, while the southward range of northern forms is restricted by the mean temperature during the very hottest portion of the year. Isotherms have been determined and our conti- nent plotted and mapped into zones, called Arctic, Hud- sonian, Canadian, Transition, Upper Sonoran, Lower No. 646] THE DISTRIBUTION OF LIFE 431 Sonoran and Tropical, some of which are known by other terms in the eastern part of the country. Roughly, the position of an isotherm, as well as the temperature of a region at other times of the year, depends upon latitude, altitude and distance from the sea. Hence it is that the winter temperature of parts of Montana at a considerable altitude and far from the sea reaches a lower figure than has been recorded on the coasts of the Arctic Ocean. The coldest temperature ever known upon the face of the earth—minus 92 degrees F. in the in- terior of Siberia—is much lower than has ever been found by any of the ‘‘farthest north’’ expeditions. We may safely infer that the degree of winter cold, below a certain point, is largely immaterial, for it makes no difference to a tree whether the thermometer is ten or sixty degrees below zero, nor to the lesser vegetation and many rodents safely protected by a deep blanket of snow. Even to the few species of birds which habitually spend the winter in high latitudes, very low tempera- tures are seldom disastrous, but rather is it due, when numbers perish, to a failure of the food supply during sieet storms or long blizzards. Neither birds nor mam- mals migrate so much because of cold as because their usual foods are not to be obtained in adequate amounts during the winter. Certain forms of life may have to contend, in rela- tively low latitudes and altitudes, with conditions which approximate those to be found much farther north. W. T. Shaw has but just brought to our attention the fact that in eastern Washingtor, where Upper Sonoran con- ditions are the rule, estivation and hibernation of the Townsend Ground Squirrel (Citellus townsendi) are so long continued that this animal enjoys but four months of activity during the year. The squirrels emerge as soon as the first growth starts in the spring, but retire to their burrows for the long sleep when the arid condi- tions of early summer cause a desiccation of their food supply. To a torpid animal in its nest below ground it 432 THE AMERICAN NATURALIST ' [Vor. LVI makes no difference whether it is summer or winter above, and so these squirrels seem to lead an existence closely similar to that of their near kin at the Aretie Circle, but with the probable difference that the northern forms experience an actually greater number of hours of daylight throughout the long aretie summer months. In the plains section of the interior, zonal divisions are acted upon by comparatively few modifying agencies, and their boundaries are rather regular and easily de- fined, but in parts of the three Pacifie Coast states, whose shores are bathed by warm ocean currents, and where the topography is decidedly irregular, the problem of zonal definition is often extremely complicated. In the eoast region of northern California, for instance, there is but slight daily and seasonal change of temperature, and a number of Boreal forms are able to occur there beeause the summers are cool enough for them, while certain Sonoran species are also able to exist because the mean temperature of the breeding season is high enough for their needs. The result is a confusion of zonal in- dices that is extremely puzzling at first glance. To these three widely-recognized zonal factors, when operative in certain regions, should undoubtedly be added character of the coastal sea currents—whether warm or cold—and direction of the prevailing winds. Faunal conditions depend largely upon humidity as well as upon all zonal factors. The chief cause of a hu- mid climate is, of course, ample precipitation, either rather evenly distributed throughout the year, or else supplemented during the drier season by heavy fogs and dense forests to retard evaporation, while a cool climate is often helpful. Precipitation may be largely dependent upon the position of adjacent mountain masses with respect to the prevailing winds, for, as is well known, moisture-laden air is cooled upon contact with an elevated land mass, and precipitation results; but little moisture will then be left in the clouds for rain in the trans-montanic sections. This fact is beautifully No. 646] THE DISTRIBUTION OF LIFE 433 . shown by the humid and heavily forested coast and mountain areas of northwestern Washington, in eontrast to the bare, arid plains east of the Cascade range. Associational temperature is induced by many causes, and although limited in extent it profoundly influences local zonal boundaries. Even associational factors other than temperature may raise barriers to distribution that are insurmountable to many organisms. Insolation, or the relative amounts of sun and shade received by a species in its habitat, is sometimes of paramount importance. This may be influenced by cloudiness, by the amount and density of surrounding vegetation or by the character of the topographical en- vironment. In illustrating this point, we may mention as extremes the bottom of a'deep, narrow, forested gulch, and the top of a warm, bare ridge; the face of a steep north slope, and one facing south. A gully on a north slope may be so situated as never to receive the rays of the sun, while at a certain optimum angle one facing towards the south will receive forty per cent. more solar heat than will a level surface. Hence, zonal boundaries upon the two slope aspects will be found to oceur at very different altitudes. Soil conditions are of great impor- tance in influencing the temperature immediately above its surface, and its character helps to control both the amount of evaporation and the degree of moisture which it is capable of retaining. A light-colored soil is con- siderably cooler, other things being equal, than a dark, rocky one, which will absorb and retain more heat. The importanee of the ehemieal eomposition as well as the mechanical condition, with amount of humus, acid or alkali, in the soil need be no more than mentioned. The temperature of the soil and the atmosphere above it is often greatly influenced by near-by cold mountain streams, and in places zonal boundaries may be de- pressed one or two thousand feet in altitude by this ageney. Large snowbanks and glaciers have a similar. effect, though usually less pronounced or, rather, more 434 THE AMERICAN NATURALIST [Vor. LVI locally restricted. A forest fire or avalanche, by de- stroying ground shade with the consequent raising of the soil temperature, will usually cause an area to grow up to plants and trees of the zone immediately below, to be gradually restored, in future years, to its original zonal status. Base level has its effect, for the foot of a mountain mass rising from a plain five thousand feet in altitude will have lower zonal tendencies than will the five thousand foot level of a mountaix rising from a plain with an elevation of but one thousand feet, because the higher plain accumulates more heat. Similarly, a large mountain mass is less influenced by the conditions which surround it than is an isolated peak. A steep slope will carry a certain zone to a greater height than will a gentle one, because the former will receive, during the day, more of the warm air arising from the lowlands, and the cold air which descends during the night will flow off more rapidly. However, this rule is often nulli- fied by the steep slope being so situated that it receives less sunlight than the more gentle gradient. These points are finely illustrated on most of the mountains of the southwest. Plants and trees of the Transition Zone often flourish on the bottom of a north-facing canyon, while the Sonoran sagebrush extends a couple of thou- sand feet higher upon the steep slopes with southern ex- posures. Protective eover is important to most of the more re- tiring forms of active life, and to such it is not only necessary as a sereen during their daily foragings, but they must have holes into which they may dart at the approach of danger and safe retreats in which to rear their young. To very few vertebrates is the actual char- acter of the soil of great moment, but there are excep- tions, as instanced by the large kangaroo rat, Dipodomys deserti, the front feet of which are so weak that it seems able to burrow only in deposits of æolian or other loose sand, and it is useless to expect to find this species in hard soil. Needless to say, character of food, both gen- No. 646] THE DISTRIBUTION OF LIFE 435 eral and specific, is a powerful determining factor of dis- tribution, and with this should be elassed not only the manner of feeding but the methods employed in secur- ing sustenance. The search for a favorite food item will even, in time, indirectly change a mammal from a ter- restrial to an arboreal type, as it evidently has the tree mouse, Phenacomys longicaudus, of the coasts of Oregon and northern California, which, so far as known, feeds exclusively upon the needles of coniferous trees. The question of enemies, it seems to me, should be given much more weight in distributional problems than it usually receives. This factor may be divided into ac- tive and passive enemies. By the latter term is meant competitive forms, as the more robust growth that chokes out a tender seedling, or an organism which, being more adaptable to a variety of conditions, forges ahead of less plastic forms whose habits are competitive. It is the opinion of the writer that such competition constitutes the real remorseless struggle for existence which most species are obliged to carry on in order to survive, rather than their efforts to elude their active enemies. Al- though these passive enemies are not spectacular and are apparent only after scrutiny by an understanding person, they are, nevertheless, always present and opera- tive. Active enemies may be divided into irritating and ex- terminating types, and in certain sections the former may constitute a formidable barrier to dispersal. Few of the larger parasites directly cause death, but the pres- ence of great quantities of aphis, scales, ticks or intesti- nal worms upon their respective hosts may so handicap a species that it is forced to the wall by the competition of more favored forms. Unusual numbers of horse flies in a mountainous section may so harass stock that they utterly refuse to dwell in such regions. Exterminating agencies may consist of directly pre- daceous organisms, such as carnivores which consume the flesh of their victims, or rodents whose presence in 436 THE AMERICAN NATURALIST [Vor. LVI great numbers seriously interferes with the propagation of certain plants. The overstocking of a range with cattle or the presenee of a vast colony of prairie dogs may actually extirpate certain grasses in those districts, and hordes of some rodents will prevent reforestation in spots because all tree seeds are eaten as fast as pro- duced. Poisonous plants work great havoe among range stock at times, and although the amount of such devasta- tion among wild forms has seldom or never been investi- gated, it is doubtless an appreciable factor. In some regions, bacteria and disease, including the smaller para- sites, play a most important role. The tse-tse fly in por- tions of Africa has rendered it utterly impossible for certain herbivorous mammals to be kept in the infested districts; the Stegomyia mosquito that is instrumental in the spread of yellow fever probably caused the Mayan survivors of this dread disease to abandon the ancient civilization of Yucatan, which was at one time so densely populated, and many ailments, comparatively harmless to white men, who have developed a degree of immunity to them, are largely responsible for the decrease in the numbers within recent years of the more savage peoples. From time to time either totally new bacterial diseases appear or else old ones suddenly aequire new virulence, and throughout the ages, such have undoubtedly killed off certain species from faunal divisions; and it is: not at all improbable that during the course of bacterial evo- lution whole genera, or even families, have been extermi- nated by this agency. It seems advisable to append to the ek paper a chart, or key, to the factors chiefly responsible for the distribution and restriction of the ranges of living forms, but this is submitted with considerable hesitancy: Most of the factors mentioned are so interdependent upon others that it is merely a matter of personal opinion as to which heading they should be placed under. For in- stance, it is impossible to decide whether the effect of a cold mountain stream should better be listed under zonal No. 646] THE DISTRIBUTION OF LIFE 437 or associational conditions, for it is operative in both connections. It should be understood, therefore, that the arrangement is only tentative, and that the list has been made to eonform to the viewpoint of a vertebrate zoologist. FACTORS TO BE CONSIDERED IN THE DISPERSAL OF LIFE Life Types: Active Forms. olant. Sedentary Form Character r ot Habita an dn seeding or reproducing. Direct Physical tan Oceans, Et a ee land forms). s, mountains, ete. Me aquatie forms). oss plains, deserts, Protective cover. Regulation by Temperature: Zonal Latitude. Mean is io during Altitude. reproduetion, Proximity to sea. Mean maximum. cean currents. Mean minimum, Prevailing winds. Delimiting temperatures (as frost to tender species unal. Humidity. cua ptr eation of near mountain masses, if any. "fuccum of near bodies of water, if any. Associational. ree of insolation. Effects of fires and avalanches. Presence of cold streams or glaciers, Topographical situation. Slope a i Slo Base level. Soil. Chemical and mechanical character. 438 THE AMERICAN NATURALIST [Vor. LVI Moisture. d: General and specific character. d habits. Enem Paiaive (competitive forms). Active. Directly predaceous. Poisonous foods Bacteria, protozoa, ete. LITERATURE Allen, J. A. 1892. The oe Distribution of North American Mammals. Bw r. Mus. Nat. Hist., IV, No. 1, Art. 14; 199 Clements, F. E. 1916. Plant Suecession. Carnegie Inst. Wash., Pub. 242. 1920. Plant Indicators. Carnegie Inst. Wash., Pub. 290. Grinnell, 1914. Barriers to Distribution as Regards Birds and Mammals. AMER. Nat. 8 1917. visent ‘Meats of Theories Romig Distributional Control. . Nat., Vol. 51, p. Hall, H. M. t: de nnell, J. 1919. Life-Zone fndicstón in California. Proc. Ca if. Acad. Sci., 4th Ser., IX; 37. Merriam, C. 1890. Results fi a ngage Survey of the San Francisco Mountain d of the Little "TT Arizona. U. S. Dept. grem ` SOR Amer. Fauna, N 1892. The Geographical Distribution of Life in "North America with Special Reference to the Mammalia. Proc. Biol. Soc. Wash., MH: 1898. Life Meg and Crop Zones of the United States. U. S. Dept. . Bull. No. 10. 1899. Pun ‘of a Biological Survey of Mount Shasta, California. U. S. De ept. Agric., North Amer. Fauna, No. 16. Shaw, W. T. 1921. Moisture and Altitude as Factors in Determining the Seasonal Aetivities of the Townsend Ground Squirrel in Washington. Ecology, II; 189. THE TAPEWORM INFECTION IN WASHINGTON TROUT AND ITS RELATED BIOLOGICAL PROBLEMS PROFESSOR NATHAN FASTEN OREGON AGRICULTURAL COLLEGE, CORVALLIS, OREGON Ix the whole realm of nature man is the only creature whose ailments have seriously occupied the attention of experts. Leta disease break out amongst the human fam- ily in some corner of the globe and almost immediately the affliction becomes the target for the trained minds of our ablest pathologists. Not so, however, with the maladies of the lower forms. Man’s only interest in them has been one of selfish exploitation, and he has done little to encourage investigations along any other lines except those which bring him immediate monetary returns. It is, therefore, not at all surprising that we possess such meager and fragmentary knowledge concerning disease amongst the lower animals. | It is almost superfluous to say that this attitude must . change if we are to intelligently conserve the lower crea- tures as natural resources. In the last few years we have been hearing a great deal about the conservation of nat- ural resources, and yet very few of us realize the full meaning of conservation. To my mind real conservation implies a thoroughgoing knowledge of the objects to be conserved, coupled with an intelligent application of the factors controlling their preservation. We must possess more knowledge concerning the diseases of the lower animals because it is of prime importance in all conserva- tion programs, in that it may be helpful in preventing great losses of animals which are beneficial to man. In the state of Washington, as well as in the other states of the Pacific coast, fish afford a natural resource of tremendous importance to the welfare of a large pro- K 439 440 . THE AMERICAN NATURALIST [Vor. LVI portion of the eitizens, and yet eomparatively little is known regarding the diseases which affect these aquatic animals. We become alarmed when the fish begin to diein great numbers, and only then are we in any manner con- cerned with finding out what ails them. During the summer of 1919 it was my good fortune to be chosen by the Washington State Fish Commission as a special investigator for the purpose of studying the parasites of the fish in some of the fresh-water lakes and streams of the state of Washington. Prior to undertak- ing these investigations reports had been coming in to the fish commissioner's office that the fish were dying in the mountain lakes and streams of Kittitas county and, therefore, it seemed advisable to spend most of my time in this region studying the nature and extent of the disease. Itis with this epidemie in partieular that I wish to deal in the present paper. Incidentally, I desire to point out some of the interesting biological problems with which the question is intimately linked up. On arriving in Kittitas county the writer found that the people, especially the sportsmen, were very much dis- turbed about the mortality of their lake trout, for they depended upon these fish to yield them spawn for their county hatcheries. They were particularly distressed about the dying of the trout in Cooper lake, and therefore this lake was the first one which I visited. | Cooper lake is situated in the heart of the Cascade mountains about thirty miles outside of Roslyn. Fig- ures 1-3 show various views of the lake. It is a clear body of water, filled with cut-throat trout. The county game commissioners closed the lake some six years ago in order to obtain a plentiful supply of fish for breeding purposes, and as a result of this the trout have multiplied very rapidly within its waters. For the first few years the results obtained were excellent, but within the last two years the fish commenced to die at an alarming rate, so that all spawning operations had to be abandoned. An examination of the cut-throat trout of this lake No. 646] TAPEWORM INFECTION 441 showed them to be heavily parasitized with larval tape- worms which attack the abdominal eavity. From all ap- pearances these larve somewhat resemble those described by Professor Linton in 1889 for the trout of Yellowstone National Park, and, undoubtedly, belong to the genus Dibothrium or Diphyllobothrium, but are probably of a General view of Cooper : lake. Fig. 2. d a d Goaper lake show =~. racks, a favorite place for the blue her Fic. 3. (— -n of Cooper iake ar shore affording an ideal sting place for fish- aceti birds. 442 THE AMERICAN NATURALIST [Vor. LVI Fig. 4. Tapeworm larva in cyst, x 20. Fic. 5. Numerous free-boring and encysted tapeworm larve, X B. Fic. 6. oe Pepe of acaso larva, x 2í FIG. Head end of tapeworm larva, x 65 different species from Dibothrium cordiceps Leidy, the ones discussed by Linton. According to Professor A. R. Cooper, of the University of Illinois College of Medicine, to whom specimens of the tapeworm larvæ were sent for identification, *' the placing of these larvæ specifically is a matter of the working out of the life histories of the species in question.’’ The larval tapeworms under consideration may be en- No. 646] TAPEWORM INFECTION 443 eysted (Figs. 4 and 5) along the walls of the digestive traet, particularly on the stomach, or they may be found burrowing freely amongst the visceral struetures, or within the surrounding muscular walls. In appearance they are translucent, whitish or yellowish-white organ- isms which may vary from a few millimeters to about twenty millimeters in length. They are long, slender and worm-like in character (Figs. 5 and 6). At the anterior end is the head (Fig. 7), which possesses two lateral slits. This head end is constantly changing its shape in the liv- ing specimens, becoming slender and spear-like at one time and stouter and knob-like at another time. The body proper of the larva may undergo periodie contractions and extensions. Covering its entire outer surface are stiff, bristle-like structures which, at first glance, seem to resemble cilia, but which do not possess any independent motion. Posteriorly the body tapers off into a blunt rounded margin (Fig. 6). The damage done to the fish by these larval — is considerable. In the first place, the fish lose their healthy appearanee, becoming much thinner and paler in hue. The parasitic larve undoubtedly produce injurious toxins which interfere with the proper functions of the host. Then, again, the burrowing habits of these para- sites injure the tissues of the fish, causing them to become mushy. And finally, secondary infections of a serious sort may develop within the injured portions. As.a re- sult of all this damage great numbers of the fish die. The life history of these larval tapeworms is extremely interesting. "Those who are familiar with tapeworm in- feetion know that ordinarily two organisms are necessary for the eompletion of the life history. The adult tape- worm lives in one animal ealled the primary host, where- as the larval tapeworm dwells in another animal called the secondary host. The primary host becomes parasi- tized by eating the infected portions of the secondary host. In the case of the tapeworm under consideration it is quite obvious that the trout acts as the secondary host. 444 THE AMERICAN NATURALIST [ Vou. LVI The primary host, however, is not definitely known. Professor Linton found that in the case of the infection of the trout of Yellowstone Park the white pelican acted as the primary host, and, in the light of this finding, it is quite probable that some similar fish-eating bird is the primary host of the larval tapeworm under discussion. While at Cooper lake a canvass was made of the com- mon fish-eating birds which visit, the lake, and it was found that the blue heron is the most frequent visitor. Since no pelicans are known to come to the lake, I rather strongly suspect that the blue heron acts as the primary host for the larval tapeworms of the trout. If this should prove to be the case then the life history, in all prob- ability, would be as follows: The adult Diphyllobothrium tapeworm develops in the intestinal tract of the blue heron, and when the segments become mature they are periodically passed out with the feces. These mature segments contain large numbers of developing embryos and if they are deposited in a stream or lake the embryos are swallowed by the fish, in which they develop into the larval tapeworms already described. When a blue heron captures one of these infected fish, the larve attach them- selves to the bird’s intestinal wall and shortly develop into adults capable of carrying on the life cycle. My visit to Cooper lake convinced me that it was pure folly to entirely elose down a lake for more than a year or two. In the first place, closing down a lake makes for a rapid inerease of fish so that the available food supply soon becomes inadequate for maintaining all of them, with the result that a fierce struggle for existence ensues, in whieh many of the weaker, but nevertheless desirable, fish are killed off. Even those which survive in the struggle appear to be starved. Secondly, when a lake is closed its shores afford an ideal, undisturbed nest- ing place for such fish-destroying birds as the blue heron, kingfisher and the like. These birds not only destroy . large numbers of fish, but they may be the means of dis- seminating parasitic infections. And lastly, in the light No. 646] TAPEWORM INFECTION 445 of the experience in other states, it is a useless waste of money to depend on the fish in a large natural body of water for spawn, because it is very difficult to control the factors which insure success. Two other mountain lakes were next visited: Lost lake on Roaring creek, near Keechelus, and Fish lake. Lost lake is stocked with eastern brook and cut-throat trout, with the former predominating in much larger numbers. The lake has been closed for several years and was utilized by the county game commissioners as a place for obtaining eastern brook-trout spawn. From this lake seventy-six brook-trout and two cut-throat trout were examined, and with the exception of two brook-trout all the fish were found to be clean and healthy. The two exceptions mentioned were each parasitized with a single larval tapeworm cyst. The situation at Lost lake seemed very striking as well as significant, and it suggested the possibility that per- haps the brook-trout are more resistant and immune to the parasitism of the tapeworm larve. At any rate, this is worth while testing out much more thoroughly. One other point which the trip to Lost lake strength- ened was in regard to what has already been said con- cerning the food supply of a closed lake. The fish in this lake, although they were nearly all healthy, were never- theless very thin. The most prominent parts of them were their heads. In two cases the fish were so hungry that they captured field mice which probably attempted to swim across the lake. These were found partially di- gested within the stomachs of the fish. At Fish lake one hundred and nine trout were caught, mainly of the cut-throat species, and' a careful examina- tion revealed the fact that they were all healthy and clean. There wasn’t a single indication of tapeworm infection. Fish lake was an open body of water and this probably accounts for the healthy state of the fish. When sports- men can get into a stream they are a source of disturb- ance to the blue heron and other fish-eating birds, and. 446 THE AMERICAN NATURALIST [Vor. LVI therefore, these birds are prevented from nesting along the shores, thereby protecting the stream from becoming infected with the tapeworm disease. At the termination of the investigations in Kittitas county, the writer made the following specifie recom- mendations to the county game commissioners: l. Not to close lakes for more than a short time, say a year or two, and only for the purpose of conserv- ing the fish. When a lake is closed for many years the normal multiplication of fish is such.that the food supply within the lake is greatly diminished, resulting in a starvation process. Furthermore, unless adequate watch is maintained, the heron and other fish-destroying birds will live along the shores of these closed lakes and serve as a con- stant source of infection for the fish. 2. Not to depend on the closed lakes for spawn, but in- stead to develop a hatchery or a series of hatch- eries with numerous outdoor ponds where they can place many of the healthy trout from Lost and Fish lakes, which will give them a constant supply of healthy spawn. They will not only save money by such a projeet, but their efforts will not be wasted. After the completion of the above studies the writer ex- amined fish from various places in King county, in which he has found the same larval tapeworm infection. Num- erous cut-throat trout of Klause lake near Snoqualmie falls were examined and found to be heavily parasitized. Also, the silver salmon and the so-called red fish or silver trout (which are nothing more than land-locked sockeye salmon) were found to be heavily infected with the same parasites. The striking thing about the parasitism of these last-named fish was that they were more heavily parasitized than any of the fish previously examined in which the tapeworm larvae were found to dwell. No. 646] TAPEWORM INFECTION 447 The observations recorded in the present paper make it obvious that a good many of our fish and game eultural practices are utterly wasted because we are ignorant of those factors which ought to insure success. What is ur- gently needed in the state of Washington as well as in the neighboring states of the Northwest is a series of ** Biological Surveys ’’ for the purpose of studying and mapping out the various ecological factors of the regions in which fish or game are to be planted. We ought to know a good deal about such faetors as available food supply, oxygen content, temperature variations, pred- atory and parasitic organisms, etc., of a place before any kind of animals or plants are introduced into it. Know- ing these eonditions we ean then intelligently fit each organism into that partieular environment where it will thrive best. But without this knowledge we are simply groping in the dark and are powerless to do any real good. MIGRATIONS AND AFFINITIES OF THE FOSSIL PROBOSCIDEANS OF EU- RASIA, NORTH AND SOUTH AMERICA, AND AFRICA. (SIXTH CONTRIBUTION ON THE EVOLUTION OF THE PROBOSCIDEA ) DR. HENRY F. OSBORN AMERICAN Museum or NATURAL History Dr. Hrxosutcutro Matsumoto, of the Tóhoku Imperial University, Sendai, Japan, has recently been studying the Fayüm collections of primitive proboscideans and hyracoids in The American Museum of Natural History, followed by a visit to the British Museum where he has been making comparisons with the types of these mam- mals, described by Dr. C. W. Andrews in his series of papers beginning in 1901. In 1918 Doctor Matsumoto! published a series of five papers on the elephants, turtles, sirenians, cervids, and bisons of Japan compared with those of India. He pointed out that the Japanese archi- pelago was an integral part of the continent from the be- ginning of the Miocene to the middle of the Pleistocene, and that the period of separation seems to have dated from the recent Pleistocene. Consequently its relations with the animal life of southern China and with the more distant peninsula of India are very close. The ancient Japanese proboscideans are chiefly of three kinds, of which the most numerous are the forest- living stegodonts, closely related in their specific phases to the stegodonts of China, such as the species Stegodon sinensis. There also occurs in the early Pleistocene the 11. **On a New Archetypal Fossil Elephant from Mt. Tomuro, Kaga.’’ 2. '*On a New Fossil Trionyx from Hokkaido." 3. “A Contribution to the Morphology, Paleobiology and Systematic of PE m A COR a New Archetypal Fossil Cervid from the Prov. of Mino. 5. **On Some Fossil Bisontines of Eastern Asia.’’ Sei. Rep. Tóhoku Imp. Univ., See. Ser. (Geology), Vol. III, No. 2, 448 No. 646] THE FOSSIL PROBOSCIDEANS 449 great Loxodon antiquus namadicus, the straight-tusked elephant, which ranged all over southern Eurasia and probably arose originally in the African continent. In the early formations, such as the Middle Pliocene of Tomuro, Kaga, we meet the Elephas aurore, regarded by the author as an intermediate type between the stego- donts of the Upper Pliocene of India and Elephas plani- frons, which in turn is related to the true mammoths (Elephas primigenius) and wandered all over southern Europe in Upper Pliocene time, namely, Bessarabia, . Austria, and southern France. In still earlier deposits, such as the Upper Miocene of Kuji, occurs a mammal whieh the author refers to Stegodon latidens, an ances- tral stegodont of Burma, India. In the Lower Miocene of the Provinee of Mino occurs a form very similar to the Trilophodon angustidens of the Middle Miocene of France, ancestral to all the long-jawed proboseideans. The Stegodon itself is peculiar to India, China, Japan, and the larger islands of the Malayan archipelago, such as Sumatra, Java, and Borneo. The author notes that there is a marked difference between the sexes, so that the stegodonts of each geologic period seem to have re- ceived two specific names, one applied to the female, the other to the male form. Among these couples are S. Cliftii-bombifrons, dating from the Upper Pliocene and from the Lower Pliocene of India; S. ganesa-insignis, dating from the Upper Pliocene sd from the Postplio- cene of the same area; S. sinensis-orientalis, dating from the same strata of China and Japan; S. airawana-tri- gonocephalus from the Postpliocene of Java. This sex dimorphism is very marked, especially in the great dis- parity of size of the upper tusks, which are much smaller and more slender in females than in males. This tusk structure in turn affects the entire form of the head. The Bison occidentalis of Japan seems to spring from the B. sivalensis of the Upper Pliocene of India. It is similar in faet to the bison found in the ancient Pleisto- cene of Kansas, in the basin of the Ohio River, in Alaska, 450 THE AMERICAN NATURALIST [Vor. LVI and in the region of the Yenisei River in Siberia. Ac- cording to the author, in the Transbaikal region the same species oecurs in association with the giant woolly rhinoceros (Diceros antiquitatis), with the hairy mam- moth (Elephas primigenius), and with the heavy-horned bison (Bison crassicornis). Quite a different order of distribution has the remark- able Desmostylus, a sirenian or sea cow peculiar to the coasts of the Pacifie Ocean, first deseribed from the Cali- fornia coast many years ago by Professor Marsh and more recently recorded from Japan. The Japanese species is much more specialized and of larger size than the forms occurring on the Oregon and California coasts, which points to a general migration from east to west, that is, from the Orient to the Pacific coast of North America. From this series of papers we are able to draw up the following table showing the principal distribution of the species of mammals in the descending order of the de- posits in Japan: Postpliocene of Shózu-shima (Sanuki): Stegodon sinen- sis, S. orientalis, Loxodon antiquus namadicus, Bi- son occidentalis, Cervus (Sika) ef. nippon. Upper Pliocene of Ikadachi-mura (Omi): Stegodon si- nensis, S. orientalis, Buffelus sp. Middle Pliocene of Tomuro (Kaga): Elephas aurore. Upper Miocene of Kuji (Hitachi) : Stegodon cf. latidens. Middle Miocene of the Provinces of Teshio, ete.: Des- mostylus japonicus. Lower Miocene of the Province of Mino: Trilophodon cf. angustidens, Teleoceras sp., Amphitragulus mino- énsis. The present researches of Doctor Matsumoto on the rich Fayüm collections of the American and British Museums have enabled him to draw an important dis- tinction in northern Africa between the true forest-liv- ing mastodons, which appear to be directly descended No. 646] THE FOSSIL PROBOSCIDEANS 451 from the genus Paleomastodon of the Fayüm, and the long-jawed mastodons, which appear to be directly de- scended from Phiomia of the Fayüm. This interesting diseovery, whieh was partly anticipated in Doctor An- drews's own papers, enables us to trace the American mastodon far back into Upper Eocene times of northern Egypt. In this connection may be mentioned also a series of five papers? by the present reviewer on the ‘‘ Evolution, Phylogeny, and Classification of the Proboscidea’’ which have appeared successively since 1918. The writer is attempting to give an iconographic revision of the en- tire group of proboseideans, including the progenitors of Africa and Eurasia and the highly developed descend- ants of North and South America, which together make up the most remarkable family history of which we have record. In 1900 Osborn predicted that the source of the mam- malian order of the Proboscidea would probably be dis- covered in Africa. In 1901 Beadnell and Andrews re- vealed, through the Geologieal Survey of Egypt, the rich fauna of the Fayüm, southwest of Cairo, in which were found the remains of three proboscidean genera, named by Andrews Maritherium, Paleomastodon, Phiomia, and confirmed by subsequent exploration and research to be the oldest proboseideans thus far known. Animals simi- lar to Meritherium and Phiomia have since been re- ported by Pilgrim in southern Asia. These animals are ? The first paper in this series is entitled ** A Long-jawed Mastodon Skele- ton from South Dakota and Phylogeny of the Proboscidea,’’ Bull. Geol, Soc. Amer., XXIX, March, 1918; the second paper, ** Evolution, a and Chimifontion of the Probóseldes, '" Amer. Mus. Novitates No, 1, January 31, 1921 (Osborn, 1921. “m the third paper, ‘í First Appearance p the "ror Mastodon in Ameriea," Amer. Mus. Novitates No, 10, June 15, 1921; the fourth r appears in the Bulletin of the Geological Society of America, under the title ** Evolution, Phylogeny, and Classification of the PAPIERA in ; the fifth paper, ‘‘ Adaptive Radiation and Classification of the Proboscidea,’’ was read before the National Academy of Sciences, April 26, 1921. The present is the auk paper. The Ieonographie Type Revision will form one of the Memoirs of the Ameriean Museum of Natural istory. [Vor. LVI THE AMERICAN NATURALIST 452 "IS6T 'eunf "eopIoSoqo.,[ ef} JO uorjerpei 9Apjdupe oq 0} SB 109g] JuIserd Əy} Fumos uwage TDA 3N39304 'sodd1 'enuuwnpane fuaiv]q lissmar) $1s240[ n 4 10 $7114 (1T pu D s] S2407 Ae 7 T S qul KA Ya žo E£. NAI 1 s ^ 320] A oP KKO o = 1 J1N3209I 10 Po 7909 qu t J A and ^ È » vtt da? "i prepa : =, ^ ; / A d f 3 m / 7 (3 Re Tat / ; Ape f I y ae] JINIJOIN dig A AY | Ped [2 LR ! Pa N i : BA I [ / 4 L / A / o lps 2 i o AA Ke 3JN320I ld S. : " GY a 7 Y oy ~X ae Q X vy x 8 6s S (= Os V/s V UNA waosa e | $| VAY [A : ia LO A 6 1N393H yy ja budo] 7 u voafy No. 646] THE FOSSIL PROBOSCIDEANS 453 now found to belong respectively to three distinct lines of the Proboscidea, namely, the meritheres, the true mastodonts, the long-jawed bunomastodonts, as indicated in black on the accompanying diagram. They point, however, to a long antecedent origin and radiation. This is part of the evidence for an ancient adaptive radia- tion proeess by which it now appears that the probos- cideans, like other hoofed mammals, were broken up into several great primary stocks way back in Eocene times, namely: An amphibious stock, adapted to rivers and swamps, of limited migration (= Meritherium, Dinotherium). A mastodont stock, adapted to forests and savannas, of wide migration (= mastodonts, trilophodonts). A stegodon-elephant stock, adapted to southern forests, to grassy plains, to savannas and steppes, of wide migration (— Stegodon, Loxodon, Elephas). These primary stocks gave off from two to six branches each, so that the Proboscidea as a whole are not two branched (i.e., mastodonts and elephants), as formerly supposed, but many branched or polyphyletic. The forest and savanna browsers and the grazers of the plains and steppes were the long distance travelers and from an Afriean or Asiatie center in Eocene times they reached in the Middle and Upper Miocene all the conti- nents of the world except Australia, while the amphibi- ous forms remained in Afriea and southern Eurasia. Certain of these branches, like the true mastodons, are of very great geologie antiquity. Intelligent, independ- ent, well defended, resourceful, adaptive, we find eleven very distinct branches of proboscideans persisting into Upper Pliocene times, five of the least hardy of which became extinct during the colder conditions of the Lower Pleistocene. The known lines of evolution are shaded on the accompanying diagram; the unknown are left in white. The theoretic adaptive radiation may be ex- pressed in a formal classification as follows: 454 THE AMERICAN NATURALIST [Vor. LVI Amphibious and swamp-living stock I. MCERITHERIOIDEA (M«eritheres) 1. Moeritheriinij amphibious or swamp-living forms own in the Upper Oligocene of Africa. IL. DINOTHERIOIDEA (Dinotheres) 2. Dinotheriini,* large amphibious forms frequenting the 4 Ibid. rivers of southern Eurasia throughout the Miocene to the close of the Pliocene. Forest and savanna grazers III. MASTODONTOIDEA (Mastodonts and Bunomasto- nts) MasToDoNTIDAE or “ true mastodonts,” including the sub- families 3. Mastodontine, springing from Palwomastodon of the Oligocene of North Africa, and terminat- ing with Mastodon americanus of the Pleisto- cene forests of North America; grinders lophodont, lacking trefoils. 4. Serridentine,® first known in the Middle Miocene of France and Switzerland, spreading over into India and North America; lacking the trefoils. BUNOMASTODONTIDÆ, the bunomastodonts, springing from orms similar to the Phiomia of North Africa and separating into four main divisions: 6. Longirostrine, typical long-jawed bunomasto- donts arising in North Africa (Phiomia), spreading all over southern Europe, Asia, and North America 5. Notorostrins, a special branch entering the An- dean region of South America and spreading over the South American continent, distin- - guished by the loss of ios lower tusks and the abbreviation of the ja T. pcr imn beaked AERE EE known in the southern United States and north- ern Mexico, with powerful downturned upper and lower tusks. 8. Lon ui short-jawed bunomastodonts, which tate both the true mastodonts and the emend in the abbreviation of the lower jaw and the early loss of the inferior tusks. 3 Herluf Winge, 1906, p. 172, 5It is a question whether this subfamily is nearest the Mastodontide, with whieh its members are generally plaeed by European paleontologists. No. 646] THE FOSSIL PROBOSCIDEANS 455 These animals wandered all over Europe, Asia, and western North America. IV. ELEPHANTOIDEA (the Elephant stock) . Stegodontins, the original members of which were doubtless ancestral to all the higher elephants, persist as an independent branch into the Lower Pleistocene of eastern Asia. 10. Loxodontine, embracing the great African division of the elephants beginning with varieties of the Loxodon antiquus of the Upper Pliocene, which wandered all over southern Eurasia and radiated widely over Africa. 11. Mammontine, including (a) the Southern Mam- moths (Elephas planifrons of India and FE. meridionalis of Europe), from which is derived E. imperator of North America, and (5) the Northern Mammoths, which probably include Æ. columbi and the widespread E. primigenius of the northern steppes; (?) tetradactyl pes. 12. Elephantine, the true elephants (E. indicus of India), which do not appear until the Upper Pleistocene; pentadactyl pes. This twelve-fold branching of the proboscideans is similar to the adaptive radiation which the author has traced in the evolution of the horses, of the rhinoceroses, and of the titanotheres, carrying the fundamental lines of separation back to the Middle Miocene as the most recent date, and to the Middle or Lower Eocene as the most remote date. It will be observed from the diagram (Fig. 1) that the shaded areas represent those probos- eidean phyla of which remains have been discovered. The large unshaded area includes the entire Oligocene, Miocene, and Lower and Middle Pliocene history of the Elephantide which is still unknown but which is likely ` to be revealed at any time by discoveries both in Africa and in central Asia. A very striking fact is that the early member of the Elephantoide, the Elephas plani- frons of the Upper Pliocene of India and the apparent ancestor of the mammoths, is now antedated in geologic time and in its transitional structure by the Elephas aurore (i.e., of the rising sun region) of Japan. BOOKS AND LITERATURE The Conservation of the Wild Life of Canada. By Dr. C. GORDON Hewirt, late Dominion Entomologist and Consulting Zoologist. York: Charles Seribner's Sons, 344 pp., illustrated. vd book was in manuseript before the untimely death of Doetor Hewitt, February, 1920, and has been prepared for publieation by his wife. Mrs. Hewitt has also written a beautiful prefaee whieh ean perhaps be fully appreciated only by those who had the rare good fortune to eount Hewitt as a personal friend. To get the proper perspeetive on this book, one should know that Doetor Hewitt was a zoologist of broad training. Previous to coming to Canada he had worked not only on insects but also on several problems on birds and their control of insect pests. The reeord of his work as Dominion Entomologist from 1909 until his death is a brilliant one. Throughout this period he was frequently consulted regarding various zoological problems which eame before the Advisory Board on Wild Life Protection and in 1916 he was appointed Dominion Consulting Zoologist which broadened his official interest. The work recorded in the book under discussion was done chiefly during the last four years of his life. For so busy a man to undertake a task of this size and to cover the field so well in so short a time is an enviable accomplish- ment. The reading of this book is like a trip to the North Woods, but with a scientist as companion rather than a record-breaking hunter of big game. Although the title might properly include fur-bearing animals and other natural groups, the discussion is chiefly limited to the larger wild mammals and birds of Canada. The information regarding the present distribution and abun- dance of the several species is accumulated from many sources and constitutes a valuable inventory of the remaining but di- minishing resources of the Dominion. As might be expected of one who understands the dangers of promiscuous and ignorant hunting and who appreciates wild life, the dangers and economic loss of unrestricted shooting are constantly set forth, and the re- sults of inadequately controlled slaughter in the United States 456 No. 646] BOOKS AND LITERATURE 457 are used as an ignoble example. It is not too late in Canada to profit by mistakes south of the boundary, and Doetor Hewitt's book should serve as a timely warning. Considerable progress has already been made in establishing government and private reserves in Canada and the record of this movement as given in this book is one of the most valuable features of the work. The author took a lively interest in this movement, and in efforts to conserve the wild life of the Domin- ion he did everything possible, from revising the game laws of the Northwest Territories to the instruction of Boy Scouts in bird protection. Previous to his work the game laws of the Northwest Territories had not been revised for many years, and he succeeded in the diffieult task of getting through a revision that is a great improvement over the former regulations. His successful effort to bring about the Migratory Birds Treaty between the United States and Canada was an accomplishment of high order. There were, of course, other earnest men on both sides of the boundary who assisted in this work, but to the author of this book fell some of the most aggravating and ability-testing tasks. If the full history of this effort is sometime written, Doctor Hewitt’s part will appear as a large one. The discussion of the periodic fluctuations of Canadian fur- bearing animals in Chapter IX is perhaps the best example of scientific method in the book. These fluctuations have long at- tracted the attention of scientific and commercial men and they are here diseussed from abundant data and from a biologieal point of view. This posthumous book is an additional monument to the seien- tifie skill and personal abilities of the author. It should serve as a valuable warning to Canadians and will be of value to readers everywhere in giving a summary of the resources of the Domin- ion in one of its most interesting and economically valuable assets. Beeause of the wide interest in big game it should attraet temporarily or permanently to Canada those who retain a whole- some love for the outdoors. E. F. PHILLIPS SHORTER ARTICLES AND DISCUSSION THE PROBABILITY ESTABLISHED BY A CULTURE OF GIVEN n THAT A MATING e Sages OF RODUCING ONLY DOM NT INDIVIDUALS To distinguish individuals heterozygous from those homozygous for a given dominant factor is a matter of mere inspection when the simplex condition is somatically distinct from the duplex condition, as is the case with the mottling factor in the Adzuki Bean. Generally, however, the degree of dominance is such that a breeding test must be resorted to in order to distinguish these two types. A homozygous dominant will breed true for the character whether selfed or back-crossed to the recessive, whereas a heterozygous individual will give 3 : 1 and 1 : 1 ratios respec- tively when similarly treated. The common breeding practice is to eonsider the parent homozygous when, if selfed or back- erossed, it fails to produce any recessive individuals in a reason- ably large number of offspring. Just what is to be considered an adequately large number of offspring has in the past been determined by the personal judg- ment of the individual investigator, and the diffieulty of obtain- ing offspring in large numbers. There has been no general agreement based on mathematical considerations, probably be- cause large numbers of offspring have not been found necessary in order to distinguish a homozygous dominant from a heterozy- gous parent producing such ratios as 3 : 1 and 1 : 1. The need of a statistieal eriterion of what is an adequately large number of offspring was realized when it beeame neeessary in tetraploid races of the Jimson Weed (Datura Stramonium) to distinguish between matings which should produce only dominant purple offspring and those which should produce a 35 : 1 ratio of purples to whites. In distributions which are so asymmetrical as those given by sampling from the 35 : 1 ratio, we are hardly justified in using the ordinary theory of probable errors. Special tables have, therefore, been computed for use in work under way at the Station for Experimental Evolution. Since other investiga- tors will probably meet with the need for similar criteria, it seems worth while to give tables showing the number of offspring 1 Jour. Hered., 8, 125-131, Fig. 10, 1917. 458 No. 646] SHORTER ARTICLES AND DISCUSSION 459 which should be considered in order to distinguish matings which should give_all dominant individuals from those which may pro- duce recessives, The theory is of course quite simple. It is assumed that the expected ratio of dominant to recessive is known, and is p : q, where p+q=—1. The distribution of the chances of obtaining dominant and recessive individuals in the frequencies n : 0, (n—1l) : 1, (n—2) : 2, ete., when n individuals are grown is (p+ q)”. To ascertain the probability of securing all dominant individuals in a eulture whieh should show a definite ratio of dominant to recessive offspring we have merely to table p" against n. Ifthis value is very small, it is reasonable to assume that in practice a culture of n individuals all of the dominant type represents a parent or parents eapable of produeing only off- spring of the dominant type. Thus, for example, if seeds which should produce dominant and recessive individuals in a 5 : 1 ratio were sown, a culture of 35 all dominant individuals should be obtained only 17 times in 10,000. Hence, if a sowing is made to distinguish between a mating eapable of producing only domi- nants and one which should give recessives in a 5 : 1 ratio, and there results a eulture of 35 individuals all of the dominant type, it is altogether reasonable to assume that the mating in question is ineapable of produeing recessives. Tables have been formed to inelude the 3 : 1 and 1 : 1 ratios familiar in ordinary disomie inheritanee, the 2 : 1 and 8 : 1 ratios found in trisomie inheritanee in the mutant Poinsettia, and the 5 : 1, 11 : 1, and 35 : 1 ratios found in tetraploids in Datura. Some of these ratios are suggested by published data on @nothera Lamarckiana and Primula sinensis, and will prob- ably be found ultimately by those studying other forms. The tables enable one to deeide how large a culture is neces- sary on a probability basis. If it is felt that only 1 chance in 1,000 of the mating being capable of producing a recessive is sufficient evidence that the culture represents only dominants, then, to distinguish a mating which ean produce only dominants from one which should give a 1 : 1 ratio, a culture of at least 10 individuals is necessary. If the 3 : 1 ratio is the one in question, then 24 individuals are necessary ; while if a 35 : 1 ratio is con- sidered, 244 individuals are required. In other words, cultures of 10, 24 and 244 individuals are of equal value in distinguishing matings which should produce only dominants from those which 460 THE AMERICAN NATURALIST [Vor. LVI should give, respeetively, 1 : 1, 3 : 1, and 35 : 1 ratios of domi- nants to recessives. A. F. BLAKESLEE, JOHN BELLING, J. ARTHUR Harris. TABLE I VALUES OF p” FOR 1: 1, 2 : 1, 8 : 1, AND 5 : 1 RATIOS N is 2:1 SI 5:1 N 3:1 5:1 I. .5000 6667 .7500 8333 19 .0042 0313 Bes -2500 .5625 .6944 20 .0032 0261 dus .1250 .2963 .4219 .5787 21 .0024 0217 4... .0625 .1975 .3164 .4823 22 .0018 0181 D. .0313 .1317 2373 .4019 23 .0013 0151 6.3. .0156 .0878 1780 .3349 24 .0010 0126 T. .0078 .0585 .1335 .2791 25 0105 $2. .0039 .0390 .1001 .2326 26 — .0087 9. .0020 .0260 . .0751 .1938 27 — .0073 10... 0010 .0173 .0563 .1615 28 — .0061 IL. — .0116 .0422 .1346 29 — .0051 12.5 — .0077 .0317 .1122 30 — .0042 ge — .0051 .0238 .0935 31 — .0035 14... — .0034 0178 .0779 32 — .0029 IB. — .0023 .0134 .0649 33 — .0024 16.. — .0015 .0100 .0541 34 — .0020 An — .0010 .0075 .0451 35 — .0017 18.. — — .0056 .0376 36 — .0014 TABLE II VALUES OF p^ FOR 8 : 1 RATIO N 0 1 2 3 4 5 | 6 7 8 9 Li .3079 | .2737 | .2433 | .2163 | .1922 .1709 | .1519 | .1350 | .1200 06 Be. E 4 .0843 | .0749 | . .0592 | .0526 | .0468 | .0416 | .0370 | .0329 a. .0292 | .0260 | .0231 | .0205 | .0182 0162 | .0144 | .0128 | .0114 | .0101 (eee z .0080 | .0071 -0056 | .0050 | .0044 | .0039 | .0035 | .0031 D o .0028 | .0025 | .0022 | .0019 | .0017 0015 .0014 | .0012 | .0011 | .0010 TABLE III VALUES OF p? FOR ll : 1 Ratio N 0 1 | 2 3 E 5 6 7 8 9 T5. .4189 | .3840 | .3520 | .3227 711 | .2485 | .2278 | .2088 | .1914 2. .1755 | .1 .1475 | .1352 | .1239 | .1136 | .1041 | .0954 | .0875 | .0802 See -0735 | .0674 | .0618 | .0566 | .0519 | .0476 6 | .04 .0366 36 a. .0308 0282. .0259 | .0237 | .0217 | .0199 | .0183 | .0167 | .0154 | .0141 &... .0129 | .0118 | .0108 | .0099 | . 0083 | .0077 | .0070 | . 0059 A. .0054 | . .0045 | .0042 |. i .0032 | .0029 | .0027 | .0025 [ens .0023 | .0021 | .0019 | .0017 | .0016 | .0015 | .0013 | .0012 | .0011 | .0010 No. 646] SHORTER ARTICLES AND DISCUSSION 461 TABLE IV VALUES OF pn FOR 35 : 1 RATIO | | | N 0 1 2 | 3 4 | b LR 1i Ri 9 CUR 4295 | .4176 | .4060 | .3947 | .3837 | .3731 | 3627 .3526 | .3428 | .3333 4.... .| -3241| .815 .2978 | .2895 | 2815 | | -2737 | .2661 | .2587 | .2515 5.....| -2445 | .2377 | .2311 | .2247 | .2184 | .2124 | .2065 | .2007 | .1952 | .1897 6.....| .1844 | .1793 | .1744 | .1695 E 1558 | .1515 | .1473 | .1432 NI. 1392 | .1353 | .1316 | .1279 | .1244 | .1209 | .1175 1143 | 1111 | .1080 8 1050 | .1021 | .0993 | .0965 |.0938 | 0912 0887 | .0862 | .0838 | .0815 9. 0792 | .0770 | .0749 | .0728 | .0708 | .0688 | .0669 | .0651 | .0632 | .0615 HE. 0598 | .0581 | .056 9 | .0534 | .0519 | .0505 | .0491 | .0477 | .04 11 0451 | .0439 | .0426 | .0414 | .0403 | .0392 | .0381 | .0370 | .0360 | .0350 id ou 0340 | .0331 | .0322 | .0313 | .0304 | .0296 | .0287 | .0279 | .0272 | .0264 13 0257 | .0250 0243 | 0236 |.0229 | .0223 | .0217 | .0211 | .0205 | .0199 hb 0194 | .0188 | .0183 | .0178 | .0173 | .0168 | .01 159 | .0155 | 15 0146 | .0142 | .0138 | .0134 | .0131 | .0127 | .0123 | .0120 | .0117 | .0113 Hu 0110| .0107 | .0104 | .0101 | .0098 | .0096 | .0093 | .0091 | .0088 M 0079 | .0076 .0074 | .0072 | .0070 | .0068 | 0065 Hn 0063 | .0061 | .0059 | .0058 | .0056 | .0055 | .0053 | .0052 | .0050 | .0049 15 55. 7 | .0046 | .0045 | .0044 | .0042 | .0041 | .0040 | -0038 | .0037 20 0036 | .0035 | .0034 | .0033 | .0032 | .0031 | .0030 | .0029 | .0029 | .0028 Me oo 0027 | .0026 | .0025 | .0025 | .0024 | | .0023 | .0022 | .0021 | -0021 22.....| .0020 | .0020 | .0019 | .0019 | .0018 | .0018 | .0017 | .0017 | .0016 | .0016 233 .0015 | .0015 | .0015 | .0014 | .0014 | .0013 | .0013 | .0013 | .0012 | .0012 24.....| .0012 | .0011 | .0011 | .0011 | .0010 | .0010 | .0010 | .0010 | .0009 | .0009 LINKAGE BETWEEN BRACHYSM AND ADHERENCE MAIZE ADHERENCE first appeared in the seeond generation of a braehytie x Boone Co. White hybrid and seemed to be linked closely with normal stature) Subsequent progenies indicated that there was no very close linkage between these characters and possibly none at all? "The relationship of these two inter- esting eharaeters has been tested now in more detail and it seems certain that their genes are located on the same chromo- some. A cross was made between a non-adherent brsehyiie plant and an adherent plant of normal stature, both plants being segregates in the F, of the braehytie-Boone hybrid. The first generation segregated with respect to the brachytic culms, ap- proximately half the plants being of normal stature, but none exhibited a tendency toward adherence. From the behavior of the F, plants it is apparent that the adherent parent of the cross was heterozygous with respect to the brachytie character. 1 Kempton, J. H., ‘‘A Braehytie Variation in Maize,’’ U. S. Dept. of Agri. Bull. 925, Feb., 1921. 2Kempton, J. H., ‘‘Heritable Characters in Maize V. Adherence,’’ Journal of Séredüly, Vol. XI, No. 7, Sept.-Oct., 1920. 462 THE AMERICAN NATURALIST [Vor. LVI Three F, plants of normal stature were self-pollinated and three were baek-erossed on the double reeessive (adherent- braehytie). The six ears were planted separately at Arlington, Virginia, but the resulting F, populations were not as large as eould be desired. The eombined self-pollinated progenies gave the following distribution: | No. Plants | Per Cent. QUNM Nor. | Br. Ad. | Br.-Ad. | Br. | A | Crossover | Q. | | 217 | 91 | 8 | 4 | 23.9 | | F 22.2 + 3.4 | 798 + .06 and the plants of the combined back-crossed progenies are dis- tributed as follows: Nor. | Be, | Ad. | Brad. | Br. | Ad. | Crossover 86 | 188 | 178 | fas | 49.5 | 47.6 | 30.0 + 1.35 These distributions clearly indicate that crossing over between these two factors occurred in from 20 to 30 per cent. of the gametes. Additional evidence of linkage between these characters is afforded by the second generation of a cross between an ad- herent plant of normal stature and a ramose-braehytie plant. The F, of this cross was normal with respect to all three char- acters, and they all reappeared in the progenies of the second generation. Five F, plants were self-pollinated and the result- ing ears planted separately. Unfortunately in most of the F, progenies there is a deficiency of adherent plants and for the combined progenies the departure below the expected 25 per cent. is 7.8 + .87, a deviation too large to be ascribed to chance. Whether this deficiency represents seedling mortality is not known, but at the time the plants were classified many of the progenies contained late plants strikingly smaller and weaker than their mature sisters. Some of these plants consisted of a small cluster of grasslike leaves with inflorescences hardly de- veloped beyond the embryonic stage. Such plants could not be classified with respect to adherence, though in many cases it was possible to determine satisfactorily whether they were ra- mose or brachytic. With respect to these last two characters No. 646] SHORTER ARTICLES AND DISCUSSION 463 the small late plants approximated the familiar 9-3-3-1 group- ing. If the assumption is made that all these late plants were adherent, the percentage of adherent plants in most of the prog- enies would then approximate the expeeted. For the present analysis of the relationship of brachytie and adherent, the low pereentage of adherent plants is not important, sinee the per- eentage of erossovers ean be determined from the ratio of normal to braehytie plants or by the use of Yule's Coeffieient of Association.’ Combining the five progenies the distribution of plants is as follows: NUMBER OF PLANTS | | | | Ad.-Ra. | Small Nor. | Ad. | Hw | Be, |} Ad. Ra, Ad.-Br. | Ra.-Br. | e d Pants 361 | 135 | 117 | 193 | 18 po. qu $ i a PER CENT. Ad. m and Small Plants Ra. | Br. 17.8 + .87 | 23.2 + .92 | 21.9 + .94 | 27.9 + 1.0 PER CENT. OF CROSSOVERS hate. Ad.-Br. Ra.-Br. Q. | % Q. | om Q. | % 37 + .06 | 38.5 £1.8 | .886 + .03 | 16.8 +19 | .05 +.06 | 49.9 + 0.4 . It is seen that the progenies of this hybrid indieate about 17 per eent. of erossing over while the three self-pollinated prog- enies of the other hybrid, involving brachytie and adherent, indicate 22 per cent. and the back crosses 30 per cent. It seems inadvisable to combine the self-pollinated progenies from the two hybrids to arrive at a single figure for the percentage of crossovers since the degree of crossing over between two fac- tors often varies greatly in different progenies. It seems certain from these two hybrids that these two characters are located in the same chromosome separated by a distance varying from 18 to 30 units, thus making a linkage series of brachytic, adher- ent and pericarp color. 3 Yule, G. Udney, **On the Association of Attributes in Statisties,’’ Phil. Trans. Roy. Soc., London, S. A., Vol. 94, pp. 257-319, 1900. 464 THE AMERICAN NATURALIST [Vor. LVI The progenies of the braehytie-adherent-ramose hybrid fur- nish evidence that the ramose character may belong to the same linkage series, though the linkage is rather loose. Although the tassels of ramose plants are much larger than those of normal plants and it seemed not unreasonable to ex- pect adherent-ramose tassels to present a large thickened mass, nothing of the sort was found and the ramose-adherent plants could be separated from the normal-adherent plants only by examining the ears. White and colored seeds were planted separately, but the percentage of the three characters are essentially alike, as is shown by the following figures indieating that all three are in- dependent of one of the aleurone faetors: 95 Adherent 95 Ramose 95 Brachytic White seeds planted ....... 16.4 + 1.65 24.3 + 1.92 31.5 + 2.04 Colored seeds planted...... 18.3 + 1.00 21.1 + 1.07 27.9 + 1.17 Dione. e r 1.9 + 1.93 3.2 + 2.2 3.6 + 2.34 J. H. KEMPTON BUREAU OF PLANT INDUSTRY, U. S. DEPARTMENT OF AGRICULTURE A GENE FOR THE EXTENSION OF BLACK PIGMENT IN DOMESTIC FOWLS! Tue results of recent experiments on the inheritance of plumage colors in fowls indicate that varieties in which black pigment extends to all or nearly all of the plumage (e.g., self black) differ by one dominant autosomal gene from varieties in which black pigment is restricted to the hackle, flight and tail feathers (e.g., Columbian and buff varieties). This gene has been called ‘‘extension of melanie pigment’’ and has been as- signed the symbol E", The evidence is derived from reciprocal crosses between Black Orpington and Columbian pattern (Light Brahma) fowls. Whichever way the cross is made the F, chicks are all black in the down. As adults, the males from the reciprocal crosses are alike. They are black with white-bordered hackles and saddle feathers; white-bordered and splashed or stippled wing coverts and narrow white borders on the upper breast feathers. They resemble fairly typical Dark Brahma or Duckwing males. 1 Contributions in Poultry Geneties, Storrs Agr. Experiment Station. No. 646] SHORTER ARTICLES AND DISCUSSION 465 The females from the reciprocal crosses are unlike in adult plumage. From the cross of Black male by Columbian female, the daughters are self black. From the cross of Columbian male by Black female, the daughters are black with white bor- ders on the feathers of head, hackle, and upper breast. They resemble the pattern known as Birchen When backerossed with Columbian tole the F, males have produced black, Columbian and buff chicks in the ratio of 4: 3: 1; or approximately equal numbers of chicks with extended and restricted black pigment. When crossed with buff females the same F, males have produced chicks in approximately the ratio 2 black: 1 Columbian: 1 buff; again showing equality be- tween the extended and restricted Gasses. The F, black fe- males crossed with buff males have produced equal numbers of black and buff chicks, while the F, Birchen females have pro- duced, when crossed with buff males, black, Columbian and buff chicks in a ratio approximating 2: 1: 1. The ratios as quoted above have all been obtained and will be reported in full when the adult classifications have been completed. All of these erosses represent matings of fowls heterozygous in extension (E"e") with fowls recessive in extension (e"e"). The expecta- tion is equal numbers of black (extended) and non-black (re- stricted) chicks. The experimental numbers at present are 99 black (E"): 98 non-black (Columbian or buff e"), A clear monohybrid segregation is evident between extension (E") and restriction (e") of black pigment. The above results are all explained on the following hypothe- ses: l. The black fowls have the dominant allelomorph of an auto- somal gene (E") which determines the extension of black pig- ment to all parts of the plumage. The reeessive allelomorph (e") of this gene is present in the Columbian and buff fowls. 2. The black fowls contain the recessive allelomorph of the dominant sex-linked gene S (silver). This dominant gene in- hibits the produetion of buff ground eolor and eauses the pro- duetion of silver or white ground eolor;"? it is known to be present in Columbian fowls, while the recessive allelomorph characterizes the buffs. As regards these two genes the blacks used in these experiments have proved to be E"E"ss in composi- tion, the Columbians e"e"SS (male) or e"e"S-(female), and the 1 Sturtevant, A. H., 1912, Jour. Ezp. Zool., 12: 499—518, 2 Dunn, L. C., 1922, Amer. NaT., 56: 242-255. 466 THE AMERICAN NATURALIST [Vor. LVI buffs e"e"ss, Blacks are therefore genetically buffs with an epistatie gene for the complete extension of black pigment. 3. Extension of black is incompletely epistatie over silver so that in fowls of the genotype E"e"Ss (male) or E"e"S-(fe- male) silver appears in certain parts of the plumage, produ- eing a pattern like that of the Dark Brahma. Collateral evidence indicates that the gene for extension of black pigment (or one with similar effeets) is present in Barred Plymouth Rocks and White Plymouth Rocks (as a eryptomere) ; and that it is absent in Columbian and buff varieties and in Rhode Island Reds. Hurst? was probably dealing with the same gene in his erosses between Black Hamburgs and Buff Cochins, since F, from this cross consisted of all black chicks, while in F, black and buff chicks occurred in the proportions of 88 black: 31 buff. In in- terpreting the results of this eross Morgan* states that either = or two pairs of factors may be involved; which is right **eould only be determined by an F, ratio." Yet the F, ratio is given by Hurst (p. 138) and is surely a sufficiently close ap- proach to a monohybrid ratio. The ratios obtained in our ex- periments agree throughout with a mono-factorial interpreta- tion. It is believed that this gene will be found to characterize many color varieties of fowls in which black, either as a self color or as a component of a pattern extends to all or nearly all of the plumage. Concerning its origin no direct evidence can be offered at present. Its occurrence as a discrete unit indicates, however, that its origin was discontinuous and that black varie- ties probably had their genesis in mutation rather than in selee- tion of particolored types toward black. L. C. DuNN THE EFFECTS OF SO-CALLED CONJUGATION IN SHELLED RHIZOPODS : THE phenomenon of conjugation in the Protozoa is regarded as the forerunner of sexual reproduction in the higher animals, 3 Hurst, C. C., 1905, Reports to the Evo. Comm., II, pp. 138-39. . * Morgan, T. H., 1919, Publ. Carnegie Inst. No. 285, p. 24. The experimental work on this problem was carried on in the Zoological Laboratory of the Johns Hopkins University. I wish to thank Dr. H. S. Jennings, of that institution, for suggesting the problem and for his aid in pursuing the investigations. No. 646] SHORTER ARTICLES AND DISCUSSION 467 in which there is a union of a sperm and an egg. This latter proeess is fundamental in the life of at least all the higher Metazoa. By this union the race is perpetuated and hereditary characters are intermingled. It would seem that studies of a similar process in the unicellular forms might throw light upon the basic relations of the sexual preron to the life of proto- plasm. Various investigators have shown that conjugation is a rela- tively common occurrence among the more complex Protozoa, such as the Paramecium, and that hereditary characters are intermingled in this way. During conjugation two Paramecia fuse by their oral surfaces and there is an interchange of nu- clear material between the individuals. The latter then sepa- rate and reproduction by division occurs. The races resulting from such divisions show the effects of modifications due to hereditary characters coming from both the conjugating indi- viduals. In the ease of the more primitive Protozoa, the Ameba and other Rhizopods, not a great deal is known. It has been ob- served that sometimes two individuals unite, but it is uncertain whether or not true conjugation occurs with interchange of nuelear material and subsequent modification of the offspring. If sueh proves to be the ease, we shall have shown that the phe- nomena of sex are found in the very lowest animals, and are of general fundamental importance in the life process. The present investigation was undertaken in the effort to throw some light on this question. An attempt was made to test the matter by indueing eonjugation between various indi- vidual Rhizopods and noting if any inherited differenees arose from these unions. (A eytologieal study of the behavior of the nuelei of sueh individuals must be made before the evidence ean be fully weighed.) A shelled Rhizopod, Difflugia corona, was used, since the shell exhibits marked characteristics which vary among the different races of the species. The ordinary shellless Amceba presents so few characters of a permanent na- ture that it is unsuited for a study such as this. The Dif- flugia is an Amoba which builds from microscopic sand grains a shell shaped much like an old-fashioned soap kettle with the legs represented by spines projecting from the rounded aboral surface. The animal lives inside this shell and thrusts its pseudopodia from the oral opening. It reproduces by division, as other Protozoa, half the body being extruded from the oral 468 THE AMERICAN NATURALIST [Vor. LVI opening of the shell and a new shell being formed over the ex- posed part of the body, the two shells, old and new, lying mouth to mouth. The body then divides and the new individual moves away to lead a separate existence. At times two Difflugias have been observed attached mouth to mouth and have been thought to be conjugating. When di- viding the new shell is much lighter in eolor than the old and in the eases supposed to be eonjugating both shells were of the same shade, that is both dark, so that the phenomenon did not seem to be that of division. It was this attachment or so-called conjugation which I endeavored to investigate. It was found by Dr. Jennings (not published as far as I am aware) that two Difflugias eould sometimes be made to attach themselves together by keeping them in a drop of distilled water for a few hours. At his suggestion, I used this method, endeavoring to hasten the proeess somewhat by pushing the individuals together by means of a fine glass rod. In several cases, the members of the pair became quite firmly united and remained so for some time. "These pairs I then put into a drop of culture medium: in the concavity of a hollow ground slide. In some eases such individuals never separated but died. In several instanees they did separate and lived and began to re- produee by division. Briefly, my procedure then was this. The number of spines varies in different races of Difflugia, sometimes averaging 2 or 3, sometimes as high as 5 or 6. The size of the shell also varies. I used the diameter across the widest part of the shell. I se- lected, for example, a small individual with a few spines and kept it under observation on a hollow ground slide in culture me- dium until it had produced an offspring by division. (The eul- ture medium consisted of tap water containing the microscopic débris washed from the leaves of Elodea.) I then isolated the offspring and allowed it to produce a race, keeping each indi- vidual separate on a slide, and measuring and counting the number of spines of each. The original Difflugia I tried to mate, so to speak, with a large Difflugia with many spines, which had previously produced an offspring which I had iso- lated and allowed to found a race. If the large and small Dif- flugias became attached, supposedly conjugating, and were suc- cessfully separated, I then isolated each one on a slide and al- lowed each to start a line of progeny. I then had 4 lines going, one from each Difflugia before conjugation and one from No. 646] SHORTER ARTICLES AND DISCUSSION 469 each after conjugation with the other. The following diagrams may serve to make my meaning clear. G. 20. 55 lsp ¥9P ` osR sp Biro SP SP Fag ad : D. 3 70 D.38 O26 D? RO 3 D O-O ‘ D.33 p^ v ^^ qd `q d » a bi £ d € f| s h Fb 107| 104 79). 1/2 102| 196 2 : € ox : yt 222] 36} 366 30| 298] 363) 32 Ave.No.. i ; Sp. 277] 239] "^ #28 253) ajas 4 293| wr] vo Diagrams illustrating the manipulation of Difflugia in the experiments with so-ealled conjugation and the results attained in three cases cles at the top cf each diagram represent the individual Difflugias which became attached to each other, D., diameter- Sp., spines. ind., individuals. Ave., average In each figure the circles at the top represent the Difflugia which were paired. For example, in A an individual with a diameter of 29 units founded a line a and then was paired with a Difflugia with 4 spines with a diameter of 33. This had previ- ously started line d. After the separation of the Difflugias, they founded lines b and c respectively. The dotted cross lines represent the influence which might be expected to come from the other member of the pair if there were actual conjugation. Line a was carried on until 103 individuals were produced, the average number of spines being 2.79 and the diameter 29.6 units. Line b, started after the attachment of the progenitor : to the other Difflugia, was carried for 107 individuals, with an average of 2.85 spines and a diameter of 29.2. The diagrams show the results gained from earrying out the lines in the other experiments. Unfortunately line i died before producing any individuals and line h produced only 7 before dying out. By eomparing the four lines in each experiment with each other, evidence may be gained relative to whether or not, dur- ing the supposed conjugation, the Difflugias exerted an effect upon each other suffieient to cause modifieation of the diameter and number of spines of the offspring in the direetion of the line started by each before the attachment. That is, for ex- ample, line b should be more like line d than is line a, and line c should be more like a than is line d, and so on throughout the experiments. 470 THE AMERICAN NATURALIST [Vor. LVI A study of the data set forth in the diagrams will show that there is praetieally no modifieation of any line springing from a Difflugia, after the supposed eonjugation, in the direetion of the characters of the line founded by the other member of the pair before the union. Any slight changes seeming to show such modification are offset by just as marked variations away from that line. Comparing line g with h, it would seem that there is a marked modifieation of g in the direetion of line e. It must be noted, however, that line h consists of only 7 indi- viduals, so that the average is untrustworthy. In no other ease is there any signifieant leaning toward the other line, although there are very slight tendeneies in that way in the spine num- bers in experiment A. Line a has 2.79 spines, b 2.85, c 4.05 and d 4.22. On the other hand, in experiment B the spine num- ber in line e is 3.58 and in f only 2.12, to be compared with 4 inline A. It would be expected that line f would show a greater number of spines than line e. In experiment C line k shows a slight increase in spine number instead of the expected de- crease. In general there are no apparent modifications of the offspring as the result of the pairing. The experiments are open, of course, to the criticism that the attachment between the Difflugias was brought about by keeping the individuals in distilled water to bring them to the state of partial starvation whieh seems usually to be the fore- runner of eonjugation in the Protozoa. It is possible that under these somewhat unnatural eonditions, the preliminary steps incident to conjugation would be inaugurated but that the proeess would stop before eompletion. In answer to this, I ean only state that the Difflugias remained attached from 12 to 24 hours, thus allowing suffieient time for nuelear changes to have oeeurred. They became separated only when placed in eultures eontaining food. o summarize then, as far as it is safe to base conclusions on the results of this limited number of experiments, it seems that the offspring of the Difflugias were not influenced by the attaehment or so-ealled eonjugation of the parents. From this fact it appears probable that this phenomenon of attachment sometimes observed in the shelled Rhizopods is not a true con- jugation and that there is no interehange of nuclear material between the individuals taking part in it. E. P. CHURCHILL, JR. UNIVERSITY OF SOUTH DAKOTA No. 646] SHORTER ARTICLES AND DISCUSSION 471 ON chi oe SEXUALLY MATURE PLATYNEREIS EGALOPS FROM EGGS I THE literature affords so few cases of marine animals reared under laboratory conditions that the writer ventures to com- munieate his successful attempts to carry through to sexual maturity the nereid, Platynereis megalops, from eggs laid in the laboratory. This work had its origin in a suggestion made by Dr. F. R. Lillie in 1911 that the capacity for cross fertilization between Nereis limbata and Platynereis megalops be tested. At that time, however, since we knew so little of the life history of these forms, we felt that it was necessary to get all data possible on each life history in order to have a standard of comparison for the life history of the hybrids. So far all efforts to cross these nereids have failed. The difference in the breeding habits of Nereis and Platynereis is so striking that this alone might ac- count for the failure of cross fertilization. Nereis sheds eggs into the sea-water where fertilization takes place; Platynereis lays in- seminated eggs soon after copulation. However, this very differ- ence is calculated to enhance the interest attaching to the cross fertilization. It might be possible to study the inheritance of the egg-laying reactions. In addition, early observations revealed that the young Platynereis megalops closely resemble Nereis dumerilii. Since, as is well known, Nereis dumerilii has a com- plex life history, we felt that the life history of Platynereis might well repay study for its own sake. II PLATYNEREIS MEGALOPS REARED UNDER LABORATORY CONDITIONS TO SEXUAL MATURITY The writer has found that it is possible to rear Platynereis megalops to sexual maturity under laboratory eonditions. This was first accomplished in 1913-1914, repeated in 1920-1921, and again in 1921-1922. The results may be briefly recounted. Methods Males and females eaught with a hand net in the evenings of the July and the August full moon are kept in separate dishes. In the laboratory as shortly after capture as possible a male and 472 THE AMERICAN NATURALIST [Vou. LVI a female are placed in a finger bowl of clean sea-water. After copulation and egg-laying the animals are removed from the finger bowl. After the jelly has been extruded by the eggs, the supernatant sea-water is poured off, leaving the eggs in the mat of jelly stuck to bottom of the bowl. At first cleavage, which is invariably one hundred per cent., the jelly-mass is gently broken up and the eggs equally distributed among seven to ten finger bowls of elean sea-water. Early the next morning the sea-water is changed. At this time all eggs that possess fewer or more than four oil drops, one in each maeromere, are discarded. Only those larve that possess four oil drops evenly distributed among the four macromeres give rise to normal swimming larve. As the trochophores rise to the surface in each dish they are pipetted off. Trochophores that fail to swim at the surface in twenty-four hours laek the viability of those that rise earlier. The young larve are kept in subdued light a few in each dish because they tend to aggregate in such dense masses that many . die off. This tendency to collect in one spot makes it easy to change the water and thus avoid too great rise in temperature, which is fatal to the animals. The larve will reach the stage of three swimming segments without the addition of food. en the segments appear, the larve must now be baid very carefully in order that food may be given at the proper time. The criterion for the initial feeding is the complete dis- appearance of the oil drops from the entoderm cells. In the eggs of both Nereis and Platynereis there is at the time of fertilization a girdle of some eighteen to twenty-two oil drops in the equatorial zone. These oil drops in the maturation stages following insemination move to the vegetative pole. During cleavage the number of oil drops is reduced to four large glob- ules which normally are distributed to the cells of the gut. Be- ginning with the third or fourth day after laying, the oil in the gut cells of the larvæ begins to form an emulsion of smaller and smaller drops. It is thus possible to follow the history of the oil drops very fully in these ereatures that make veritable living test tubes in a fat-digestion experiment. If food is given the worms before the oil has been completely-used, they are killed in large numbers. On the other hand, food must not be withheld too long after the disappearanee of the oil. The first feeding consists of ten c.c. of a diatom culture known by previous ex- amination under the microscope to be free of metazoa or larve, No. 646] SHORTER ARTICLES AND DISCUSSION 473 strained through three thicknesses of bolting silk of very fine mesh. As the larve add segments more food is given. When the larve build their tubes both food and mud are added until the bottom of the dish is well covered. The method of preparing the diatom culture may now be considered. In 1911 I procured food according to the method deseribed by Hempelmann in his study of the life history of Nereis dumerilii at Naples. This method consists prineipally in seraping the: growths from the live tables. The bottom and sides of aquaria under running sea-water frequently show a felt-like growth of diatoms and protozoa. Serapings from such aquaria suspended in sea-water will give food suffieient to keep a few young worms of both Nereis limbata and Platynereis. The method does not allow the rearing of any large number of worms. Similarly, at- tempts at a pure eulture of diatoms gave poor results. In 1913 through the kindness of Dr. Caswell Grave I procured a remarkably fine eulture of diatoms from Beaufort, N.C. With this, the first sexually mature worms were obtained. But obvi- - ously, Platynereis at Woods Hole must live on food got in Woods Hole waters. I, therefore, made various attempts to get an ade- quate diatom culture from the immediate vicinity. The success- ful method follows. At the beginning of the season mud is taken from Eel Pond, near-by flats, or scraped from eel grass, together with animal and plant life. This is placed in jars with the addition of an equal volume of sea-water. The jars are then covered with glass plates and set aside in subdued light. In a day or so all metazoa— worms, crustacea, ascidians, ete.—are dead. After a period of putrefaction, the culture purifies itself and a rich growth of diatoms begins. For young worms a suspension of diatoms strained through several thicknesses of bolting silk is used. The diatoms for this purpose are previously examined under the microscope, one e.c. at a time; usually no metazoa are found. The suspension is then made up in filtered sea-water. As the larve increase in size and vigor food is added in greater quantities. A brief summary of the three cultures of Platynereis megalops reared to sexual maturity may now be given. The Larval Cultures The 1911 and 1912 larve were not kept after September first. The 1913 Culture.—During August, 1913, from eggs laid in 474 THE AMERICAN NATURALIST [Vor. LVI the laboratory by twenty females over 100,000 larve were reared. On the eve of my departure from Woods Hole the larve were all carefully removed from their tubes and placed in half-gallon Mason jars. Each jar contained 500 ¢.c. of the rich Beaufort diatom culture and sea-water to within ten centimeters of the top. The jars were then tightly covered and set aside. After the worms had had time to build new tubes, the jars were shipped -to Washington, D. C. The worms at this time averaged about 10 mm. in length. Early in June, 1914, these worms were shipped from Washington to Woods Hole. Relatively few survived the journey. The largest worms (females), brought carefully in a hand bag, died before the journey was half made. Since during the winter hundreds of these worms had been killed periodically for future study, the number left from the 1913 culture had been greatly reduced. Some animals of this culture were carried through the summer of 1914. They never reached sexual matu- rity. They were taken back to Washington at the end of 1914. In 1915 they were brought back to Woods Hole and returned to Washington that fall. During this period they still showed no change. . In Washington the animals were kept in the clamped jars with- out any change of water or additions of distilled water. One culture was kept in a battery jar covered with a glass plate. Nor was any addition ever made to this culture. The jars were kept at room temperature in subdued light. To avoid contamination worms removed for study were from the culture in the battery jar only. After observing the worm I never replaced it, but killed it for in toto mount or sectioning. In 1917 several very fine cultures of worms were started, but they died in transit to Washington. In 1918 and 1919 no worms were reared. The 1920 Culture—In August, 1920, very beautiful cultures of about 50,000 larve were started: unfortunately, the majority of these died very suddenly late in August. About a thousand worms survived. These were distributed among twelve dishes with food and left over winter in the heated laboratory at Woods Hole. The dishes were left covered with glass plates exposed to north light. No.change was made in the water or any additional food given during the period September 1, 1920, to June 1, 1921. On May 17, 1921, about 200 worms of different sizes were found in the dishes. Of these some were preserved from time to time. No. 646] SHORTER ARTICLES AND DISCUSSION 475 The history of the others shows that the first female with ripe eggs appeared June 5. She was discovered slowly crawling around the dish near the surface of the water. In color and in form this animal resembled the females collected during the breeding season. In size she was rather below the average and somewhat more sluggish. The eggs in size, color and form were identical with the eggs got from animals captured during the breeding season. Subsequently, mature females were found at intervals through the summer. No males appeared until late in June. Eventually, thirty-two females and twelve males fully mature were got from this 1920 culture. Males got from the sea copulate with females reared in the laboratory ; such females lay normal eggs that give rise to larve of a high degree of viability. Males reared in the laboratory copulate with females taken from the sea. The eggs are perfectly normal. On only one occasion did I find a male and female from this 1920 culture sexually mature at the same time. They copu- lated in normal fashion. The eggs laid were normal in every respect and gave rise to larve that I kept for two weeks before discarding. These larvæ could not be distinguished from larve resulting from eggs laid by animals taken from the sea. The 1921 Culture.—At this writing only two mature animals (females) have appeared—one May 1 and one May 6, 1922." The Rate of Growth Some idea of the rate of growth in these worms may be ob- tained from data collected from the 1913 culture. This was the only culture on which I had the opportunity to make continuous observations. Life History Observations so far made on eultures of Platynereis megalops reared in the laboratory from eggs laid by animals taken from the sea do not reveal any indication of a sexually mature inter- mediate form. So far, all eggs obtained appear to be identical with those got from animals in nature. This would seem to sug- gest that the life history of Platynereis is simple—without the complexity of form and sexual condition found in Nereis dumer- ilii, which Platynereis so closely resembles. It must be clearly 1 Since the above was written, 23 animals have reached maturity—17 females and 5 males. One reason for this sex ratio is that the males have difficulty in getting out of their tubes; their mortality is therefore high. ® 476 THE AMERICAN NATURALIST [Vor. LVI stated, however, that on this point the observations so far made . are not eonelusive. In order to determine fully that the eggs laid by worms in the cultures in the laboratory are from the same worms started in the eulture and not from an intermediate form and are the only eggs laid, it would be necessary to make continu- ous observations on isolated worms. So far it has not been feasible to do this, since it would mean praetieally eontinuous residence at Woods Hole through the winter. TABLE I RATE OF GROWTH OF PLATYNEREIS MEGALOPS FROM EcGGs LAID ON THE VENING OF JuLY: 21, 1913 Segmen ate pelo Mode Length in Mm. July 28-1913 4 co eser Vs 3 July 29. 2 surdis (Nodo detit 4 to 5 July 280. tenet as oy wins d s 5 July Sb 075. ay ren hen ERR oe 6 Aarons Io | 75 V re 10 Aupub 0 eae ades 16 2 Anpust 8 505. ey ee as 22 4 Amt 48 o aa ay wale 26 5 PROGRES oe eee 24 T a 7 | Wo AG INIS ULT E RM 14 Oet 305. 46v VPE Cees 18 Nov. dU c CLAVES OC US 20 Dee. Ac Sieg ks Cie P B 25 Jan, 10, 1814... s ers 30 Feb. B OO oie a 33 NAMES 1o eat gan tee 40-45 April E-E 0 Nata eee ees 40-50 May 8 — Sa E ee ees 50-60 May NR o IR Mh ee er aes 40-50 On the other hand, it is just barely possible that in a state of nature the life history is more complex than in the laboratory cultures. Under operation of changes in such factors as density of the sea-water, food, and temperature, the life history of the . worms may be modified. That this possibility deserves some con- ‘sideration we may conclude from the sex ratio, if such meagre data will allow. In the laboratory eultures females appeared first in all three years and they outnumber the males. In nature just the reverse is true. Whatever our conclusions as to the interpretation of these ob- servations, it seems to the writer that the life history of this interesting nereid is worthy of further study. No.646] SHORTER ARTICLES AND DISCUSSION 477 A Comparison with Other Forms The method used for rearing sexually mature Platynereis from | - the fertilized egg has been used to rear other worms through to the adult stage: namely, Pectinaria gouldii, Diopatra, Nereis limbata, and Chetopterus. In all cases the worms were reared from eggs cut out of the females and inseminated in sea-water. In no ease were the worms kept beyond September 15 (from one to three months). Though it is usually stated that artificial in- semination of Diopatra eggs is not possible, every attempt made by the writer in 1911, 1912, 1913, 1914 and 1915 was successful. There is one danger to avoid with these eggs—initiation of de- velopment by mechanical shock. The worms reared from Di- opatra eggs are if anything more hardy than those of Platy- nereis. In 1913 I reared Diopatra in a watch glass to a length of four centimeters. Pectinaria gouldii are likewise readily reared from eggs in- seminated in the laboratory. These eggs are extremely beautiful, small, and almost wholly transparent. They are easy to handle. I have found them the best eggs in my experience for study under.high power (oil immersion lenses). The specimens used were from the Eel Pond and are normally smaller than Pectinaria found outside of Eel Pond. They are infested with a distome and an interesting eiliate; the latter I did not find in the larger speeimens (1911). This, if it be generally true, together with the size of the Eel Pond specimens makes an interesting ease from the point of view of ecology. Among the shed spermatozoa of Pectinaria are many in bundles that break up after a short time in the sea-water. In addition to these one ean always get bundles of spermatoeytes, immature sperm, ete., by puncturing the body wall. It is a very excellent form to use for the study of eytoplasmie inelusions: it is pos- sible to get the whole history of the sperm on one slide. My objeet in studying these ova was to try to learn if size, opacity, and yolk influence the ease with which the animals ean be reared under laboratory eonditions. I found no correlation. Thus, the egg of Platynereis is almost transparent; it measures 180-200 u. Nereis egg has more color and measures about 100 y. The Nereis egg is the hardest of all to earry through. The egg of Pectinaria is small and almost wholly transparent. It is readily reared. The Chetopterus egg has more color than that of Nereis and is smaller. It is easier to rear than the egg of Pectinaria. The Diopatra egg is wholly opaque; it is the largest 478 THE AMERICAN NATURALIST [Vor. LVI of the five eggs and perhaps the easiest to rear. The eggs may be arranged according to size, depth of color and ease with which they may be reared as follows: Size Depth of Color Ease of Rearing Diopatra Diopatra Diopatra P'atynereis Chetopterus P d Nereis Nereis Chetop Chetopterus P atynereis sae ahaa Pectinaria ectinaria Nereis As in the ease of Platynereis the essential point in rearing these annelids is to give them food at just the right time in the larval stage. This time varies somewhat with each form. Briefly, food must not be given before the yolk and oil are wholly used up. One needs but to watch the larve, note the disappearance of the oil from the gut, and then add diatoms. E. E. Just HOWARD UNIVERSITY . REFERENCES Hempelmann, F. 'll. Zur Naturgeschichte von Nereis dumerilii Aud. et Edw. Zoologica, ns 25, Lief 1 (Heft 62). Just, E. 14, a ae of Platynereis megalops at Woods Hole, Mass. Biol. Bull. Lillie, F. R. and das A E. '13. Breeding Habits of the Heteronereis Form of Nereis limbata at Woods Hole, Mass. Biol. Bull., AN OBSERVATION ON THE ‘“‘CLUSTER-FORMATION” OF THE SPERMS OF CHITON + WHILE engaged with an inquiry into the natural history of the chitons, in 1918,? I several times made an observation which may have a bearing on the significance of sperm-clusters, and on the mechanism of their formation. The matter could not at the time be adequately investigated, but since I shall not soon be in a position to examine it further my observations are here related for what they may be worth. The species concerned is Chiton tuberculatus Linn., an intertidal form quite abundant at Bermuda. It i is PAHPIREN to note, first, certain features of the breeding process, which seems to me to have heretofore been 1 Contributions from the Bermuda Biologieal Station for Research. No. 119. 2 The corrected proof and manuscript of this artiele were returned to the publisher Aug. 18, 1920; but the corrected article was accidentally taken out of type in the office of the printers. The author has now re- written the paper. E. L. M. No. 646] SHORTER ARTICLES AND DISCUSSION 479 somewhat misunderstood. In another connection I shall de- seribe several aspects of the reproductive activities of these ani- mals, the present remarks having to do merely with the act of feeundation. . Although it has commonly been held that the liberation of eggs by a female chiton is due to the reception of spermatic fluid diffusing into her respiratory water-currents from a near- by male, the proeess of fertilization would appear in faet to be initiated in a quite different manner. Stated briefly, the pres- enee of one or more neighboring females serves in some way to aetivate the diseharge of sperm by the males, the spermatie substances secondarily inducing the liberation of eggs. Nor- mally this occurs only at those periods when the flow of the tide begins just before sunrise, the shedding of the genital products commencing as the chitons become covered by the sea. The discharge of sperm can, however, be induced artificially at certain times, in the laboratory, even a month or more before the eggs are matured. A method which several times yielded this result consisted in keeping some male chitons in a damp, darkened vessel for about 14 hours, then covering them with sea water and admitting light. It should be noted here that C. tuberculatus is an animal nicely fitted for observations of this kind, because the differential pinkish tint of the soft tissues of the females permits the quick and accurate identification of sex.* In May, a month before ripe eggs are seen, it was noticed that when sperm diffusing from a male, in a glass dish, was taken up between the etenidia of a female, it issued from the posterior ends of the etenidial channels in an altered state, for the sperm-stream was then seen to eontain numerous aggluti- nated masses of active sperms, which persisted in sea water for at least half an hour. During natural feeundation, however, no sperm-balls are formed. The thick glutinous stream of spermatozoa passes under the girdle of a female, is somewhat diluted with sea water 3 That the discharge of sperm is under nervous control is indicated by the behavior of male Chetopleura following strychninization (cf. Crozier, 1920, Jour, Gen. Physiol., Vol. 2, pp. 627-634) 4 See Crozier, W. J., 1920, ‘‘Sex-correlated Coloration in Chiton,"" AMER. Nart., Vol. 54, pp. 84-88. Tidal, or rather lunar, periodicity in the libera- tion of gametes has been observed also in Chetopleura; I was able to note a probable lunar periodieity in this genus, in 1919, at Woods Hole, and the point is dealt with at length in a recent paper by Grave, B. H., 1922. Biol, Bull., Vol. 42, pp. 234-256. 480 THE AMERICAN NATURALIST [ Vor. LVI by the traetive eurrent, and emerges posteriorly in eompany with numerous large greenish eggs, about whieh, under the mieroseope, it ean be seen that many sperms are gathered. But no real ‘‘eluster-formation’’ takes place. The body juiees of the ripe female, whether or not diluted with.sea water, do not eause agglutination of sperm suspen- sions. But ovarian extraets from (mature) eggs in sea water do induce decided and apparently typical agglutination. So far as I know, sperm-agglutination by ovarian extracts has not previously been seen in molluscs.” Sea water into which ripe eggs have been shaken from an ovary and the whole allowed to stand for half an hour has a similar agglutinative effect. Coneerning the signifieanee of the cluster formation, then, these two points seem significant: (1) the absence of such a process in normal feeundation, and (2) its conspicuous occur- rence when sperm, before the real onset of the breeding season, has passed through the etenidial channels of males or immature females. It could not be discovered whether or not the mature female in a non-spawning interval would cause this cluster pro- duction, because at such times the consistent response of a fe- male to an impinging current of sperms was to depress the girdle to the substratum, thus cutting off the water current carrying sperms, and, by reducing the volume of the etenidial channel, violently to expel from below the sperms already ad- mitted. These observations do not, of course, merit analysis of the rôle of egg-substances in fertilization of chiton, but do serve to point the contention that mere evidence of sperm agglutina- tion (cluster formation) may well have no bearing on such anal- ysis. It is possible that the sperms set free at a period before the natural ripening of eggs are in some degree immature, their surface perhaps more sticky, or liable to be made so by slight external changes experienced in passing between the gill fila- ments of another individual. W. J. CROZIER ZOOLOGICAL LABORATORY, UTGERS COLLEGE 5 Loeb, J., 1916, ** The Organism as a Whole," x + 379 pp., New York. Woodward, A. E., 1918, ‘* Studies on the Physiological Significance of Certain Precipitates from the Egg mcm of Arbacia and Asterias,’’ Jour. Ezper. Zool., Vol. 26, pp. 459-5 Just, E. E., 1919, ** The teen. Renetion 1 in Echinarachnius parma,’’ II, Biol. Bull. Vol. 36, pp. 11-38. THE AMERICAN NATURALIST Vor. LVI. .. JNovember- December, 1922 No. 647 THE PROGRESSION OF LIFE IN THE SEA! DR. E. J. ALLEN, F.R.S. THE method we usually follow in the ordinary course of zoologieal work is to make first, with the unaided eye, a general examination of the animal that interests us, and then study in detail its separate parts with a simple lens, with a low power of the microscope, with gradually inereasing powers, until, finally, minute portions are ex- amined with the highest oil-immersion lens. The suc- cessful research worker is generally one who, whilst car- rying to the utmost limit he ean achieve his search into detail, maintains as by instinet a true sense of proportion and holds firmly to the idea of the organism as a whole. In diseussing the living organisms of the sea I shall try to follow a similar plan, thinking of the life of the sea as a whole, built up of individual plants and animals, each in intimate relation with its surroundings, and all interdependent among themselves. But even this is not enough, for we must take still a wider view and keep in mind not only the life of the waters, but that also of the land and of the air, for both, as we shall see, have a bear- ing on our theme. Deep oceans, coastal waters, shallow seas, rivers and lakes, continents and islands, all play their part in one scheme of organic life—life which had, it seems, one origin, and, notwithstanding migrations and transmigrations from water to land, from land to air, and from land and air back again to the water, remains one closely interrelated whole. 1 Address of the President of the Section of Zoology of the British As- sociation for the Advancement of Science, Hull, September, 1922. 481 482 THE AMERICAN NATURALIST [Vor. LVI Both Brandt* and Gran* have recently emphasized the faet that it is in the coastal waters and shallow seas, which receive much drainage from the land, that plant and animal life are most abundant, the more open oceans far from land being relatively barren; as Schütt puts it, the pure blue of the oceans is the desert color of the seas. This increased production in the coastal waters is due principally to the presence of nitrogen compounds and compounds of phosphorus derived from terrestrial life. From forest, moor and fen, wherever water trickles, the life of the land sends its infinitesimal quota of these es- sential foodstuffs to fertilize the sea. When, however, we go back to the beginning of things, we shall probably be right if we say that any influence of terrestrial life upon life in the sea must be left out of account. Different views are still held as to where life in the world had its origin, but no one questions that it began in close connection with water. That it began in the sea, where the necessary elements were present in appropriate concentrations and in an ionized state, is an idea which appeals to many with increasing force the more closely it is examined. This view has been devel- oped recently by Church * in his memoir on ** The Building of an Autotrophie Flagellate,’’ in which he boldly at- tempts to trace the progression from the inorganic ele- ments present in sea-water to the unicellular flagellate in the plankton phase, floating freely in the water. The autotrophic flagellate, manufacturing its own food, he regards as the starting-point from which all other organ- isms, both plants and animals, have sprung. To under- stand the first step i in this progression, the passage from the dead inorganic to the living organic remains, as it has always been, one of the great goals of science, not of biological siendo alone, but of all science. Recent re- search has, I think, thrown much light on the fundamental problems involved. In a paper published last year, Baly, 2 Wissensch. Meeresunters. Kiel, 18, 1916-20, p. 187. 3 Bull. Planktonique. Cons. Internat., 1912 (1915). * Biological Memoirs I. Oxford, 1919. No. 647] PROGRESSION OF LIFE IN THE SEA 483 Heilbron, and Barker,’ extending and correcting previous work by Benjamin Moore and Webster,’ have shown that light of very short wave-length (A= 200 me), obtained from a mercury-vapor lamp, acting upon water and ear- bon dioxide alone, is capable of producing formaldehyde, with liberation of free oxygen. Light of a somewhat longer wave-length (A= 290,44) causes the molecules of formaldehyde to unite or polymerize to form simple sugars, six molecules of formaldehyde, for example, uni- ting to form hexose. The arresting fact brought out in these researches is that the reactions take place, under the influence of light of appropriate wave-lengths, with- out the help of any catalyst, either organic or inorganic. Where a source of light is used which furnishes rays of many wave-lengths, the simple reaction of the formation of formaldehyde is masked by the immediate condensa- tion of the formaldehyde to sugar, but this formation of sugar can be prevented by adding to the solution a sub- stance which absorbs the longer wave-lengths, so that only the short ones which produce formaldehyde are able to act. When the formation of sugars is postulated, the intro- duction of nitrogen into the organic molecule offers little theoretical difficulty ; for not only has Moore‘ shown that nitrates are converted into the more chemically active nitrites under the influence of light of short wave-length, but he maintains that marine alge, as well as other green plants, can under the same influence assimilate free nitro- gen from the air. Baly® also has succeeded in bringing about the union of nitrites with active formaldehyde in ordinary test-tubes by subjecting the mixture to the light of a quartz-mercury lamp. It will be admitted that these three reactions: (1) the 5 Journ. Chem. Soc., London, Vols. 119 and 120, 1921, p. 1025. Nature, Vol. 109, 1922, p. 344. 6 Proc. Roy. Soc. B., Vol. 87, p. 163 (1913), p. 556 (1914); Vol. 90, p. 168 (1918). | * Proc. Roy. Soc. B., Vol. 90, p. 158 (1918); Vol. 92, p. 51 (1921). 8 Baly, Heilbron and Hudson, Journ. Chem. Soc., London, Vols. 121 and 122, 1922, p. 1078. 484 THE AMERICAN NATURALIST [Vor. LVI formation of formaldehyde, H.CO.H, from carbonic acid, OH.CO.OH, with liberation of free oxygen, or, to put it more simply, the direct union of the carbon atom of CO, with a hydrogen atom of H-O; (2) the formation of sugars from formaldehyde, and (3) the formation from nitrites and formaldehyde of nitrogenous organic sub- stances, are the most fundamental and characteristic re- actions of organic life. It is true that light of such short wave-lengths (A= 200 »») as were required in Baly's ex- periments to synthesize formaldehyde does not occur in sunlight as it reaches the earth to-day; but, as we shall see later, the same author has shown that, in the presence of certain substances known as photocatalysts, the reac- tion can be brought about by ordinary visible light; and from Moore and Webster’s work it appears that colloidal hydroxides of uranium and of iron are suitable photo- catalysts for the purpose. If these results of the pure chemist are justified, they go far towards bridging the gap which has separated the inorganic from the organic, and make it not too presump- tuous to hazard the old guess that even to-day it is possible that organic matter may be produced in the sea and other natural waters without the intervention of living organ- isms. We may note here, too, that if we take account of only the most accurate and adequately careful work, the actual experimental evidence at the present time requires the presence of a certain amount of organie matter in the culture medium or environment before the healthy growth of even the simplest vegetable organism can take place. This was illustrated in some experiments made by myself some years ago when attempting to grow a marine diatom, Thalassiosira gravida, in artificial sea-water made up from the purest chemicals obtainable dissolved in twice- distilled water. Even after nutritive salts, in the form of nitrates and phosphates, had been added, little or no growth of the diatom occurred. But if as little as 1 per cent. of natural sea-water were added, excellent cultures resulted, in which the growth was as healthy and vigor- ous as I was able to obtain when natural sea-water was No. 047] PROGRESSION OF LIFE IN THE SEA 485 used entirely as the basis of the culture medium. There was clearly some substance essential to healthy growth contained in the 1 per cent. of natural sea-water, and from further experiments it became praetieally certain that it was an organic substance. When, for instance, the nat- ural sea-water was evaporated to dryness,.the residue slightly heated and redissolved in distilled water, 1 per cent. of this solution added to the artificial culture me- dium was as potent in producing growth of the diatom as the original natural sea-water had been. When, on the other hand, the residue after evaporation was well roasted at a dull red heat and redissolved in distilled water, the addition of this solution to the artificial culture medium produced no effect and growth did not take place. Growth could also be stimulated by boiling a small frag- ment of green seaweed (Ulva) in the artificial culture medium, the weed being removed before inoculation with the diatom. All this points to the necessity for the pres- ence of some kind of organic matter in the solution before growth can take place. One must not dogmatize, how- ever, for there are many pitfalls in the experimental work and the necessary degree of accuracy is difficult to attain. My own experience of these difficulties culmi- nated when I discovered, covering the bottom of my stock bottle of distilled water—water which had been carefully redistilled from bichromate of potash and sulphuric acid in all-glass apparatus—a healthy growth of mold. Let us then assume that we are allowed to postulate 1n primitive sea-water or other natural water organic com- pounds formed by the energy of light vibrations from ions present in the water, and see how we may proceed to picture the building up of elementary organisms. Without doubt the evolutionary step is a long and elab- orate one, for even the simplest living organism is al- ready highly complex both in structure and in function. As the molecules grew more complex by the progressive linkage of the carbon atoms of newly formed carbohydrate and nitrogenous groups, we must suppose that the or- ganic substance, for purely physical reasons, assumed * 486 THE AMERICAN NATURALIST [Vor. LVI the colloidal state, and at the same time its surface-ten- sion became somewhat different from that of the sur- rounding water. With the assumption of the colloidal state, the electric charges on the colloidal particles would produce the effect of adsorption and fresh ions would be attracted from the surrounding medium, producing a kind of growth entirely physical in character. We thus ar- rive at the conception of a mass of colloidal plasma dif- fering in surface-tension from the water and increasing in size by two processes, the one chemical, due to linkage of carbon atoms; the other physical, brought about by the adsorption of ions by the colloidal particles. The difference of surface-tension would tend to make the surface a minimum and the shape of the mass spher- ical. On the other hand, maximum growth would demand maximum exchange with the surrounding medium, and hence maximum surface. From the antagonism of these two factors, surface-tension and growth, there would fol- low, firstly, the breaking up of the mass into minute par- ticles upon the slightest agitation, and, secondly, changes of form wherever growth involved local alterations of surface-tension, which changes of form would represent the first indication of the property of contractility. So far we have considered only the process of the building up of the elementary plasmic particles, the anabolic: process. Church, whose memoir already re- ferred to I am now closely following, points out that these anabolic operations must from the beginning have been subject to the alternations of day and night, for dur- ing the night the supply of external energy is removed. ** If during the night,’’ he asks, ‘‘ the machine runs down, to what extent may it be possible so to delay the onset of molecular finality that some reaction may continue, at a lower rate, until the succeeding day?" And his answer is: ** The successful solution of this problem is defined physiologically by the introduction of the conception ‘katabolism,’ 'as implying that energy derived from the ‘breaking down’ of the plasma itself... may be re- garded as a ‘ secondary engine,’ functional in the absence No. 647] PROGRESSION OF LIFE IN THE SEA 487 of light, and evolved as a last resort in failing plasma." Katabolism persists as the ultimate mechanism in the physiology of animal as contrasted with plant life, but if the suggestion just quoted is sound, it originated, as the first ** adaptation " of the organism, to meet the factor of recurring night and day. That the problem was suc- cessfully solved we know, but as to the mechanism of its solution we have no key. It is at this point again, to use Church’s words, that the ** plasma, previously within the eonnotation of chemieal proteid matter, becomes an auto- trophie, inereasingly self-regulated, and so far individu- alized entity, to which the term ‘ life’ is applied." The elementary plasma is thus now fairly launched as an individual living organism, and the great fundamental problems of biology —memory, heredity, variation, adap- tation—face us at each step of our further progress. We see in broad outline the conditions the advaneing organ- ism had to meet, we see the means by which those condi- tions were in faet met, we know that only those individuals survived which were able to meet them. Further than this we, the biologists of to-day, have not advanced. The younger generation will pursue the quest, and, with pa- tient effort, much that now lies hidden will grow clear. The differentiation of the growing particles of plasma into definite layers, which followed, seems natural; first the external layer, in molecular contact with the sur- rounding water, from which it receives substances from outside in the form of ions, and to which it itself gives off ions; beneath this the autotrophic layer to which light penetrates, and in which, under the influence of the light, new organic substance is built up; in the center a layer to which light no longer penetrates. This central region, the nucleus, depends entirely on the peripheral layers for its own nutrition, and becomes itself concerned only with katabolie processes, those processes of the organism whieh depend upon the breaking down, and not the build- ing up, of organie substance. At an early stage in the development of the individual organism the spherieal shape, which the organie plasma 488 THE AMERICAN NATURALIST [Von. LVI was compelled to assume under the influence of surface- tension, underwent an important modifieation, the effect of which has impressed itself upon all later develop- ments. A spherical organism floating in the water and growing under the direct influence of light would ob- viously grow more rapidly on the upper side, where the light first strikes it, than it would on the lower side away from the light. There followed, therefore, an elongation of the sphere in the vertieal direction, and the definite establishment of an anterior end, the upper end which lay towards the light and at which the most vigorous growth took place. In this way there was established a definite polarity, which has persisted in all higher organ- isms, a distinetion between an anterior and a posterior end. With the concentration of organic substance which took the form of nucleus and reserve food supply, the specifie gravity of the plasma would become greater than that of the surrounding water and the organism would tend to sink. The necessity, therefore, arose for some means of keeping it near the surface, that it might con- tinue to grow under the influence of light. The response to this need, however it was attained, came in the de- velopment of an anterior flagellum. This we may regard as an elongation in the direction of the light of a contrac- tile portion of the external layer, moving rhythmieally, which by its movement counteracted the action of gravity, and acting as a tractor drew the primitive flagellate up- wards towards the surface layers, into a position where further growth was possible. That this speculation of Church’s represents what was actually accomplished, ‘even though it does not make clear the means by which it was brought about, is shown by the interesting re- searches of Wager?’ on the rise and fall of the more highly organized flagellate Euglena. Euglena is a somewhat pear-shaped flagellate, the tapering end being anterior and provided with a single flagellum, which acts as a tractor drawing the organism towards the light. The 9 Phi!. Trans. Roy. Soc., Vol. 201, 1911; and Science Progress, Vol. vi. October, 1911, p. 298. No. 647] PROGRESSION OF LIFE IN THE SEA 489 posterior end carries the nucleus and most of the chloro- phyll and food reserves. The whole organism has a specific gravity of 1.016, being slightly heavier than the fresh water in which it lives. When dead, or when the flagellum is not moving, it takes up, under the action of gravity alone, a vertical position in the water, with the pointed anterior end uppermost, and the heavier, rounded, posterior end below, and sinks gradually to the bottom. In a very crowded culture a curious phenomenon is seen, because the organisms tend to aggregate into clus- ters beneath the surface film, and when they are crowded together in these clusters the flagella cease to work. This makes the whole cluster sink to the bottom under the action of gravity. When the bottom is reached the in- dividuals are spread out by the action of the downward current, and, when they are sufficiently widely apart, the flagella again begin to move, carrying the organisms in a more diffuse stream once more to the surface. The whole culture vessel becomes filled with a series of ver- tical lines of closely aggregated falling organisms, sur- rounded by a broad cylinder of disseminated swimming ones, rising to the surface by the action of their flagella. If the conditions are kept uniform, such a circulation of Euglenas, falling to the bottom by gravity when the fla- gella are stopped and returning to the surface under their own power, will continue for days. The flagellum in this species, therefore, retains its most primitive function of drawing the organism to the light in the surface layer. With the establishment of the flagellum an organ is produced which shows remarkable persistence in both the animal and vegetable kingdoms, and from the existence of the flagellated spermatozoon in the higher vertebrates, in accordance with Haeckel’s bio- genetic law that the individual in its development repeats or recapitulates the history of the race, we conclude that they also in their earliest history passed through a plank- ton flagellate phase. Exactly at what stage in the history of the autotrophic flagellate the first formation of chlorophyll and its allied 490 THE AMERICAN NATURALIST [Vor. LVI pigments took place we have no means of determining, but it may have been before even the flagellum itself had begun. This advance and the subsequent concentration of the pigments into definite chromatophores or chloro- plasts doubtless immensely increased the efficiency of the organism in producing the food which was necessary to it. The recent work of Baly and his collaborators be- comes here again of the first importance, and though the subject of the part played by chlorophyll in photosyn- thesis belongs rather to botany and chemistry than to zoology, I may perhaps for the sake of completeness be allowed to refer to it very briefly. . I have already said that Baly brought about the synthesis of formaldehyde from CO, and H.O under the influence of rays of very short wave-length (A= 200) from a mereury-vapor lamp. He was also able to show that when certain col- ored substances were added to the solution of carbon dioxide in water the same reaction took place under the influence of ordinary visible light. His explanation of this process is that the colored substance known as the photocatalyst absorbs the light rays and then itself radi- ates, at a lower infra-red frequency corresponding to its own molecular frequency, the energy it has absorbed. At this lower frequency the energy thus radiated is able to activate the carbonic acid, so that the reaction leading to the formation of formaldehyde can and does take place. In the living plant this synthesized formaldehyde probably at once polymerizes to form sugars. Malachite green and methyl orange, as well as other organic compounds, were found to act as photocatalysts capable of synthesizing formaldehyde, and Moore and Webster’s work had previously shown that inorganic substances, such as colloidal uranium oxide and colloidal ferric oxide, can do the same. Chlorophyll in living plants may with some confidence be assumed to operate in a similar way, though no doubt the series of events is more complex, since the green pigment itself is not a single pigment, and others, such as earotin and xantho- phyll, are also concerned. No. 647] PROGRESSION OF LIFE IN THE SEA 491 We have tried to pieture the gradual building up from elements occurring in sea-water of a chlorophyll-bearing flagellate, capable of manufacturing its own nourishment and able to multiply indefinitely by the simple process of dividing in two. If we assume only one division during each night as a result of the day’s work in accumulating food material, such an organism would be able in a com- paratively short space of time to occupy all the natural waters of the world. But here we are met by a difficulty which is not easily overcome. Chlorophyll, the photo- catalyst, the most essential factor in the building up of the new organic matter, is itself a highly complex organic substance, and in any satisfactory theory its original for- mation and its constant increase in quantity must be accounted for. Lankester?? has maintained that chloro- phyll must have originated at a somewhat late stage in the development of organie life, and has suggested that earlier organisms may have nourished themselves like ani- mals on organie matter already existing in a non-living state. An alternative hypothesis, which in view of the recent work seems more attractive, is to suppose that the earlier organisms were either activated by some simpler photocatalyst, or that they received the necessary energy at suitable frequency directly from some outside source. It must not be forgotten, also, that at the time these developments were taking place the conditions of the en- vironment would in many ways have been different from those now existing in the sea. One suggestion of special interest that has been made” is that the concentration of carbon dioxide in the atmosphere, and hence also in nat- ural waters, was very much greater than it is to-day. Free oxygen, indeed, may have been entirely absent, and all the free oxygen now present in the air may owe its existence to the subsequent splitting up of carbon diox- ide by the action of plant life. With such possibilities of differenees in the conditions in this and in so many 10** Treatise on Zoology,'' Part I, Introduction. London, 1909. 11See Carl Snyder, ‘‘Life without Oxygen,’’ Science Progress, Vol. vi, 1912, p. 107: 492 THE AMERICAN NATURALIST [Vorn LVI other directions, may we not be well satisfied if, for the time, we can say that the formation of carbohydrates and proteids has been brought within the category of ordi- nary chemical operations, which can occur without the previous existence of living substance? To return once more, however, to the free-swimming, autotrophic flagellate. In the early stages of its history the loss caused by sinking, and so getting below the in- fluence of light and the possibility of further growth, must have been enormous. We may conceive a constant rain of dead and dying organisms falling into the darker re- gions of the sea, and thus a new field would be offered for the development of any slight advantages which par- ticular individuals might possess. Under such conditions we may suppose that the holozoie or animal mode of nu- trition first began in the absorption of one individual by another one, with which it had chanced to come into con- tact. If the one individual were more vigorous and the other moribund, we should designate the process ** feed- ing,’’ and the additional energy obtained from the food might well cause the individual to survive. If the two individuals which coalesced were both sinking from loss of vigor, the combined energy of the two might make pos- sible a return to the upper water layers, where under the influence of light growth and multiplication would pro- ceed, and we should, I suppose, designate the coalescence ** conjugation,’’ or sexual fusion. Other individuals, again, sinking in shallow water, would stick to solid objects on the sea-floor, whilst the flagellum continued to vibrate. The current produced by the flagellum under these conditions would draw towards the organism dead and disintegrating remains of its fel- lows, and again we should have ingestion and animal nutrition. At this stage we witness the definite passage from plant to animal life. A further stage is seen when a cup-like depression to receive the incoming particles of food is formed at the base of the flagellum, to be followed still later by a definite mouth. Any roughening of the external surface of the swim- No. 647] PROGRESSION OF LIFE IN THE SEA 493 ming flagellate, such as we so often find brought about by the deposition of ealeareous plates or silicious spicules or the produetion of ridges or furrows, would tend to slow down its speed of travel, from increased friction with the surrounding water. This would have a similar effect to actual fixation in drawing floating particles by the action of the flagellum, and also lead to animal nutri- tion. Still another development would oecur when the fallen flagellate began to creep along the sea-floor by contractile movements of the plasmic surface, losing its flagellum, and adopting the mode of life of an ameba. That amoba and its allies, the Rhizopods, are descended from a flagellate ancestor is a view suggested by Lan- kester'* in 1909, which was adopted by Doflein,? and is now strongly advocated by Pascher™ as a result of much new research. The transformation from the plant to the animal mode of feeding we can see in action by studying actual organ- isms which exist to-day. In the course of my work al- ready referred to on the culture of plankton organisms there has on several occasions flourished in the flasks a small flagellate belonging to the group of Chrysomonads, which was first described by Wysotzky under the name of Pedinella hexacostata, and to which I drew the attention of Section D at the Cardiff Meeting in 1920. The general form of Pedinella resembles that of the common Vorti- cella, but its size is much smaller. The body, which is only about 5» in diameter, is shaped like the bowl of a wine glass, and from the base of the bowl, which is the posterior end, a short, stiff stalk extends. From the center of the anterior surface there arises a single long flagellum, surrounded at a little distance by a circle of short, stiff, protoplasmic hairs. Arranged in an equa- torial ring just inside the body are six or eight brownish- green chromatophores or chloroplasts. In a healthy eul- 1? Lankester, ‘‘ Treatise on Zoology,’’ Part I, London, 1909, p. xxii. 13 Doflein, ‘‘ Protozoenkunde,’’ 1916. 14Pascher, Archiv f. Protistenkunde, Ba. 36, 1916, p. 81, and Bd. 38, 1917, p. 1. 494 THE AMERICAN NATURALIST [Vor. LVI ture Pedinella swims about freely by means of a spiral movement of the flagellum, which functions as a tractor, the stalk trailing behind. The chromatophores are large, brightly colored and well developed, and the organism is obviously nourishing itself after the manner of a plant, like any other Chrysomonad. But from time to time a Pedinella will suddenly fix itself by the point of the trail- ing stalk. The immediate effect of this fixing is that a eurrent of water, produced by the still vibrating flagel- lum, streams towards the anterior surface of the body, and small particles in the water, such as bacteria, become caught up on the anterior surface, the ring of fine stiff hairs surrounding the base of the flagellum being doubtless of great assistance in the capture of this food. One can clearly see bacteria and small fragments of similar size engulfed by the protoplasm of the anterior face of the Pedinella and taken into the body. The organism is now feeding as an animal. In some of the cultures in which bacteria were very plentiful nearly all the Pedinella re- mained fixed and fed in the animal way, and when this was so the chromatophores had almost disappeared, though they could still be seen as minute dark dots. We can, as it were, in this one organism see the transition from plant to animal brought about by the simple process of the freely swimming form becoming fixed. In the group of Dinoflagellates, also—the group to which the naked and armored peridinians belong—the same transition from plant to animal nutrition ean be well followed by studying different members of the group. In heavily armored forms, with a rich supply of chro- matophores, nutrition is chiefly plant-like or holophytie. In those with fewer ehromatophores there is, on the other hand, often distinct evidence of the ingestion of other organisms, and nutrition becomes partly animal.like. Amongst the naked Dinoflagellates such holozoie nutri- tion is very much developed, and in many species has entirely superseded the earlier method of carbonic acid assimilation. It is really surprising how many structural features No. 647] PROGRESSION OF LIFE IN THE SEA 495 . found in higher groups of animals make their first ap- pearance in these naked Dinoflagellates in conjunction with this change of nutrition, and we seem to be led directly to the metazoa, especially to the Coelenterata. In Poly- krikos there are well-developed stinging cells or nema- toeysts, as elaborately formed as those of Hydra or the anemones. In Pouchetia and Erythropsis well-developed ocelli are found, consisting of a refractive, hyaline, some- times spherieal lens, surrounded by an inner core of red pigment and an outer layer of black; the whole structure is comparable to the ocelli around the bell of a medusa. In Noctiluca and in the allied genus Pavillardia a mobile tentacle, which is doubtless used for the capture of food, is developed. Division of the nucleus, with the formation of large, distinet chromosomes, has also been described in several of these Dinoflagellates. With the tendency of the cells in certain species to hold together after divi- sion and form definite chains, we seem to approach still nearer to the metazoa, until, finally, in Polykrikos we reach an organism which may well have given rise to a simple pelagic eclenterate. It is difficult to resist the suggestion put forward by Kofoid* in his recent mono- graph, that if to Polykrikos, with its continuous longitu- dinal groove which serves it as a mouth, its multicellular and multinucleate body and its nematocysts, we could add the tentacle of Noctiluca, and perhaps also the ocellus of Erythropsis, ** we should have an organism whose ‘struc- ture would appear prophetic of the Coelenterata and one whose affinities to that phylum and to the Dinoflagellata would be patent." Or it may be that the older view is the correct one here, and that the first cclenterate came from a spherical colony of simple holozoic flagellates, ar- ranged something on the plan of Volvoz, in which the posterior cells of the swimming colony, in whose wake food particles would collect, had become more speeialized for nutrition than the rest. Before proceeding, however, to consider the further 15 Kofoid and Swezy, ''The Free-living Unarmored oe M Mem. Univ. California, 496 THE AMERICAN NATURALIST [Vor. LVI progress of animal life, we must pause for a moment to ask in what direction plant life in the sea developed, from which the inereasing animal life derived its nourish- ment. Here the striking fact is the lack of progress in the free, floating, plankton phase. The plant life of the plankton has never proceeded beyond the unicellular stage, for the plankton diatoms, which with the peri- dinians form the great, fundamental vegetable food sup- ply of the sea, are only autotrophic flagellates which have lost their flagella, having acquired other means of flota- tion to keep them in the sunlit region of the upper water layers. Deriving their food, as these plants do, directly from molecules in the sea-water, the factor which is for them of supreme importance is the exposure of maximum surface directly to the water. Hence the minute unicel- lular form has been the only one to survive as phytoplank- ton. The marine region in which plant life has succeeded in making some progress is the narrow belt along the shores, where a fixed life is possible, but this belt, limited by the amount of light which penetrates, extends only to a depth of about 15 fathoms. The available area is further restricted to rocky and hard bottoms, and is therefore nowhere great. This is the wave-lashed region of the brown and red seaweeds. In the brown seaweeds a history ean still be traced, from the fixture of an auto- trophie flagellate to the building up, by laying cell on cell, of the essential structures which afterwards, on trans- migration to the land, reached their climax in the forest tree. But if the flagellate thus rose and gave origin to the flora of the land, it also degenerated, for it adopted a parasitie habit, living in and direetly absorbing already formed organic matter. In this way the bacteria arose, whose activities in so many directions influence the life of to-day. This view exceeds in probability, I think, the suggestion often put forward,” that it is to the simpler bacteria we must look for the first beginnings of life. 16 Church, Botanical Memoirs, No. 3. Oxford, 1919. 17 Osborn, ‘‘The Origin and Evolution of Life," 1918. Waksman and Joffe, **Miero-organisms concerned in the Oxidation of Sulphur in the No. 647] PROGRESSION OF LIFE IN THE SEA 497 After this digression on the botanieal side we must return to the primitive eclenterate and see on what lines evolution proceeded in the animal world. As a purely plankton organism, swimming freely in the water, the progress of the ecdlenterate was not great, and reached, as far as we know, no further than the modern Cteno- phore. The Ctenophore seems to represent the culmi- nating point of the primary progression of pelagie ani- mals, which derived directly from the autotrophic flagellate. Further evolution was associated with an abandonment by a eolenterate-like animal of the pelagic habit, and the establishment of a connection with the sea bottom, either by fixing to it, by burrowing in it, or by creeping or running over it. At a later stage many of the animals which had become adapted to these modes of life developed new powers of swimming, and thus gave rise to the varied pelagic life which we find in the sea to-day; but this must be regarded as secondary, the pri- mary pelagic life, so far as adult animals were concerned, having ended with the evolution of the Ctenophore.'* Sueh is the teaching of embryology, the history of the race being conjectured from the development of the in- dividual. In group after group of the animal kingdom, when the details of its embryology become known, the indications are the same—first the active spermatozoon, reminiscent of the plankton flagellate, then the pelagic larval stage, recalling the celenterate, and then a bottom- living phase. The primitive, free-swimming celenterate, adopting a fixed habit and becoming attached mouth upwards to solid rock or stone, gave rise to hydroids, anemones and Soil,’’ Journal of Bacteriology, VII, 2, March, 1922. The authors claim that Thiobacillus thiooxidans will grow in solutions containing no organic matter. In view of the minute traces of organic matter that suffice for the growth of bacteria and molds, care must be taken, however, in draw- ing conclusions from experiments made in flasks or tubes closed in the ordinary way with cotton-wool plugs and subsequently sterilized in flowing steam. 18 There is perhaps a possibility that further knowledge of the embry- ology of Sagitta and its allies might make it necessary to modify this suggestion. : 498 THE AMERICAN NATURALIST [Vor. LVI corals, typical inhabitants of the coastal waters, for the sands and muds at greater depths offered few points of attachment suffieiently stable. A Volvox-like colony of simple holozoie flagellates, ac- cording to MacBride,” commenced to feed upon micro- scopic organisms lying on the sea bottom, and under these circumstances only the cells of the lower half of the colony would be effective feeders. The upper cells, there- fore, lost their flagella and became merely a protective layer, which finally grew downwards outside the others and fixed the colony to the ground. In this way a sponge was formed. The collar cell, so typical of the group, had been developed already by the flagellates, its first incep- tion being perhaps a circle of protoplasmic hairs such as we find in Pedinella. But this adoption of a fixed habit, as it were mouth downwards, did not lead very far, and though there has been much elaboration within the group itself, the sponges have remained an isolated phylum, unable to develop into higher forms. It is in a Ctenophore-like ancestor that we find the line of development to higher animal groups, and this an- cestor must have been at one time widely distributed in the seas. Its immediate descendants are familiar to every zoological student in the well-known series of pe- lagie larval forms. Müller's larva, taking to the bottom, and in its hunt for food gliding over hard surfaces with its eilia, led to the flatworms; the Pilidium, developing a thread-like body and creeping into cracks and crevices to transfix its prey, gave rise to the nemertines. A Tro- chophore, burrowing in soft mud and sand, developed a segmented body which gave it later the power of running on these soft surfaces, and became an annelid worm. An- other Trochophore, developing a broad, muscular foot, crept on the sand, and afterwards buried itself beneath it as a lamellibranchiate molluse, or migrated on to harder surfaces as the gastropod and its allies. Pluteus, Bipinnaria, Auricularia, first fixing, as the crinoids still do, and developing a radial symmetry, afterwards broke free and wandered on the bottom as sea-urchin, star-fish 19'fTextbooks of Embryology. Invertebrata." London, 1914. No. 647] PROGRESSION OF LIFE IN THE SEA 499 and cucumarian. Tornaria developed into Balanoglos- sus, whose structure hints to us that the ascidians and vertebrates came from a similar stock. All the phyla thus represented derive directly from the free-swimming Ctenophore-like ancestor, and only one considerable group, the Arthropods, remains unaccounted for. The evolutionary history of an Arthropod is, however, not in doubt. Its marine representatives, the Trilobites and Crustacea, came directly from annelids, which, after their desertion of a pelagic life to burrow in the sea-floor and run along its surface, again took to swimming, and not only stocked the whole mass of the water with a rich and varied life of Copepods, Cladocera and Schizopods, but gave rise to Amphipods, Isopods, and Decapods, groups equally at home when roaming on the bottom or swim- ming above it. Another important addition to the pelagic fauna we should also notice here. From the molluses, creeping on solid surfaces, sprang a group of swimmers, the Cephalo- pods, which have grown to sizes almost unequaled amongst the animals of the sea. All these invertebrate phyla had become established and most of them had reached a high degree of develop- ment in the seas of Cambrian times. Amongst animals then living there are many which have survived with little change of form until to-day. One is almost tempted to suggest that the life which the sea itself could produce was then reaching its summit and becoming stabilized. Since Cambrian times geologists tell us some thirty mil- lion years” have passed, a stretch of time which it is really difficult for our imaginations to picture. During that time a change of immense moment has happened to the life of the sea; but if we read the signs aright, that change had its origin rather in an invasion from without than in an evolution from within. Whence came that tribe of fishes which now dominates the fauna of the sea? It would be rash to say that we can give any but a speculative reply to the question, but the probable an- swer seems to be that fishes were first evolved not to meet 20 Osborn, ‘‘ Origin and Evolution of Life,’’ 1918, p. 153. 500 THE AMERICAN NATURALIST [Vor. LVI eonditions found in the sea, but to battle with the swift eurrents of rivers, where fishes almost alone of moving animals can to this day maintain themselves and avoid being swept helplessly away." It was in response to these conditions that elongate, soft-bodied creatures, which had penetrated to the river mouth, developed the slender, stream-lined shape, the rigid yet flexible muscu- lar body, the special provision for the supply of oxygen to the blood to maintain an abundant stock of energy, and all those minute perfections for effective swimming that a fish’s body shows. The fact that many sea-fishes still return to the rivers, especially for spawning, sup- ports this view, and it is in accordance with Traquair's classical discoveries of the early fishes of the Scottish Old Red Sandstone, which were for the most part fresh- and brackish-water kinds. Having developed, under the fierce conditions of the river, their speed and strength as swimmers, the fishes returned to the sea, where their new-found powers en- abled them to roam over wide areas in search of food, and gave them such an advantage in attack and defense that they became the predominant inhabitants of all the coastal waters, and as such they remain to-day. The other great migration of the fishes, also, the migra- tion from the water to the land, giving rise to amphibians, reptiles, birds and mammals, must not be left out of ac- count. The whales, seals and sea-birds, which after de- veloping on land returned again to the waters and became readapted for life in them, are features which can not be neglected. | And so we are brought to the pieture of life in the sea as we find it to-day. "The primary produetion of organie substance by the utilization of the energy of sunlight in the bodies of minute unicellular plants, floating freely in the water, remains, as it was in the earliest times, the feature of fundamental importance. The conditions which control this production are now, many of them, known. Those of chief importance are (1) the amount of light which 21 Chamberlin, quoted in Lull, ‘‘Organie Evolution," New York, 1917, p. 462. No. 647] PROGRESSION OF LIFE IN THE SEA 501 enters the water, an amount which varies with the length of the day, the altitude of the sun, and the clearness of the air and of the water; (2) the presence in adequate quantity of mineral food substances, especially nitrates and phosphates; and (3) a temperature favorable to the growth of the species which are present in the water at the time. Experiments with cultures of diatoms have shown clearly that if the food-salts required are present, and the conditions as to light and temperature are satis- factory, other factors, such as the salinity of the water and the proportions of its constituent salts, can be varied within very wide limits without checking growth. The increased abundance of plankton, especially of diatom and peridinian plankton, in coastal waters and in shallow seas largely surrounded ‘by land, such as the North Sea, is due to the supply of nutrient salts washed directly from the land by rain or brought down by rivers. An exceptional abundance of plankton in particular locali- ties, which produces an exceptional abundance of all ani- mal life, is also often found where there is an upwelling of water from the bottom layers of the sea. These condi- tions are met with where a strong current strikes a sub- merged bank, or where two currents meet. Food-salts which had accumulated in the depths, where they could not be used owing to lack of light, are brought by the upwelling water to the surface and become available for plant growth. The remarkable richness of fish life in such places as the banks of Newfoundland and the Agul- has Banks off the South African coast, each of which is the meeting-place of two great currents, is to be explained in this way. Our detailed knowledge of the steps in the food-chain from the diatom and peridinian to the fish is increasing rapidly. The Copepod eats the diatom, but not every Copepod eats every diatom; they make their choice. The young fish eats the Copepod, but again there is selection . of kind. Even adult fishes like herring and mackerel, which were formerly supposed to swim with open mouth, straining out of the water whatever came in their way, are now thought largely to select their food.” 22 Bullen, Journ. Mar. Biol. Assoc., 9, 1912, p. 394. 502 THE AMERICAN NATURALIST [Vor LVI A result of extraordinary interest in connection with the food-chain has recently been brought to light by two sets of investigators working independently. In seeking to explain certain features which he had found in connec- tion with the growth of the cod, Hjort” undertook a study of the distribution in marine organisms of the growth stimulant known as vitamin. Fat-soluble vitamin was already known to be present in large quantities in eod-liver oil, and is what probably gives the oil its medic- inal value. Hjort was able to trace the vitamin, by means of feeding experiments on rats, in the ripe ovaries of the eod, in shrimps and prawns, which resemble the animals on which the cod feeds, and in diatom plankton and green alge. Jameson, Drummond and Coward* cultivated the diatom Nitzschia closteriwm, and by a simi- lar method to that used by Hjort showed that it was extraordinarily potent as a source of fat-soluble vitamin. We thus conclude that this substance, so essential to healthy animal growth, is produced in large quantities by plankton diatoms, and passed on unchanged to the fish through the crustaceans which feed on the diatoms. In the fish the vitamin is first stored in the liver, and with the ripening of the ovary passes into the egg, to be used to stimulate the growth of the next generation. Again we see the fundamental importance of the food-producing activities of the lowest plant life. Attention has already been drawn to the suggestion that fishes developed their remarkable swimming powers in rivers, in response to a need to overcome the currents, and that they afterwards returned to the sea, where they preyed upon a well-developed and highly complex inverte- brate fauna already fully established there. Their speed enabled them to conquer their more sluggish predecessors, whilst they themselves were little open to attack. With the exception of the larger cephalopods, which are of comparatively recent origin, and were probably evolved after the arrival of the fishes, there are few, if any, in- vertebrates which capture adult fishes as part of their 23 Proc. Roy. Soc., May 4, 1922. 24 Biochemical Journal, 1922. No. 647] PROGRESSION OF LIFE IN THE SEA 503 normal food. Destructive enemies appeared later in the form of whales and seals and sea-birds, which had devel- oped on the land and in the air. And now in these last days a new attack is made on the fishes of the sea, for man has entered into the struggle. He came first with a spear in his hand; then, sitting on a rock, he dangled a baited hook, a hook perhaps made from a twig of thorn bush, such as is used to this day in villages on our own east coast. Afterwards, greatly dar- ing, he sat astride a log, with his legs paddled further from the shore, and got more fish.. He made nets and surrounded the shoals. Were there time we might trace step by step the evolution of the art of fishing and of the art of seamanship, for the two were bound up together till the day when the trawlers and drifters kept the seas for the battle fleet. There can be little doubt that in European seas the at- tack on the fishes in the narrow strip of coastal water where they congregate has become serious. A consider- able proportion of the fish population is removed each year, and human activity contributes little or nothing to compensate the loss. We have not, however, to fear the practical extinction of any species of fish, the kind of extinction that has taken place with seals and whales. Fishing is subject to many natural limitations, and when fishing is suspended recovery will be rapid. There is evi- dence that such recovery took place in the North Sea when fishing was restricted by the War, though the in- crease which was noted is perhaps not certainly outside the range of natural fluctuations. Until the natural fluc- tuations in fish population are adequately understood, their limits determined, and the eauses which give rise to them discovered, a reliable verdict as to the effect of fishing is diffieult to obtain. If such problems as these are to be solved, the investi- gation of the sea must proceed on broadly conceived lines, and a eomprehensive knowledge must be built up, not only of the natural history of the fishes, but also of the many and varied conditions which influence their lives. The life of the sea must be studied as a whole. FAMILY RESEMBLANCES AMONG AMERICAN MEN OF SCIENCE DR. DEAN R. BRIMHALL SECRETARY OF THE PSYCHOLOGICAL CORPORATION Tue fascinating problems that concern the causes of individual and group differences among human beings are still with us. Since Galton set out to prove that ‘‘a man’s natural abilities are derived by inheritance under exactly the same limitations as are the form and physical features of the whole organic world ”’ the biological sci- ences have made many and notable contributions to the fund of knowledge concerning the derivation ‘‘ of the form and physical features of the organic world.’’ But the influences by which individual and group differences come about, particularly differences in intellectual per- formance, seem to have been singularly neglected. The problem has been avoided partly because of the nature of the material. Human beings are not only com- plicated in organic construction, but they are mongrel in breed, the period between generations is long and direct experimentation impossible. Few scientific problems are more likely to be disturbed by the bias of the experi- menter. The American millionaire or European aristo- erat explains differences in the wealth gathering or keeping performance of human beings in terms of innafe ability. To the socialist the expression ‘‘ royal minds ”’ has little basis in fact. Laws, taboos, economic and social conditions are thought to be the proper explanations of differences in human behavior. So deeply do the facts concern the fundamental concepts about the organization of society that debate with its anecdotal method should be supplanted by objective method of the best sort avail- able. The method of approach used in this study is statisti- eal. The investigation represents an attempt to deter- mine some of the differenees or resemblances, the causes No. 647] FAMILY RESEMBLANCES 505 of which have been so often the subject of debate and argument. It is necessary to determine what resem- blances or differences exist before causes can be explained. This is a study of family resemblances in intellectual per- formance, particularly in the field of science. By limiting the problem to the measurement of resem- blances the advice of a gifted and successful worker in this field is followed. He writes: It is impossible at present to estimate with security the relative shares of original nature, due to sex, race, ancestry and accidental variation, and of the environment, physical and social, in causing the differences found in men. e can only learn the facts, ai them with as little bias as possible, and try to secure more By this same limitation it is hoped that a serious and common error will be entirely avoided, namely, the fail- ure to realize the twofold nature of the problem of in- heritance as ordinarily discussed. The following analy- sis is so clear and the need of it so general that it is given at length. Most sociological writers and some biologists are confused in their use of the concept of heredity. When there is discussion of the rela- tive influence on performance of heredity and environment, by hered- ity there is sometimes understood the original constitution of the individual and sometimes his resemblance to parents and other rela- ives. It is conceivable that the original constitution of son and father might be exactly the same and yet the individual be so plastic to environment that under different conditions there would be but slight similarity between their performances. It is also conceivable that there might be no similarity between the original constitution of son and father, and yet the performance of each be determined by his original constitution almost without influence from environment. Under which of these extreme hypotheses would the current sociologist call heredity strong or weak? The word heredity should be reserved for resemblance due to a common germ plasm and some other word found for the constitution of the fertilized ovum or zygote; perhaps the best that can be done is to use this uncouth word. We can then discriminate between the two distinct questions: What is the resem- blance between the zygotes of two brothers? How far does the zygote of an individual determine his performance as an adult? 2 1 Thorndike, ‘‘ Educational Psychology,’’ Vol. III, p. 310. 2 J. McKeen Cattell, ‘‘ Families of American Men of Science,’’ Pop. Sci- ence Monthly, May, 191 15. 506 THE AMERICAN NATURALIST [Vor. LVI It must not be inferred that this study is an attempt to determine ‘‘ how far the zygote of an individual de- termines his performance ’’ or ‘‘ what is the resemblance between the zygotes of two brothers." It is primarily a statistical measurement of resemblance in performance with particular reference to performance in science. With the results of these measurements at hand, some- thing about the resemblances of the zygotes of near rela- tives and about how far the zygote of an individual de- termines his performance may be estimated. By resemblance is meant, not identity, but degree of similarity. Galton, with the idea of particulate inherit- ance in mind, early insisted on measuring resemblances as deviations from an average. He justly claims to have been the first to ‘‘ introduce the law of deviations from the average into the discussions on heredity."? Almost forty years later he is found insisting that ‘‘ proba- bility is the foundation of eugenics.’’* It is here that the statistical method avoids the pitfalls of proof by anec- dote. Given a group, selected for some particular sort of performance, the number in any particular degree of relationship showing similar performance may be de- termined and compared with a similar group of the gen- erality. This is the method used in this study. The group studied consists of approximately 1,000 American men of science and their families. The wives and the near relatives of the wives of the men of science are included in the data, and the results of the compari- sons offer perhaps the most unique contribution of the investigation. The statistical data include only relatives of a degree no more remote than first cousin. These rela- tives are: brothers and sisters, sons and daughters, fathers and mothers, nieces and nephews, uncles and aunts, grandparents and first cousins. It may be well here to anticipate a criticism of the use of the term resemblance. Since the group of men of sct- ence is made up of persons of known distinction one since 3 ‘f Hereditary Genius,’’ Preface, 1869. 4 Spencer lecture, 1907. No. 647] FAMILY RESEMBLANCES 507 resemblance of near relatives is measured in terms of the number of the latter who become distinguished, it is as- sumed that distinction in any intellectual performance is evidence of resemblance. Thus, a psychologist whose cousin is a psychologist may be said without much fear of contradiction to resemble that relative, but does he really resemble his brother, who is a well-known judge? Does he resemble his distinguished father, a former presi- dent of the University of Michigan, a gifted adminis- strator?® Apparently, that has been assumed. The fact is, they do resemble each other in so far as they vary in the same direction from the average person in the direc- tion of performance. The selection of the group of scientific men used has become a classic in individual psychology and need not be related in detail here. Briefly, it may be said, the Workers in science were grouped in twelve general divi- sions as follows: Anatomy, anthropology, botany, chem- istry, geology, mathematies, pathology, physies, physiol- ogy, psychology and zoology. The workers in each Science were arranged in an order of merit by ten leading men in each of those sciences. The average position as- signed each man, together with the probable error, was computed. Thus each man's position with the reliabil- ity of the figure describing his position was determined, and a thousand of the leading men of science were selected as a group for study. Two selections were made, the first in 1903, the second in 1910. The lists varied some- what in composition due to deaths, changes of position within the group, removal to a foreign country and the like." The positions attained were not published, that 5 Since the above was written the psychologist, James Rowland Angell, has become a member of the National Academy of Sciences and the presi dent of Yale University. The question banman less pertinent but the case more dramatic. 6J. McKeen Cattell, ‘‘ American Men of Seience,’’ appendix, second edition, 1910. A third selection has now been made and will furnish material for a continuation of this study. See ‘‘ American Men of Science,’’ third edi- tion, J. McKeen Cattell and Dean R. Brimhall, 1921. 508 THE AMERICAN NATURALIST [Vou. LVI .information being confidential, but the names of those achieving a position in either selection were marked with an asterisk in the handbook in which they were published. They are referred to in this study as the starred group. The members of the starred group were asked to re- port among other data the names of relatives as follows: Relatives who have done scientific work with ‘designation of relation- ship and direction of work. Relatives (with designation of relationship and direction of work) sufficient to warrant inclusion in a book such as “ Who’s Who,” say among the first twenty thousand of a hundred million popula- tion. Relatives (with designation of relationship and direction of work) who have done scientific work or work of distinction in other directions Polas Cattell, to whom sole credit is due for gath- ering the original d writes concerning requests and replies: Of one thousand one hundred and fifty-four scientific men from whom information in regard to their families was requested 1,036 replied and 118 did not. Of the replies 16 were blank, sometimes ac- companied by the explanation that the information was not readily attainable or the like, 7 were to the effect that the information would be sent later or the like, 13 were received too late, 25 were very im- perfect, 975 were usable and in most cases complete. This is an unusually full reply to a questionnaire. For example, in answer to an inquiry in regard to noteworthy relatives addressed to 467 fellows of the Royal Society, Sir Francis Galton received 207 useful replies, and the completely available returns “scarcely exceeded 100 Following a thorough investigation of that vet of the data concerning relatives, in an attempt to supplement and correct them by the use of biographical and genea- logical handbooks, the writer sent 186 letters to as many of the men of science, with a report of what had been found in the way of additional information, and asked for corrections and additions. Second and third requests were sometimes sent and in some cases a personal visit to the man of science or near relative was made. As a result the number of usable replies for this sue proved to be 956. 8 ‘Families of American Men of Science," Popular Science Monthly, May, 1915. No. 647] - FAMILY RESEMBLANCES 509 Of these 956 records 22 were incomplete in the case of . relatives more remote than parents, children, brothers and sisters. Twenty-three were incomplete in kinships more remote than those mentioned. Fully half of the replies of those found to have distinguished relatives were originally incomplete either in names or designation of relationship of names or both. The brother relation- ship, where it would be supposed complete information would be available, had 84 eases in the original data. This number was raised to nearly 150 through consulta- tion of the handbooks mentioned below. Not more than 25 of those added were found to have first biographical mention at a date later than that of the request for in- formation. It is unlikely that the ten per cent. who failed to reply .did so beeause of lack of relatives to report; it is un- likely, beeause that information was a relatively small part of the total requested. Two hundred and fifty-six, or about one fourth of the number replying, were found to have relatives of distinetion or relatives who were scientific men; since the other three fourths replied, though they had no relatives to report, it seems reasonable that those who did not reply did not represent a select group. This is further shown to be the case in the num- ber of cross relationships between the two groups. There are found to be brother and cousin relationships that were reported by some of those replying that would have been reported if the others had replied. The objective criterion used is biographical inclusion in one or more of the three following handbooks: ** Amer- ican Men of Science," ‘‘ Who's Who in America,’’ ** Ap- pleton’s Cyclopedia of American Biography." Both editions of ‘‘ American Men of Science,’’ that is, the edi- tions of 1903 and 1910, were consulted, and biographies found in either were counted. Those found in the orig- inal edition of ** Appleton's Cyclopedia,’’ published in 1887-88, together with the appendix of 1900 and all ten volumes of ** Who's Who in America," covering the pe- 510 THE AMERICAN NATURALIST [Vor. LVI riod from 1910. to 1918, were also included. The first edition of ‘‘ American Men of Science’’ contains more than 4,000 men of science, of entire North America, the second edition about 5,500 names. *' Appleton's Cyclo- pedia of American Biography (1887-88)"' contains ‘above 15,000 prominent native and adopted citizens of the United States, including living persons, from the earli- est settlement of the country." In the appendix of 1899-00 ‘‘ will be: found nearly 2,000 notices of Ameri- cans who won renown in the war with Spain . . . and of persons of the New World who have become prominent in the peaceful activities of life during the decade,’’ be- tween the appearance of the two publications.. The ten volumes of ‘‘ Who’s Who in America’’ contain 36,915 biographical sketches. The first volume contains 8,602 biographies, while Volume 10 has 22,968. It is evident that the three publications have varying standards of selection, and it becomes necessary to get some statement of the degree of fineness of selection represented by each. If the reader doubts the validity of any one of the three measures he may disregard those found in that handbook because the lists of names and tables are arranged to that end. That there are biographies of persons included that are out of place is likely and that omissions of others quite deserving occur is also likely, but inclusion repre- sents unusual performance that is a reality. There is given below the biographical account of one of the persons in the study as it is given in the three dif- ferent handbooks. Besides adding reality to the data in the lists it will afford a comparison of the characteristic methods employed by the editors of the different publi- cations. The accounts give some idea of the interesting and voluminous records that would be necessary if no more than a brief history of each individual were given. The histories of the men of science and their relatives, if abbreviated in the most careful manner, would make a fair-sized volume. One need only imagine the size of the volume necessary to give an account of the unusual per- No. 647] ` FAMILY RESEMBLANCES 511 formance of an equal number of people taken at random to see the difference. BIOGRAPHICAL Account or EDWARD CHARLES PICKERING, AMERICAN MEN or SCIENCE, 0 Piekering, Prof. Edward C(harles), Harvard College Observatory, Cambridge, Mass. * Astronomy, Astrophysics. Boston, Mass, July 19, 46. B.S, Harvard, 65, A.M, 80, LL.D, 03; California, 86; Michi- gan, 87; Sc.D, Victoria (England), 00; LL.D, Chicago, 01; Ph.D, Heidelberg, 03; LL.D, Pennsylvania, 06. Instr. math, Lawrence Sci. Sch, Harvard, 65-67; prof. physics, Mass. Inst. Tech, 67-77; astron. and director, Harvard Col. Observatory, T7- Bruce Gold Medal, Pacific Astron. Soc; Rumford, Draper, Bruce, two Royal Astron. Soc. and other medals. Nat. Acad; F.A.A. (v. pres, 77); Astron. and Astrophys. Soc. (pres, 06-08); Philos. Soc. (v. pres, 09) ; fel. Am. Acad; Wash. Acad; hon. mem. N. Y. Acad; cor. mem. Berlin Acad; Soe: astron. de France; Inst. de France; Royal Soe. ‘Upsala; Soc. Lynceorum Nova; St. Petersburgh Imp. Acad; Socie- ties of Cherbourg, Palermo, etc. Stellar photometry and spectros- copy. ACCOUNT GIVEN IN WHo's Wuo 1n AMERICA, Vor. 6, 1910-11 Pickering, Edward Charles, astronomer; . b.. Boston, July 19, 1846. s. Edward and Charlotte (Hammond) P; brother of William Henry P. (q.v.) ; ed. Boston Latin School; S.B, Lawrence Scientific Sch. (Harvard), 1865 (hon. A.M., 1880; LL.D., univs. of Cal, 1886, Mich., 1887, Chicago, 1901, Harvard, 1903, Pa., 1906; Ph.D., Heidel- berg, 1903; D.Se., Victoria U., Eng., 1900; m. Lizzie Wadsworth, d. Jared Sparks, Mar. 9, 1874. Instr. mathematics, Lawrence Scien- tifie Sch., 1865—7; Thayer prof. physies, Mass. Inst. Tech., 1867-76; prof. astronomy and dir. Harvard Coll. Obs. since 1876. Estab- lished 1st physical lab. in U. S.; under his direction, invested capi- tal and income of the observatory has increased fourfold. Study of light and spectra of the stars have been spl. features of his work; devised meridian photometer and made 1,400,000 measures of the light of the stars with it. By establishing an auxiliary sta. in Arequipa, Peru, Southern stars are also observed, extending the work from pole to pole, in which 200,000 photographs are included. Accompanied Nautical Almanac expdn. to observe total eclipse of sun, Aug. 7, 1869; mem. U. S. Coast Survey expdn. to Xeres, Spain, Dee. 22, 1870. Awarded Henry Draper medal for work on astron. physies; gold medals, Rumford, 1891, Bruce, 1908, Royal Astron. Soc., 1886, 1901. Mem. Nat. Acad. Sciences; hon. mem. of Socs. at Mexico, Cherbourg, Liverpool, Toronto, Upsala and Lund; mem. Royal Astron. Soc., Royal Instn. Acaad. dei Lincei, Royal Prussian, and Royal Irish socs., Royal Soc. of London, Institute de France, Imperial Acad., St. Petersburg; pres. Astron. and Astrophys. Soc. 512 THE AMERICAN NATURALIST | [Vor. LVI America, 1906-9; fellow Am. Acad. Arts and Sciences; founder and 1st pres. Appalachian Mountain Club; mem. Century Assn. New York. Author: Elements of Physical Manipulation, and 60 volumes of annals and other publieations of Harvard Coll. Observatory. Address: Harvard Observatory, Cambridge, Mass. ACCOUNT GIVEN IN APPLETON'S CYCLOPEDIA OF AMERICAN BIOGRAPHY, Pickering, Edward Charles, astronomer, b. in Boston, Mass., 19 July, 1846, was graduated in civil engineering course at the Law- rence scientific school of Harvard in 1865. During the following year he was called to the Massachusetts institute of technology as assistant director of physics, of which branch he held the full professorship from 1868 till 1877. Prof. Pickering devised plans for the physical laboratory of the institution, and introduced the experimental method of teaching physics at a time when that mode of instruction had not been adopted elsewhere. His scientific work of these years consisted largely of researches in physics, notably investigations on the polari- zation of light and the laws of its reflection and dispersion. He also described a new form of spectrum telescope, and invented in 1870 a telephone receiver, which he publicly ‘exhibited. He observed the total eclipse of the sun on 7 Aug., 1869, with the party that was sent out by the Nautical almanac office, at Mt. Pleasant, Iowa, and was a member of the U. S. coast survey expedition to Xeres, Spain, to observe that of 22 Dec., 1870, having, on that occasion, charge of the polariscope. In 1876 he was appointed professor of astronomy and geodesy, and director of the observatory at Harvard, and under his management this observatory has become one of the foremost in the United States. More than twenty assistants now take part in investigations under his direction and the invested funds of the observatory have inereased from $176,000 to $654,000 during his administration. His principal work since he accepted this appoint- ment has been the determination of the relative brightness of the stars, which is accomplished by the means of a meridian photometer, an instrument specially devised for this purpose, and he has prepared a catalog giving the brightness of over 4,000 stars. Since 1878 he has also made photometer measurements of Jupiter’s satellites while they are undergoing eclipse, and of the satellites of Mars and other plication of photography to astronomy, as a Henry Draper memorial, and the study of the spectra of the stars has been undertaken on a scale that was never before attempted. A fund of $250,000, left w Uriah A. Boyden (q.v.) to the observatory, has been utilized for the special study of the advantages of very elevated observing stations. Prof. Pickering has also devoted attention to such objects as moun- tain surveying, the height and velocity of clouds, papers on which he No. 647] FAMILY RESEMBLANCES 513 has contributed to the Appalachian Club, of which he was president in 1877, and again in 1882. He is an associate of the Royal astronomical society of London, from which in 1886 he received its gold medal for photometrie researches, and besides membership in other scientific societies in the United States and Europe he was elected in 1873 to the National academy of sciences, by which body he was further honored in 1887 with the award of the Henry Draper medal for his work on astronomical physics. In 1876 he was elected a vice-presi- dent of the American association for the advancement of science, and presented his retiring address before the section of mathematics and physics at the Nashville meeting. In addition to his many papers which number about 100, he prepared “ Reports on the Department of Physies," for the Massachusetts institute of technology, and the “ An- nual Reports of the Director of the Astronomical Observatory," like- wise editing the “ Annals of the Astronomical Observatory of Harvard College." He has also edited with notes * The Theory of Color in Relation to Art and Art Industry," by Dr. William von Bezold (Bos- ton, 1876), and he is the author of * Elements of Physieal Manipula- tion" (2 parts, Boston, 1873-6). The raw material of the investigation is found below in the lists of men of science and their relatives. Sta- tistieal treatment of these data will follow. "They are arranged so that any competent observer can test their validity. While great effort has been made to have the details reliable, it is possible that mistakes may be found in the designation of relationship, but it is thought that such mistakes are few if any. Tue MEN or SCIENCE AND THEIR NEAR RELATIVES OF DISTINCTION The names of the men of science and their near rela- tives of distinction are given at the left of the page. The name of the man of science comes first and is followed by a short dash. The names of the relatives follow and are preceded by the letter or letters which tell the relation- ship. The first name, for example, is Allis, Edward Phelps; he has a cousin (FSiS, a father's sister's son), Callahan, Henry White, whose work is in edueation. The biography of this relative is found in ** Who's Who in Ameriea." When the name of a relative is printed in small capitals it shows that the person is known for work 514 THE AMERICAN NATURALIST [Vor. LVI in science. The direction of the performance of the rela- tives is given at the right of the name. The handbook or books containing his biographical account are shown by the abbreviations at the right. These abbreviations are to be understood as follows: A.M.S, American Men of Science (if preceded by an asterisk the person is in the starred group of men of science); W.W, Who’s Who in America; A.C, Appleton’s Cyclopedia of American Biography. | Any degree of relationship can be conveniently and ac- eurately deseribed by one, or a combination of two or more, of the following seven letters: S for son, D for daughter, B for brother, Si for sister, F for father, M for mother. Thus, paternal grandfather can be written as FF, meaning father's father, maternal grandfather, MF, meaning mother's father, FBS meaning first cousin or father's brother’s son, etc. These symbols precede the name and show the kinship of the person to the man of Science. The men of science are arranged according to the sci- ence in which they were recorded as obtaining a place in the starred group. The groups are arranged alphabeti- eally, beginning with the anatomists and closing with the zoologists. penc fet ge P ists nr qus A eren ud and near relatives i of distinction sadeqene pt aP of relatives Anatomists Allis, Edward Phelps— : FSiS Callahan, Henry White Education W.W. Bardeen, Charles Russell— Bardeen, Charles William Education Ww — Robert Russe Bensley, Beijaitk Arthur Zoology A.M.S. TER Thomas— WARREN, JONATHAN MASON Surgery A.C. XY WARREN, JOHN COLLINS Surgery A.C. MBS WARREN, JOHN COLLINS Surgery *A.M.S; W.W: * AC FBS Dwight, Wilder Warfare A.C. FBS Dwight, William Warfare W.W. No. 647] FAMILY RESEMBLANCES Greenman, Milton Jay— FBS GREENMAN, JESSE MORE "e Meyer, Arthur Will B Meyer, Balthasar Henry Polit. econ, Spitzka, pate Anthony— F TZKA, Epwarp A, Neurology Anthropologists Dorsey, George Amos— Dorsey, CLARENCE W. Soil physies B DORSEY, HERBERT GROVE Physics Farrand, Livingston— Farrand, Max History B Farrand, Wilson Edueation Hough, Walter— FBS Hoven, THEODORE Physiology Astronomers Doolittle, Charles Leander— DOOLITTLE, Eric Astronomy Doolittle, Erie— DOOLITTLE, CHARLES L. Astronomy Frost, Edwin Brant— B Frost, GILMAN DuBois Anatomy Lowell, Pereival— Lowell, Abbott Lawrence Education Si Lowell, Amy Poetry MF Lawrence, Abbot Diplomaey FSiD Paid Ella Lowell Lyman Edueation MSiS R , ABBOTT LAWRENCE Physies MSiS otek, Asme Arehitecture Pickering. Edward Charles— B PICKERING, WILLIAM H. Astronomy FB ER CHARLES Ethnol; Bot. Pickering, William Henry— B PICKERING, EDWARD C. ronomy FB PICKERING, CHARLES Ethnol; Bot. Pritchett, Henry Smith— F Pritchett, Car Waller Ministry; Educ. 515 *A.M.S; W.W. W.W. A.M.S; A.C. WAW, A.M.S. W.W. W.W. *A.M.8; W.W. *A.M.S; W.W; A.C. A.C. * A.M.S. A.C. W.W. 516 THE AMERICAN NATURALIST [Vor. LVI Searle, Arthur— Astron; Religion A.M.S; W.W; B SEARLE, GEORGE MARY A. Stone, Ormond— Journ; Finance B Stone, Melville Elijah W.W AC, Wright, William Hammond— M Wright, Johanna Maynard Organization W.W. Botanists Beal, William James— MSiS STEERS, JOSEPH BEAL Zoology A.M.S; W.W. Bessey, Ernst Athern— F BESSEY, CHARLES EDWIN . Botany ; *A.M.S; WW. Bessey, Charles Edwin— S BESSEY, ERNST ATHERN Botany *A.M.S; WW, Blakeslee, Hoe Franei B Blakeslee, assent Hubbard History W.W. Blakeslee, Franeis Durbin Ministry W.W. Bray, Wil'iam— MBS or FSiS Foster, LUTHER Botany W.W. Campbell, aii xiu ms B CAM ARD DEM. Chemistry *A.M.S; W.W. B wii Mus Munroe Law W.W F Campbell, James V. Law A.C. Coker, William Chamber Coker, James Lide Manufacturing W.W. FBS COKER, ROBERT IRWIN Zoology A.M.S; W.W. Coulter, John Merle— B COULTER, STANLEY Botany *A.M.8; WW. S COULTER, JOHN GAYLORD Botany A.M.S. FBS Counter, SAMUEL Monps Botany A.M.S. Coulter, Stanley— B ULTER, JOHN MERLE Botany . *A.M.S; W.W; A.C. BS COULTER, JOHN GAYLORD Botany A.M.S. FBS CouLTER, SAMUEL MoNDs Botany A.M.S. Coville, Fred Vernon— B COVILLE, LUZERNE Medicine oe v m. Davis, Bradley Moore— FSiS Woop, RoBERT WILLIAMS Physics "AMB: W.W. No. 641] Duggar, Benjamin Minge— B DUGGAR, JOHN FREDERIC Earle, Franklin Sumne i orne, Mig n Earle F EARLE, PARKER Fernald, Merritt Lyndon— B FERNALD, ROBERT H. F FERNALD, MERRITT C. Greenman, Jesse More— REENMAN, MILTON JAY Halstead, pides David— SiS RCHILD, DAVID G. SiS pma Edwin Milton Kearney, Thomas H.— i MB Miner, Charles Wright ` Maebride, Thomas Huston— FSiS Sterrett, James M. Mec Gifford— ine FAMILY RESEMBLANCES Agronomy Fietion Horticulture Mech. eng. Edue; Physics; Math Anatomy Botany Education Warfare Ministry; Hist. hot, Amos Law; Politics * Pinehot, James W. Trade MF Eno, Amos R. Finance f Pound, Roscoe— Si Pound, Louise Philology Robinson, Benjamin Lincoln— obin ecd James H. History Shull, George Har B SHULL, eae ALBERT Zoology B Scholl, John W Literature i Chemists Acheson, Edward Goodrich— FB eson, Marcus Wilson Law FBS Acheson, Alexander M Civil Eng. FBS Acheson, Alexander W. Med; Politics FBS Acheson, Ernest Francis Journal; olities FBS Acheson, Marcus W. Jr. Law FSiS Brownson, Marcus A, Ministry FSiD Brownson, Mary Wilson Literature; Math pady Wilder Dwight— Bancroft, George History 517 AMS; W.W. W.W. A.C. A.M.S; W.W. A.M.S; W.W. *A.M.S; W.W. *A.M.S; W.W. W.W W.W. 518 THE AMERICAN NATURALIST [Vou. LVI Blair, Andrew Alexander— F Blair, Francis Preston Warfare; Polities A.C. FF Blair, Francis Preston Journal; Polities A.C. Burgess, Charles Frederick— Burgess, George H. Railway Eng. W.W. Campbell, Edward DeMille— B CAMPBELL, DovGLAs H. Botany "AMS: WW. B Campbell, Henry Munroe Law W.W. F Campbell, James V. Law A.C. Chatard, Thomas Marean— B Chatard, Franeis Silas Edue; Ministry; W.W; A.C, FB Chatard, Frederick Warfare A.C. Crafts, James Mason— ` Mason, Jeremiah Law AC. Dabney, Charles Willia F Dabney, ober: Lewis Educ; Ministry A.C. Doremus, Charles Avery— : F DongEMUS, ROBERT O. Chemistry A.M.8; W.W; FM Doremus, S. P. (Haines) Philanthropy A.C. iege ie Franeis Perry— : Keener, John C, Ministry : WW; AC. , Franklin, Edward Curtis— B FRANKLIN, WILLIAM f. Physies *A.M.S; W.W. Freer, Paul Casper— B Freer, espa W. Art (Painting) W.W. B Prem, OTTO Tie Laryngology W.W. are Eugene Woldemar— ARD, JULIUS Math; Geodesy A.C. HILGARD, THEODORE C. Biology A.C, F HILGARD, THEODORE E. Law A.C. SiS Tirrman, OTTO HILGARD Geodesy A.M.S; W.W. Jackson, Charles Loring— FSi Lowell, Anna C. J. Education A.C. FF Jackson, Patrick Tracy Manufacturing A.C. i tles MF FSiS CABOT, ARTHUR Tracy Surgery . MBS Loring, William Caleb Law W.W. FSiS Lowell, C. R. Warfare A.C. Lewis, Gilbert Newton— FB wis, Homer Pierce . Education W.W. No. 641] Lloyd, John Uri— B LLOYD, Curtis GATES Loeb, Morris— Loeb, James More, Riehard Bishop— F Moore, William Thomas MF Bishop, Richard Moore Morely, Edward Willia B Moreley, John Sidós ae Charles Edward— BS Munroe, James Phinney ne Musrot, Kirk Norris, James Flaek— B Norris, RICHARD C. Norton, Thomas Herber MB HORSFORD, "uu NORTON MF Horsford, Jerediah MBD HORSFORD, CORNELIA FBS Norton, LEWIS MILLS Noyes, -William Albert— i Davidson, Hanna M. N. Orndorf, William Ridgely— MF Ridgely, James Lot Osborne, Thomas Burr— MB BLAKE, ELI WHITNEY MF Blake, Eli Whitney Palmer, Chase— FSiS Harris, George MBS Chase, George Pellew, Charles Ernest— F Pellew, ey Edward MB Jay MF J M, iene MBS Jay, William Pond, George Gilbert— iB OND, FRANCIS JONES Reese, Charles Lee— ic Reese, Frederick Focke Riehards, Theodore William— B RICHARDS, HERBERT M. F Richards, William Trost FAMILY RESEMBLANCES Botany Finance; Archeol, Ministry; Editing ae H Polities; Trade Ministry; Edue. Mfg; Writing Fietion; Journal Surgery Chemistry Wa Are Chemistry Edueation; Lit. Law Chem; Physies Invention; Mfg. Ministry; Lit. Edue; Law Philanthropy Diplomacy Law Law Chemistry Ministry Botany ; Art (Painting) 519 W.W. WW, WW. W.W. W.W. *A.M.S; W.W. W.W; A.C. 520 Sadtler, Samuel Phili F dtler, Benjamin MB Sehmucker, B. Melanehton MB Schmucker, Samuel M MB Schmucker, Samuel D. MF Sehmucker, Samuel S. MBS ScHMUCKER, SAMUEL C. Sanger, Charles Robert— Sanger, George P. Saunders, inis Perey B SAUNDERS, ice. E. B SAUNDERS a Ai B Savausas, WILLIAM F SAUNDERS, duin Sherman, Henry Clapp— B SHERMAN, FRANKLIN, JR., FBS Sherman, Frank D. Shimer, Porter William— F HIMER, HERVEY W. Smith, Edgar Fahs— SMITH, ALLEN JOHN Stieglitz, Julius Osear— Stieglitz, Alfred Stillman, John Maxon— STILLMAN, STANLEY FB Stillman, Thomas Bliss FB Stillman, William James FBS STILLMAN, THOMAS B Stillman, Thomas Blis FB Stillman, TRAMA Bliss Stillman, William James FBS STILLMAN, JOHN Maxon FBS STILLMAN, STANLEY Tuckerman, Alfred— MB Gibbs, George MB GisBs, OLIVER Wo.corr FF Tuckerman, Joseph MF GIBBS, Gro E FSiS Broke, GEORGE F. VanSlyke, Lucius Lincoln— S ANSLYKE, DoNALD D. THE AMERICAN NATURALIST [Vor. LVI Ministry A.C Ministry A.C. Writing A.C. Law W.W. Educ; Theol. A.C. Botany A.M.S; W.W Law A.C Chemistry A.M.S. ee * A.M.S; W.W Ornith; B A.M.S. naire A.M.S Entomology M.S. Architect; Poetry W.W. Geology A.M.8; W.W. Pathology A.M.S; W.W. Photog; Chem. W.W. Surgery W.W. Manufacturing A.C. History; Journal W.W; A.C. Chemistry *A.M.S; W.W. Manufacturing A.C. History; Journal W.W; A.C. Chemistry *A.M.8; W.W: Surgery Ww History W.W. Warfare A.C. Antiquarianism A.C Chemistry *A.M.8; W.W; : A.C. Ministry A.C Geology A.C. Geology *A.M.S; W.W Chemistry A.M.S. No. 647] Venable, Francis Preston— FAMILY RESEMBLANCES F VENABLE, CHARLES S. Astronomy MF McDowell, James iti Waller, Elwyn— B Waller, Frank Architecture Geologists Ashley, George Hall— B Ashley, Roscoe Lewis Becker, George Ferdinand— MSiS Tuckerman, Bayard Brooks, Alfred Hulse— Brooks, THOMAS B. Si Paige, Gaspard B. Chamberlin, Thomas Corwin— S CHAMBERLIN, ROLLIN T. Clarke, John Mas B Clarke, ond Mason Dana, Edward Salisbury— Dana, JAMES DWIGHT MB SILLIMAN, BENJAMIN JR. MF SILLIMAN, BENJAMIN Davis, William Morris— Mott, James MM Mott, Lucretia Farrington, Oliver Cummin B a ngs— ington, ` Wallace R B Tüsxmeron, Regents Grove Kar Gilbert, e: Sheldon Grant, Ulysses Sherman— F Grant, Lewis Addison Hague, Arnold— B HAGUE, JAMES DUNCAN F Hague, William Harris, Gilbert Dennison— B HARRIS, ROLLIN ARTHUR Epwarp H. History; Econ. . Ministry Chemistry Biography; Hist eology Fiction Geology Ministry Geology Chemistry Chem; Geol. Philanthropy Min inistry (Quaker) Journalism ; Chemistry Art Law; Warfare Geology Ministry Geodesy 521 W.W; A.C. A.C. W.W; A.C. *A MS; W.W. 522 Hayes, Charles Willard— UM LLEN Hitcheock, staples Henry— F TCHCOCK, EDWARD B adu EDWARD BS HITCHCOCK, Irving, John Duer— F Irvine, ROLAND DUER Jaggar, Thomas Augustus, Jr.— F Jaggar, Thomas A Keith, Arthur— MSiS GALE, Hoyr STODDARD Mathews, a, Bennett— B Mathews, Shailer Merriam, John Campbell— B Merriam, Charles E. Merrill, Te Perkins— B RILL, Lucius H. Penrose, Richard Alex. Fullerton— B Penrose, Boies B PENROSE, CHARLES B. B Penrose, Speneer F PENROSE, RICHARD A. F, .FB Penrose, Clement B. FF Penrose, Glasi B. FBS Penrose, Stephen B. L. Pumpelly, Raphael— D Smyth, Margarette P. M Pumpelly, Mary Weller Rice, William North— S ICE, EDWARD LORANUS € yya aa , HUGH LENN io Hole punte Fins MF Hodge, Charles Smith, James Perrin— B Smith, Charles Forster Stevenson, John Jam MB Wil MBS Willson, David Burt EDWARD, JR. THE AMERICAN NATURALIST Mathematies Geol; Edue. Hygiene Hygiene Geology Ministry Bel Edue; Theol. Polities; Hist. Chemistry Polities ; Law Law Law Edue; Philos. Art (Painting) Poetry eo Theology Philology Philol; Theol, Philol; Theol. Philol; Theol. [Von. LVI W.W; A.C. W.W. W.W. A.M.S; W.W. W.W; A.C, No. 647] FAMILY RESEMBLANCES Taylor, Frank Bensley— F Stewart, Robert Vaughn, Thomas Wayland— FBS "Vaughn, Horaee Worth Weed, brun Harvey— Weed, Samuel Riehards Weller, Stewart— Tarran, J. T. White, David— MSiS Kent, Charles Foster Willis, Bailey— F Willis, Nathaniel P. i Poetry FB Willis, Richard Storrs Music Winchell, Alexander Newton— B WINCHELL, Horace V. Geology F WINCHELL, NEWTON H. Geol; Archeol. FB CHELL, ALEXANDER Geology FB Mise Bando R. Edue; Journalism Winchell, Newton Horace— B -.WINCHELL, ALEXAND Geology B Winchell, Samuel R, uc; J ournaliam S WiNCHELL, ALEX."N. Geology S WINCHELL, HoRACE V. Geology FBS Winchell, Benj. La Fon Ry. Management FB Winchell, James M Ministry Wright, id pies Daneel B WRIGHT, CHARLES WILL Geology Mathematicians Birkoff, George David— MB Droppers, Garrett Polit. Econ. Coolidge, Julian Low B Co Wie eae ibald C. History B Coo. , John Gardiner Diplomacy B Coalides: J. Randolph Jr. Architecture FB Coolidge, Jeffe Diplomaey FBS Coolidge, T. Jefferson Jr. Finance Fine, Henry Burchard— FF Fine, John Polities; Law Polities; Law Polities; Law Finanee; Lit. Law Hist; Archeol. 523 524 THE AMERICAN NATURALIST (Vor. LVI Franklin, Fabian— Heilprin, Michael History; A.C, Sociol. MF Heilprin, Phineas M. Semities x. MBS HEILPRIN, ANGELO Geology FANS Wows : A.C, MBS Heilprin,. Louis Philology : W.W; A.C. Halsted, George Bruce— : F Halsted, Oliver Spencer Politics A.C. FF Halsted, Oliver Spencer Philology A.C. Johnson, William Woolsey— ent B Johnson, cae Fred. Philology; Math. W.W. McClintock, Emory— McClintock, John Educ; Ministry A.C. T Eliakim Hastings— Moore, David Hastings Educ; Ministry W.W. Pierce, Charles Santiago Sanders— eirce, Herbert H. D. Diplomaey W.W. B PEIRCE, JAMES MILLS Math; Edue. 1MB: WOW, ; A.C F PEIRCE, BENJAMIN Mathematies A.C, -FB PEIRCE, CHARLES HENRY Chem; Med. A.C. FF Peiree, Benjamin Lines A.C. (Harvard) MF Mills, Elijah Hunt Polities A.C. Roe, Edward Drake Jr— FB oe, Franeis Asbury Warfare W.W; A.C. Slichter, Charles Summer— BS SLICHTER, WALTER I, Elee. Eng. A.M.8; W.W. Van Vleek, Edward Burr— VAN VLECK, JoHN M. Astron; Math. A.M.8; W.W; A.C. Veblen, Oswald— F VEBLEN, ANDREW A. Math; Physies A.M.S; W.W. FB Veblen, Thorstein B. Economics W.W Wilson, Edwin Bidwell— FB Wilson, Frank E, Polities; Law W.W. Pathologists Biggs, Herman Miehael— : FBS Bices, GEORGE PATTEN Pathology A.M.S, Blumer, George— LUMER, GEORGE ALDER Neurol; Med WVW. No. 647] Cabot, "ise Clarke— FBS FBS cues GODFREY L. Christian, Henry Asbur FBS Christian, dhyo L. pnus pond Williams— Cus Y SiS CREHORE, WILLIAM Dana, Charles Loomis— Dana, John Cotton Dock, George— Si Dock, Lavina L. Ernst, Harold Clarence— rnest, George um B ERNST, OSWALD MF OTIS, GEORGE ALEX. Flexner, Simon— Flexner, Abraham Hurd, Henry a jB Hur Loeb, Leo— B pup teur William George— MacCAL ` MACCALLUM, GEORGE A Mitchell, Silas Weir— S M F Musser, John Herr— MBS or FSiS Herr, Edwin M. Park, i Halloek— Hallock, William S. MSiS Johnson, William H. Putnam, James J.— Jackson, James LOEB, JACQUES ITCHELL, JOHN K. S Mitehell, Langdon E. MrTCH JOHN K FAMILY RESEMBLANCES Med; Surg. Chemistry Polities; Lit; ar Geology Law Physies; Eng. Mech. Eng. Library Medieine Law Astron; War; Eng. Surgery Edueation Psychiatry Zoology Anatomy Ornith; Med. Neurology Playwriting Chem; Med Elec. Eng. Editing; Ministry Ministry ; Philos, Clin, Med. 525 W.W. W.W. W.W. *A.M.8; W.W. *A.M.S; WW. W.W. AMS; W.W. A.C. W.W. 526 Ravenel, Mazyek Porcher— FBS RAvENEL, HENRY W. MBS PorcHer, FRANCIS P. Thayer, William Sydney— : B pach Ezra Ripley F Thayer, James Bradley MSiS titus Edward Warren, John Collins— F WARREN, JONATHAN M. FF WARREN, JOHN suai FSiS Dwicut, THOM Williams, Herbert Upha Si Williams, iis, _ Sprague FSiS Sprague, Carle Welch, William Henr THE AMERICAN NATURALIST Botany Botany; Chem. Law aw Art (Painting) Surg; Med. Surgery Anatomy Sociolo Finance; Art FSiS Cowles, John Guiteau Finanee MBS Collin, Frederick Welch MBS Collin, Charles Avery Law Physicists E Cleveland— A " ABBE, CLEVELAND, JR. S ABBE, M FSiS Smith, Guilford Abbot, Charles Greeley— FBS Abbot, Henry Larcom FBS Abbot, Edwin Hale Bauer, Louis Agricola— B Baver, WILLIAM C. Bell, Alexander Graham— F EX, MELVILLE Bell, Louis— F Bell, Louis FB Bell James FB ELL, JOHN FB BELL, LUTHER V FB Bell, Samuel Dana MB Bouton, John Bell MF ues Nathaniel FF Bell, uel FBS Bell, Peas Newell Surg; Physics Meteorol; Geog. siol. urg; inanee; Philanth. Eng; Warfare Finance; Law Elec. Eng. Physiology Warfare; Chem. Polities; Law Editing; Med Medicine L Polities; Law A.C. A.C. WW. [Vor. LVI No. 647] Buckingham, Edgar— FF Buckingham, Joseph T. Crehore, Albert Cushing— B CREHORE, WILLIAM W. MSiS Cusuine, Harvey W MSiS Cusuine, HENRY PLATT MSiS Cushing, William E. z gne Harvey Nathaniel— AVIS, NATHANIEL F. Duane, William— B Duane, Russel FF Duane, William Franklin, William Suddards— B FRANKLIN, EDWARD C. Hering, Carl— B HERING, RUDOLPH Humphreys, William Jackson— FB Humphreys, Milton W. Ives, Frederick Eugene— S Ives, HERBERT EUGENE Jackson, Dugald Caleb— B JACKSON, JOHN PRICE B Jackson, William B. FSiS Cravath, Paul Drennan Kent, Norton Adams— B Kent, William Kimball, Arthur Lalann B MBS Fisher, Samuel S., Jr. Kinsley, Carl— F Kinsley, William W. Lyman, Theodore— F LYMAN, THEODORE FF Lyman, Theodore Magie, William Franeis— F Magie, William Jay Mann, Charles Reborg— F arles H. FSi Miller, Harriet Mann FAMILY RESEMBLANCES Editing; Publish. Bridge Eng. Pathology Geology Mathematics Law Publishing Chemistry Hydraulic Eng. Philology Physics Elec. Eng. Engineering Law ; Philanthropy Finance inistry ; M Educ. Politics; Law Math; Theology Zoology Philanthropy Polities; Law Inventing; Editing; „Ministry Ornith; Writing 527 A.M.S; W.W. W.W. A.C. *A.M.8; W.W. AMS. W.W; AC. W.W. A.C. A.C. W.W. W.W. W.W. 528 Mendenhall, Thomas Corwin— MENDENHALL, C. E. Mendenhall, Charles Elwood— MENDENHALL, T. C More, Lewis Trenchard— B More, Enoch Anson, Jr. B More, Paul Elmer MF Elmer, Lueius Q. C. Northrup, Edwin Fite B Northrup, Eos J ris F Northru sa nsel J FB NongTH e. pi MB Fiteh, nie us Parson, William Barclay— B Parsons, Harry DEB. Roteh, Abbott Lawrence— B teh, hur ot _MF Lawrence, Hei FBS Rorcu s M. MSiS Lowell, cum posce MSiS LOWELL, PERCIVAL MSiD Lowell, Amy Saunders, Frederick Albert— B SAUNDERS, ARTHUR P. B SAUNDERS, CHARLES E. B SAUNDERS, WILLIAM E. F SAUNDERS, WILLIAM Stevens, Moss usps , JOHN MB one JOSEPH MF LECONTE, LEWIS MBS LECONTE, JosEePH N. Stewart, Osear Milton— STEWART, GEORGE W. Stewart, George Walter— STEWART, Oscar M. Trowbridge, Charles Christopher— Trowbridge, S. B. P. F TROWBRIDGE, W. P. THE AMERICAN NATURALIST Physics Physics Fiction oetry; Editing Polities ; Law L Pathology Editing; Edue. Mech. Eng. Architeeture Diplomaey Pediatries Edue. Astronomy Poetry Chemistry Chemistry otany Horticulture Physies Geology Botany Mech. Eng. Physies Physies Architecture Engineering Law *A.M.S; W.W. *A M.S; W.W; A E eH W.W. W.W. A.C. .W. W.W; A.C A.M.8; W.W WwW. A.M.8; W.W A.C AC. "ANB. W.W *A.M.S; WW A.C. W.W. *A.M.8; W.W A.M.S A.M.S A.M.S, A.C. W.W; AC A, A.M.S; W.W *A.M.S; W.W "AMB; W.W. W.W. A.C. [Vor. LVI No. 647] FAMILY RESEMBLANCES 529 Very, Frank tiens FB ery, Jones Lit; Ministry A.C, FSi Very, sn Louisa A. Literature W.W: AC. Wead, Charles Dasson— MB Kasson, John Adams Polities WW; A.C. Whitman, Frank P MSiS Taylor, piget Monroe Edue; Ethies W.W; AC. MSiD Bissell, Mary Taylor Medieine W.W. Wood, Robert Williams— MBS Davis, BRADLEY Moore Botany *A.M.S; W.W Wurts, Alexander Jay— B Wurts, John W.W MF Jay, JOHN CLARKSON Medieine A.C Pen John— ZELENY, ANTHONY Physics *A.M.8; W.W: : ZELENY, CHARLES Zoology *A.MB; W.W. Zeleny, Anthony— B ZELENY, CHARLES Zoology *A.M.S; W.W. B ZELENY, JOHN Physies *A.M.S; W.W. Physiologists Curtis, John wore Curtis, Edward Medicine W.W; A.C B Curtis, sans Bridgham Warfare; Eng. .C. 3B ien xam i William Lit; nblis shing W.W. B , Constan Art (Painting) W. BD du MU Music W. Dawson, Perey Millard— F Dawson, Samuel quede" FSiS Apams, FRANK Daw Hare, Hobart Amory— F Hare, William Hobart FF Hare, George Enlen MF Howe, Mark A. DeWolfe MBS Howe, Mark A. DeWolfe Henderson, Yandell— MB YANDELL, DaviD WENDEL MF YANDELL, LUNSFORD Hough, Theodore— FBS Hoven, WALTER Hist; Publishing A.C, Geology A.M Ministry A.C, Ministry. AG. . Ministry W.W; A.C. Editing W.: Surgery A.C, Geology; Med, A.C, Anthropology *A.M.8; W.W. 530 THE AMERICAN NATURALIST [vor LVI Lee, - Schiller— LEE, LESLIE ALEXANDER re Geo *A.M.8; W.W. 3 Lee, John Clarence dici W;W F Lee, John Stebbins dee Ministry W. W. pea panera Lusk, WILLIAM T. Physiol; Med. A.C, de nds. inb HM. Finanee; AC. Polities Sewall, Henry— F SEWELL, THOMAS Medicine — ALC. Psychologists Angell, James Rowland— B Angell, Alexis Caswell Law W.W. F Angell, James Burrill Educ; W.W; AC. ; Diplomacy MF CASWELL, ALEXIS Astron; Educ. A.C. FBS ANGELL, FRANK Psychology *A.M.S; W.W. Angell, Frank— FB Angell, James Burrill Educ; W.W; AC. a FBS Angell, morte Caswell W.W. ANGELL, JAMES R. Psychology *A.M.S; WW. Bently, Madison— ` ; ; Bentiy, Charles E. Ministry W.W; A.C. sie William Lowe— Bryan, Enoch Albert Edueation W.W. Cattell, ~ a a B NRY WARE Pathology A.M.S; W.W. F pa nee wii Educ; Ministry A.C. FB Ca ttell, Merce ` Finance; A.C. Polities FBS Cattell, Edward James Econ; Geog; WW. Lit. . FBS Cattell, William A. Engineering W.W. Delabarre, Edmund Burke— : Delabarre, Frank Alex. Med; Dentistry W.W. Dewey, John— B Dewey, Davis Rich Econ; Statisties W.W. Hall, Granville Stanley— MBS Brats, EpwArp ALDEN Meteorology WW: No. 647] FAMILY RESEMBLANCES Jastrow, Joseph— B Jastrow, Morris Theol; Sociol. F Jastrow, Mareus Philology Patrick, George Thomas White— Patrick, Mary Mills Edue; Writing sages Edmund Clark— SiS SHINN, CHARLES H. Botany xb SHINN, MrinicENT W. Psychology Stratton, George Maleolm— B Stratton, Frederick S. Law Strong, Charles Augustus— F Strong, Augustus H. Educ; Theology Thorndike, Edward L B Thorndike, one H. Philology Wells, Frederie Lyman— F Wells, Benjamin Willis Language; Econ. Woodworth, Robert aani 4B Woodworth , Frank G. Educ; Ministry MB Sessions, nus. R. Polities; Agrie. Zoologists Andrews, genes Allen— B DREWS, HORACE Civil Eng. FF Tct Ethan Allen Lexieography Bruce, Charles Thomas— Armstrong, William Musical Criticism 18i Benough, Elisa A. Fiction Clark, Hubert Lyman— F CLARK, WILLIAM 8. Chemistry MF Richards, Willia Education MBS WILLISTON, pinnis L. Mech. Eng. MBS Williston, Samuel : Crampton, Henry Edward— MB Miller, Charles Henry Art (Painting) Dahlgren, Ulrie— FB Dahlgren, Ulric arfare FF DAHLGREN, JOHN A. Math; Warfare Dall Wiliam Heal Dall, and Henry A. Minist Dall, Caroline Wells H. Leeturing; Lit. 531 W.W. W.W. W.W. W.W. AMS; W.W. W.W. W.W; A.C. W.W. WW; 532 THE AMERICAN NATURALIST . [Vor. LVI M Ree Charles Benediet— Davenport, William E. Sociol; Ministry W.W. us Davenport, Frances G. History W.W. Drew, Gilman Arthur— Drew, William Lineoln Law W.W. Forbes, eas Alfred— S RBES, ERNEST B. Entomology WW. BS ise RosERT H. Soil Chemistry A.M.S; W.W. Gage, Simeon Henry— Si GAGE, MARY Sanitation W.W. Gerould, John Hir Gerould, gore Hall Philology W.W. B Gerould, James Library - W.W. Glaser, Otto Charles— F GLASER, CHARLES : Chemistry A.M.S. Grave, Caswell— RAVE, BENJAMIN H. Zoology , AMS, Gulick, John Thomas— S GULICK, ADDISON Zoology A.M.S. F Gulick, Peter Johnson Ministry A.C. BS GULICK, LUTHER H. Physiol; Edue A.M.8; W.W BS Gulick, Sidney Lewis Ministry ; .W. Writing BD Jewetts, Frances Gulick Hygiene W.W. FBS Gulick, Charles Burton Philology W.W. Hargitt, Charles Wesley— S HanarTT, GEORGE T. Geology | A.M.S. Herrick, Charles Judson— ; ; B HERRICK, CLARENCE L. Neurology A.M.S, Howard, Leland Ossian— MSiS Stimson, Henry lewis Polities ; W.W. Warfare MSiD Keith, Dora Wheeler Portrait W.W. Painting Jayne, Horaee— B Jayne, Henry La Barre Law W.W. F JAYNE, DAVID Medieine; A.C, Pharmacy Lefevre, George— B Lefevre, Albert Philosophy W.W. B Lefevre, Arthur Edueation W.W. No. 647] Lillie, Frank Rattray— B LILLIE, RALPH STAYNER Lillie, Ralph Stayner— B LILLIE, FRANK R. Loeb, Jacques— B OEB, LEO Mayer, Alfred Goldborough— F MAYER, ALFRED M, FB Mayer, Franeis B, Merriam, Clinton Hart— Si BAILEY, FLORENCE M. FB Merriam, Augustus C. —€— Maynard Mayo— etealf, Irving Wight B METCALF, WiLMOT V. FBS Metcalf, Wilder S. Montgomery, Thomas Harrison— B Montg A. omery, James MB Morton, James St. Clair B MonTON, THOM MF Morron, SAMUEL G. Moore, John Perey— B Mover, HENRY FRANK Newman, Horatio Haekett— Newman, Albert Henry Nutting, Charles Cleveland— MB Hunt, Henry MB Hunt, Lewis Cass Osborn, Henry Fairfield— Osborn, William Church MF Sturges, Jonathan Peckham, George Williams— Peckham, R. Wheeler FBS Peckham, R. Wheeler, Jr. FBS Peckham, Wheeler H Rice, Edward Loranus— F Rice, WILLIAM NORTH FAMILY RESEMBLANCES Zoology Zoology Pathology Physies Art Ornitholo Archeol; ogy Finance; Ministry Theol; History y Ethnol; Path; t. Zoology History; Theol. Warfare Warfare Law; Polities Trade Philol. 533 * A.M.S. *A MS; W.W. *AMS; W.W. AMS; W.W. AMS; W.W. W.W. 534 THE AMERICAN NATURALIST [Vou. LVI gaa si Robert Wilso Shufeldt, a W. Warfare A.C. Stone, Witmer— Stone, Frederiek D. History; A.C. Library True, Frederick William— B TRUE, ALFRED CHARLES Educ; A.M.8; W.W. Agriculture F True, Charles Kitredge Ministry Verrill, Addison Emery— VERRILL, ALPHEUS H. Zoology A.M.S; W.W. Ward, cag Baldw F WARD, Bisnis H. Bot; Microscopy A.M.S; W.W; A.C, FSi Ward, Anna Lydia Ethnol; WW} AC. Lexicog. Weed, Clarence Moores— Weep, Howarp Evarts Entomology A.M.S. Zeleny, Charles— ZELENY, ANTHONY Physics *A.MS; W.W. B ZELENY, JOHN Physies "AMS: W.W. RELATIVES OF THE WIVES OF MEN OF SCIENCE The same general plan used in the listing of the men of science and their distinguished relatives is followed below. Anatomists Donaldson, "um uu Herbert— F , Calve Arehiteeture A.C. MB Me Mun. ld Architecture AC. Anthropologists McGee, Mrs. W. J.— F NEWCOMB, SIMON Astronomy "AMB; W.W: : AC, MB Hassler, Ferdinand A. Med; Literature W.W. Astronomers Doolittle, Mrs. Charles Leander— B Wolle, Fred Music W.W., S DOOLITTLE, ERIC Astronomy *A.M.S; WW. F WOLLE, FRANCIS Botany A.C. No. 647] FAMILY RESEMBLANCES Frost, Mrs. Edwin Brant— Hazard, Marshal Editing Holden, Mrs. Edward Singleton— B CHAUVENET, REGIS Mining Eng. B CHAUVENET, WILLIAM M. Chemistry F CHAUVENET, WILLIAM Mathematics Loud, Mrs. Frank Herbert— j ILEY, WALTER H. Mining Eng. Mitehell, Mrs. Samuel Alfred— F DuMBLE, Epwin T. Geology Pickering, Mrs. Edward Charles— r park rks, History; Educ. MF Silsbee, Nathaniel Polities Pritehett, Mrs. Henry Smith— À FB McAlister, Ju im Warfare FF McAllister, Nath. Hall Law FBS McAllister, Ward _ Jurisprudence ye gos William Hammond— eib, Samuel aw; Hortieulture Botanists Atkinson, Mrs. George Franeis— Kerr, W. C Geology Bessey, Mrs. Charles E.— S Bessey, ERNST A. Botany Clements, Mrs. Frederie Edward— Si Sehwartz, Julia Literature Coulter, Mrs. John Merle— S COULTER, JOHN G. Botany Coulter, Mrs. Stanley— id Roswell iban Ministry FB Post, Truma Ministry fox, Mrs. William Gilson— Si Horsford, Cornelia Areheology F HonsroRp, EBEN N. Chemistry M Horsford, Mary L. H. Poetry FF Horsford, Jedediah . Warfare; Polities Ganong, Mrs. William Francis— B arman, Bliss Poetry; Editing 535 A.M.S; W.W. *A.M.S; W.W. W.W. A.M.S. W.W. 536 THE AMERICAN NATURALIST [Vor. LVI Greenman, Mrs. Jesse More— MSiS Hartranft, John F. Polities A.C. Pinchot, Mrs. Gifford— F Bryce, Lloyd 8. Editing; W.W. Polities MF Cooper, Edward Finanee; Ww. Polities Ramaley, Mrs. Franei JACKSON, Runs Ophthalmology W.W. T 2 Joseph Nelson— : Sims, Charles R. Ministry; Edue. A.C. Stone, Mrs, George Edward— F CLARK, HENRY JAMES Botany - A.C. Wilson, Mrs. William Posell— FF Williams, Charles K. Polities A.C. Chemists Bigelow, Mrs, Samuel Lawrenee— Harrison, Joseph Railroad A.C, Building Burgess, Mrs. Charles Frederick— B Jackson, Charles F. Literature W.W. Cushman, Mrs. Allerton Seward— B Hoppin, Joseph Clark Archeology ; W.W. Art FB Hoppin, Augustus Art; Literature A.C. FB oppin, Thomas Ast: Seulptoring A.C. FB Hoppin, William Jones Stage; Editing A.C. Franklin, Mrs. Edward Curtis— B Seott, Charles Fred. Polities W.W. Gies, age William John— Tressler, David Education AC, es Tressler, Victor Edueation W.W. Kahlenberg, Mrs. Louis— ; B EALD, FRED. DeForest Botany "AMS; WVW. Long, Mrs. John Harper— FB Stoneman, George Warfare AC. Marshall, Mrs. John— F Ww ORMLEY, T. G. A.C. No. 647] FAMILY RESEMBLANCES Mine. Mrs. Charles Edward— F BARKER, GEORGE FRED. Osborne, Mrs. Thomas Burr— JOHNSON, SAMUEL WM, Palmer, Mrs. Chase— F Edwards, Howard FBD Edwards, Louise Pellew, Mrs. Charles Ernest— F CHANDLER, CHARLES F. FB CHANDLER, WILLIAM H. Richards, sare Theodore F Thayer, edi Heny Sanger, Mrs. Charles Robert— F Davis, And FB Davis, Hasbrook FB Davis, Hora FB Davis, John C. B FF Davis, John a 0 FBS Davis, John Saunders, Mrs. Arthur Perey— Es s Brownell, Silas B. Shimer, Mrs, Porter William— B Sandt, George W. Talbot, Mrs. Henry Paul— FSiS Baker, Newton D. Tassin, Mrs. Wirt— F Moran, Thomas M Moran, Mary Nimo FB Moran, Edward Peter . FBS Moran, John Leon FBS Moran, Edward Perey FBD Moran, Annette Van Slyke, Mrs. Lucius L.— 8 Van SLYKE, DONALD D. Physies Chemistry due; i Linguisties Literature Chemistry Chemistry Theology Numismaties are Manufaeturing Diplomaey Polities Law; Diplomaey Law Ministry; Editing Law; Diplomaey Art; Exploration *A.M.8; W.W; A.C. *A.MS; W.W; A.C W.W. W.W "AMS: WW: A.C. AMS; WW.: .C. W.W; A.C W.W; A.C A.C A.C. A.C, ‘AC A.C, W.W W.W. W.W. WW: AC A.C. A.C, Ww AU W.W; AC W.W; AC WwW A.M.S. 538 THE AMERICAN NATURALIST [Vor. LVI Venable, pm Francis Preston— anning, Jam Jurisprudenee W.W. FB NAE iones C. Jurisprudenee A.C. Wiechman, Mrs. Ferdinand G.— B Damrosch, Frank Musie W.W; A.C. B Damrosch, Walter Musie W.W; A.C. F Damroseh, Leopold Musie AJ F Mrs. Harvey Washington— Kelton, John C. Warfare; A.C, Sociol. Geo ogists Brooks, Mrs. Alfred Hulse— F BAKER, FRANK Anatomy *A.M.B; WW. Chamberlin, Mrs. T. C S CHAMBERLIN, R. T. Geology A.M.8; W.W. Cross, Mrs. Whitman— FBS Places Naas L Exploration; A.C. Eng. Eastman, 2» Charles— F ‘LARK, ALVAN G. Astronomy A.C. FF CLARK, ALVAN Astronomy A.C. Gannet, Mrs. Henry— FSiS Gerdon, Seth C. Surgery W.W. MSiS Howe, Lucien Ophthalmology W.W. Grant, Mrs. Ulysses Sherman— B WINCHELL, ALEX. N. Geology * A.M.8; WW. B WINCHELL, Horace V. Geology A.M.8; W.W. F WiwcHELL, NEWTON H. Geology *A.M.8; W.W; A.C. FB WINCHELL, ALEXANDER Geology AC, FB Winchell, Samuel R. Edue; Lit. W.W. .Harris, Mrs. Gilbert Dennison— B Stoneman, George Warfare A.C. Hitchcock, Mrs. Charles Henry— F Ba arrows, E. P. Linguisties A.C. Hobbs, Mrs. William Herbert— S Kimball, Alonzo Art W.W. FSiS Banister, HENRY Geol; Med. A.M.S. Mathews, Mrs. Edward Bennett— B WHITMAN, FRANK P. Physies *A.M.S; WW. No. 647] FAMILY RESEMBLANCES 539 Pirsson, Mrs. Louis Valentine— F B RUSH, GEORGE JARVIS Mineralogy *A.M.8; W.W; A.C. oe Mrs. Raphael— Smyth, Margarita P. Art W.W. vt Hill Edward B. Musie W.W. MBS Pope, Alexander Painting W.W. Rice, Mrs. William North— S Kick, EDWARD L. Zoology *A.M.S; W.W. Smith, Mrs. T Allen— F Gar and, Landon C. Education A.C. FB Bia, Hugh A. Law; Warfare A.C. FBS Garland, Hugh A Law; Warfare A.C. FBS Garland, Baud Law; Warfare AC. Stevenson, Mrs. John James— Ewing, Nathaniel Jurisprudence W.W, Vaughn, Mrs. Thomas Wayland— Upham, Charles W. Lit; Theol; A.C. lit. Weed, Mrs. Walter Harvey— F Hill, Ebenezer J. Polities W.W. Willis, Mrs. Bailey— BAKER, FRANK Anatomy *A.M.8; W.W. Winchell, Mrs. N. H.— S WINCHELL, ALEXANDER Geology *A.M.S; W.W. VH WINCHELL, Horace V. Geology A.M.S; W.W. Mathematicians Beman, Mrs. Wooster Woodruff— B Burton, Ernest DeWitt Theology W.W. B Burton, Henry Linguistics; W.W. Educ. Eehols, Mrs. William Holdin MB Tucker, H. St. George w C MB Tucker, Nathaniel B w; Literature A.C. MF Tueker, St. George Lésulsties FF inguisties A.C MBS Tucker, Joh Polities A.C. MBS Tucker, T Md B. Journal; Lit. A.C. Lehmer, Mrs. Derrick Norman— B Mitehell, Wesley C. Economics W.W. 540 THE AMERICAN NATURALIST [Vor. LVI MeMahon, Mrs. Jam . B Crane, aa F. Linguistics Moore, Mrs. Eliakim Hastings— iB YouNG, JOHN WESLEY Mathematics Peirce, Mrs. Charles Santiago S.— Si ay, Amy Musie MB Hopkins, Casper Journalism MB Hopkins, Charles J. Musie MB Hopkins, Edward A. Finanee MB Hopkins, FRED V. Medieine MB Hopkins, John H. Ministry ; Poetry MF Hopkins, John H. Ministry Roe, Mrs. Edward Drake, Jr.— : B Robinson, Maurice H. Polit. Economy Veblen, Mrs. Oswald— B RICHARDSON, OWEN W. Physies Pathologists Barker, Mrs. Lewellys— B LSEY, JOHN TAYLOR Medicine Cabot, Mrs. Richard Clarke— MB Lowell, John Jurisprudence MBS Lowell Abbot Lawrenee Educ; Hist; MBS Lowell, Franeis C. Jurisprudence MBS LOWELL, PERCIVAL Astronomy MBD Lowell, Amy Poetry Dana, Mrs. Charles Loomis— MB Opdyke, William S. Law MF Opdyke, George Finanee; Polities Edsall, Mrs. David Linn— MB Foote, Arthur Musie Ernst, Mrs. Harold Clarenee— B Frothi rothingham, Louis A. Law B Frothingham, Paul R. Mini stry FSi ieu Ellen F. Linguisties; Lit. FF Frothi bin. Nathaniel ^ Ministry MF Lunt, William P. Ministry FB Frothingham, O. B. Ministry Fitz, Mrs, Reginald Heber— F KE, Epwarp H. Medieine No. 647] Flexner, een Sim B UAE Gee M. Si heey Carey Jordan, Mrs. Edward Oakes— MBS r MSiS LOMBARD, Wee, Meyer, Mrs. Adolf— S Paige, Hildegard B. 1 B Brooks, wenns HULSE B. F Brooks, Mitchell, Mrs. S. Wier— S MITCHELL, JOHN K. S Mitchell, Langdon E. Pearce, Mrs. Richard Mills— F Mu SSER, JOHN HERR Putnam, Mrs, James Jackson— MB Channing, William H. Ravenel, Mrs, Mazych Porcher— Allston, R. F. W. g ‘John Collins— , Francis G. FF ^n Robert G. Warren, BS Shaw, Robert G. FSiS Oliver, Robert S. FAMILY RESEMBLANCES Neurology due; Literature Physiology Literature Geology Geology Neurology Playwright Pathology Ministry Polities Finance; Sociol. Fi ; are Warfare Polities Physicists Abbe, Mrs. Cleveland, Sr.— 8 BE, CLEVELAND, JR. S. ABBE, TRUMAN Bedell, Mrs. Fred— B sHING, HENRY PLATT MSiS Cushing, William E, Briggs, Mrs. Lyman James— F Cook, ALBERT JOHN MB BALDWIN, JAMES Meteorology Surg; Physiol. Elec. Eng. Zoology Surgery 541 A.M.8; W.W. W.W TAMB; WW, W.W. *A.M.S; W.W. A.C A.M.8; W.W. W.W. *A.M.8; W.W. 542 Crawford, Mrs. Morris Barker— Rice, WILLIAM NORTH BS RICE, EDWARD LORANUS Day, Mrs. William Seofield— MF Hoe, Richard Dorsey, a dm Ernst— B , EpwaRD H, B is WILLIAM Ferry, Mrs, Ervin Sidney— White, Andrew D. Gale, Mrs. Henry Gordon— F Cook, John W. MSiS Hors, Richard Goodwin, Mrs. Harry Manley— MB F G ARLOW, WM. GILSON Ives, Mrs. Frederick Eugene— E Ives, HERBERT EUGENE Jackson, Mrs. Dugald Caleb— MB JORDAN, WHITMAN H. Kimball, Mrs. Arthur Lalanne— FB Scribner, Charles MB liie, George W. MBS Sehuyler, Eugene King, Mrs. Arthur Scott— FSiS DaiNS, FRANK B. Magie, Mrs. William Franeis— FB Hodge, A. Alexander FF Hodge, Charles Mendenhall, Mrs. T. D.— 8 MENDENHALL, C. E. More, Mrs. Louis Trenchard— Si Taft, Helen Herron SiD Taft, Helen Herron Riehardson, Mrs. Owen William— B WiLSON, HAROLD A. THE AMERICAN NATURALIST Geology Zoology Inventing Meteorology Geology Edue; Diplom, Edueation Poetry Botany Physies Agriculture Publishing Finance; Lit. Diplomacy; Lit. Chemistry Ministry Theology Physics Society ` Education Physies Polit; "AALS: WW; AC *A.M.S; WW. sC W.W. A.M.S; W.W. W.W; A.C. *A.M.S; W.W. W.W. W.W. W.W. [Vor. LVI No. 647] Benj ae Pu ames D. Snow, Mrs. Stevens, Mrs. Walter LeConte— F Letcher, John TE Mrs. Charles Burton— WHITE, HENRY SEELEY Trowbridge, Mrs. Augustus— F Fulton, J. D. FAMILY RESEMBLANCES Literature Polities Mathematies Ministry; Lit. Wead, Mrs. Charles Kasson— F Pease, Calvin Philos; Edue. Webster, Mrs. Arthur Gordon— F — Robert Warfare FB Town Warfare FB Das Borir Pathology Wright, Mrs. Arthur William— : F SILLIMAN, BENJAMIN Chemistry FF SILLIMAN, BENJAMIN em; Geology MSiS Dana, EDWARD Zeleny, Mrs. John— MB ONES, FREDERICK S. Physies Physiologists T Mrs. John— MELTZER, SAMUEL J. Physiology Dawson, Mrs. Perey Millard— B Learned, Henry Barrett Lit; History Henderson, Mrs, Lawrence Joseph— Thayer, Joseph Henry Bible Henderson, Mrs. Yandell— FB Colby, Gardner Philanth ; Railway FBS Colby, Charles L. Ry. Management FBS Colby, Everett Law; Polities FBS Colby, Henry F. Ministry Hough, Mrs Theodore— FBS WmnrrEHEAD, RicHARD H. Pathology Hunt, Mrs. Ried— F Taylor, Hannis Law; Lit; Diplom. 543 W.W; A.C. A.C. "AMS; WW. A.C A.C A.C A.C A.C A.C A.C. "AMS: WW *A.M.S; W.W "AMS: W.W. W.W. W.W; A.C A.C A.C, W.W. W.W; A.C. A.M.S; W.W W.W; AC. 544 THE AMERICAN NATURALIST [Vot. LVI Lusk, Mrs. — Inventing; Art W.W; A.C. FF =e dete i Finanee; Mfg. W.W; A.C. Psychologists Angel, Mrs. Frank— F ayard, Thomas F. Diplomacy A.C, FF Bayard, James A. Diplomacy A.C. Buehner, Mrs. Edward Franklin— MB Morgan, Daniel Nash Finanee . W.W. Jastrow, Mrs. Joseph— Szo : Henrietta Language; W.W. Editing F Szold, Benjamin Ministry; Lit. W.W. Strong, nes Charles Augustus— B ockefeller, J. D., Jr. Finance W.W. Si krate es -Edith R. Wealth W.W. F Rockefeller, John D. Finanee W.W; A.C FB Rockefeller, William Finanee W.W; A.C FBS Rockefeller, W. G. Finanee W.W. Zoologists Andrews, Mrs. Ethan Allen— MF Powell, John Hare Diplomaey A.C. Calkins, igo Gary Nathan— B WILLISTON, ARTHUR Eng; Edue. A.M.8; W.W. B Williston, Samuel W.W. Child, Mrs. Charles Manning— T y n Meter, John B. Philol; Educ. W.W. Forbes, Mrs. Stephen A. S ORBES, ERNEST B. Entomology W.W. Gulick, Mrs. John Thom 8 GULICK, ju ia ADDISON Zoology A.M.S. m Mrs. Charles Wesley— ; ITT, GEORGE T. Zoology A.M.S. Jayne, Mrs. Horace— B Furness, Horace H. Literatu WW. B ESS, WILLIAM H. Anthropology A.M.8; W.W. F Furness, Horace H. Literature W.W; A.C. M Furness, r K. Literature A.C. FB Furness, W. H. Art A.C. FSi Wister, ad Ax. D Linguisties A.C. FF Furness, W. H. Anthropology ALC. No. 641] Lillie, Mrs. Frank Rattray— B rane, Charles R. F Crane, Richard Teller Linton, Mrs. Edward— B Brownson, M. Acheson Si Brownson . Wilson MB Acheson, Savers W. MBS Acheson, Alex.. W. MBS AcHESON, EDWARD G. MBS Acheson, Ernest F, MBS Acheson, Marcus W. Mast, Mrs. Samuel Ottmar— B TENNENT, DAvID HILT Mayer, Mrs. Alfred— ' Si Hyatt, Anna V. F HYATT, puse Newman, Mrs. Horatio Hackett— LILLIE, FRANK RATTRAY MSiS LILLIE, RALPH Osborn, Mrs. Henry Fairfield— F Perry, Alexander J. Peckham, Mrs. George William— B GIFFORD, HAROLD FB Gifford, Sanford R. Pratt, Mrs. Henry Sherring— MBS Woodbury, Charles H. Reighard, Mrs. Jacob— FBS "enne. LIVINGSTON FBS Farrand, Max FBS acini Wilson Sedgwiek, Mrs. William Thompson— B Rice, Riehard Austin oe Mrs. Robert LX fina — AUDUBON, JONATHAN J. Smallwood, Mrs. William Martin— MB Corr, JuDSON BOARDMAN Stiles, Mrs. Charles Wardell— Forpycr, JOHN A. MB Fordyce, Samuel W. FAMILY RESEMBLANCES Mfg; Diplomacy Mfg; Education Ministry ath; History Jurisprudence Med; Polities Chemistry Editing; Polities Jurisprudence Biology Sculptoring Geology Zoology Physiology Warfare Ophthalmology rt Art Anthropology Editing; Edue. Languages ; Hist. Ornithology Astronomy Dermatology Finance 545 *A.M.S; W.W. W.W. AC. *A.M.S; W.W. *A.M.S. W.W; A.C. W.W. A.C. W.W. TAMS: W.W: W.W. W.W. W.W. W.W. 546 THE AMERICAN NATURALIST [Vor. LVI Strong, Mrs. Reuben Myron— F Fr eeman, Henry V. Jurisprudenee W.W. MB Curtis, Pawara Lewis Literature W.W. MF Curtis, William 8. ` Education A.C. True, Mrs. Frederiek William— F PRENTISS, DANIEL W. Medicine W.W. Verrill, Mrs, Addison Emory— B MITH, SIDNEY Zoology "AMS: W.W. S VERRILL, ALPHEUS H. Zoology A.M.8; WW.: Wilson, Mrs, Henry Van Peters— MB STEDMAN, HENRY Rust Pathology A.M.S; W.W. Zeleny, Mrs. Charle MB ERIKSON, (oS ANTON Physies A.M.S. DISTINGUISHED WIVES OF AMERICAN MEN OF SCIENCE Anatomists Lewis, Warren Harmon— EWIS, MARGARET REED Biology A.M.S. i Anthropologists MeGee, W. J.— MoGxE, ANITA NEWCOMB Medicine | A.M.8; W.W. Botanists Wilson, William Powell— son, Lucy M. WILLIAMS Biology A.M.S; W.W. Chemists Norton, Thomas Herbert— Norton, Edith Eliza Ames >- Literature WW, Phelps, me King— . PHELPS, MARTHA AUSTIN Chemistry A.M.S. Geologists Grabau, Amadeus William— Grabau, Mary Antin Literature W.W. Mathematicians Franklin, Fabian— FRANKLIN, CHRISTINE LADD Psychology *A.M.8; W.W. eden weaned Apps Sanders— ee, Harriet M. Fay Literature W.W. No. 647] FAMILY RESEMBLANCES 547 Pathologists Cabot, Richard C.— abot, Ella Mont Lyman Philos; W.W. Education Physicists Barus, Carl— Barus, Annie G. Howes Literature W.W. Zoologists Andrews, Ethan Allen— ANDREWS, SARA G. FOULKE Zoology A.M.S. Calvert, Philip Powell— CALVER MELIA K. SMITH ' Zoology A.M.S. Comstock, John Henry— ; COMSTOCK, ANNA BOTSFORD Entomology ; A.M.8; W.W. Art Davenport, Charles Benedict— DAVENPORT, GERTRUDE CROTTY Zoology A.M.S. Gage, par Hen : GaG QE PHELPS Zoology “AMS: W.W. xn Horace— Jayne, Caroline G. Furness Literature W.W. Moore, John Perey— MOORE, KATHLEEN CARTER Psychology A.M.S. Nickerson, Winfield Scott— NICKERSON, MARGARET LEWIS Histology; * A.M.S. Embryol. Peckham, George William— PECKHAM, ELiz. M. GIFFORD Zoology A.M.S; W.W. Wilder, Harris Hawthorne— WILDER, INEZ WHIPPLE Zoology A.M.S; W.W. (To be continued) AUTOPHORIC TRANSPLANTATION, ITS THEORY AND PRACTISE PROFESSOR HANS PRZIBRAM BIOLOGISCHE VERSUCHSANSTALT DER AKADEMIE DER WISSENSCHAFTEN, VIENNA Ir a machine breaks down, the mechanical engineer has four ways of repairing it. He may discard the broken parts and reconstruct the whole on a smaller scale; he may fabricate the missing part and fit it into its right place again; he could also take a piece from another machine of more or less similar type, as long as an ex- change is made possible by the material of the parts, soldering the broken pieces together or fixing them by screws, wires, etc.; or lastly he may simply exchange the broken part for a whole one, first taking the former out of the broken machine at the points where it was joined, and refitting the new part, taken in like manner out of a similar machine by the same means, in the place of the first. An organism is often just as badly in want of repair as a machine of human fabric. In comparing the two I do not wish to enter here into the controversy of Mechanism versus Vitalism. No vitalist will deny that the body of an animal, let us say of vertebrate or arthropod type, is built up of various contrivances the physicist calls machines, and that its functions are best described in physical and chemical terms. It is not the machinery of organized forms that he would throw doubt on, but the mechanical or chemical nature of its driver. Now, when living machinery is broken or maimed, there are the same four possibilities of repair stated above. The organism may shed such parts as are now super- fluous for its reduced size and reconstruct itself on the basis of a proportionately diminished form, as in small pieces of planarians, a process called *' Morphallaxis,’’ by T. H. Morgan. Secondly, a missing part may be 548 No. 641] AUTOPHORIC TRANSPLANTATION 549 manufactured anew by the remaining body of the ani- mal, such ‘‘ Regeneration ’’ not being uncommon even in whole extremities of amphibians and crayfish. But, un- fortunately, in the warm-blooded vertebrates this faculty is very limited, not extending much beyond the repair of small pieces of tissue, and never including a whole organ or appendage. It has therefore long been customary in human medicine to try to replace lost parts by ‘‘ trans- plantation ’’ of cornea, skin, muscle, bone, or even nerve and blood vessel. Without regard to the composition of the injured part, small pieces or larger portions have been taken from the same or from another individual, and again without special orientation have been grafted upon the wound. All sorts of fastenings have been tried, bandages, plaster, wires, ligatures, but mostly with poor results. The same methods and many others have been applied in experimental zoology, but only when embry- onic stages which had not functioned before the operation were used have good results been achieved. Neverthe- less it has been demonstrated by A. Carrel that even whole limbs and kidneys may be again healed back in mammals and in the ease of the latter again become func- tionally active. But the tedious method of sewing every sinew, blood vessel and nerve together seems to have pre- vented till now the general application of this discovery. Carrel’s method, as also that of other surgeons, must be compared to the third method of the engineer, when he is soldering or fixing a broken piece on to another, trying to repair the machine without taking it to pieces. Now it is generally simpler to take out the injured piece of a machine, by unscrewing or unsoldering or even by striking it out of the whole by sheer force, so that its connections give way at the points of least resistance, and to replace it by a new one of exactly the same form, than to try and fix the broken parts together again at the point of breakage. Is there a possibility of applying this fourth method of the engineer to the organism? One will, perhaps, at first be inclined to doubt this proposi- 550 THE AMERICAN NATURALIST [Vor. LVI tion. The vitalist will now come forward and claim that the organism is not constituted by parts simply fastened together at certain points, that its unity is the cause of its function; the mechanist will be inclined to doubt the possibility of whole organs regaining their function by ‘‘ exchange " in animals without high regenerating power, for he has been trained to believe in the destruc- tion of function by the severing of the nerve. Let us turn to facts. Certain animals, widely distrib- uted through the animal kingdom, practise the faculty of shedding appendages or other parts of their body at cer- tain preformed breaking points. This ‘‘autotomy’’ is also observed in the Crinoid, Antedon rosaceus. Work- ing at the Naples Station in 1900 on the regeneration of ‘these Crinoids I wanted to find out if the color in regen- erating arms would be influenced by the color of the vis- ceral mass. Now Antedon shows a great variety of very distinct shades, such as bright yellow, carmine red and chocolate brown. The visceral mass, easily shed by the animal, was transplanted in proper orientation to a speci- men of different color, also void of its viscera. It was immediately accepted by the new owner and clutched tightly to the calyx, as is the usual thing with the normal animal. The connections between the new visceral sac and the body were soon restored, the exchange succeeding in every ease. Mouth and anus, both situated on the sur- face of the visceral sae, became functional again. It is clear that here there is a case of the fourth method of the. engineer, namely the replacement of a missing part by a new one of exactly the same form fixed in at the same connecting points as before. One difference is apparent: in the machine there will be little if any activity on the part of the receiver or the new part, whilst in the Crinoid the newly fixed parts are reunited by internal forces. If we want to understand the ‘‘ exchange "' followed by func- tion, it is therefore necessary to know the nature of these forces. Is it possible to account for them on the ground of our present knowledge of living matter? Can we con- No. 641] AUTOPHORIC TRANSPLANTATION 551 ceive the organism as an engineer mending his own body? When the visceral mass of Antedon is not replaced, a new sac is regenerated by the creature. As in all cases of regeneration known to me, it is nothing else than an ac- celeration of growth going on normally at slower rate, but in the same direction and sense. From this theoreti- eal standpoint, which has been proved to be correct over- and over again, we can be satisfied that there are growing forces in the Antedon sufficient to ensure the attachment of the new visceral sac. We have heard that in higher animals regeneration is not as ready to supply lost parts, and as soon as growth ceases, for instance in the imago of insects, the faculty of restoring missing limbs is lost. But a certain degree of repair has been noticed and experimentally tested even here, for instance the closing of holes pricked in the in- tegument of beetles, and even the resprouting of torn- out wings as mere skin duplicatures. In vertebrates a good deal of physiological regeneration is always going on in the tissues, and transplanted pieces of living tissue often become attached in a short time by connective tissue and blood vessels growing over and into them. Will ex- change of organs lead under certain conditions to their functional restoration also in such animals as these? The first condition must be the possibility of removing the part to be replaced always in the same place and man- ner, so as to be sure that it will comprise just the same material and fit in again in the corresponding place of the new host. Planes of preformed breakage would an- swer best to this condition, but they are generally pre- cluded by the second condition that must be fulfilled, namely retention of the implanted organ by the own forces of the recipient. Such forces may be divided into three groups: first, the natural friction of a mass pressed into a socket, also aided by atmospheric pressure; secondly, the active aid of muscle and nerve clutching the im- planted organ and preventing it from falling out of its place; thirdly, the clotting of the body fluids, gluing, as 552 THE AMERICAN NATURALIST [Vor. LVI it were, the graft to the stock. During the last two years my pupils and myself have tried to extend this method, which I now call ** autophorie ” or self-retaining trans- plantation, to other cases than the visceral sac of Ante- don, and we have found that under these conditions func- tion can be restored in a degree unknown till now, at least in developed animals. The eye of vertebrates may be described as a ball- shaped camera movable by three pairs of levers in all directions of space, connected with its supply of chemi- cals by the blood vessels and in communication with its operator, the brain, by the optic nerve. If these fixing strings are severed, there is scarcely any attachment to the surroundings save some connecting tissue of unspe- cialized sort. The ‘‘ camera "' itself will not be injured, if the whole eyeball be taken out of the orbit, and there is scarcely a possibility of altering the points of sever- ance if the enucleation be made quickly and with decision. If the eye is restored to its orbit, it will therefore be possible for all the above-mentioned connections to join again. This was observed as long ago as 1906 by Rug- gero Pardo in Triton, who made experiments on the neces- sity of the presence of the optic nerve for the regenerative process in the eye of this amphibian. Unintentionally he had excised the eyeball with the nerve and was much astonished at its reattachment to the orbit. But will eye- sight be restored with this reattachment? Pardo was not able to convince himself of this fact, although on histo- logical examination he found the optic nerve regenerated. I have suspected for some time that the vertebrate eye might furnish good material for the restoration of fune- tion by autophorie transplantation, as it will in many forms be retained in the orbit by friction and atmos- pherie pressure alone, aided also in some cases by the eyelids elosing over the eyeball, and by its great surface securing wide contact with the blood issuing into the orbit after extirpation. My own first experiments to realize this expectation in new-born rats failed. No. 647] AUTOPHORIC TRANSPLANTATION 553 In the new-born rat, as in many mammals, the eyes are tightly closed and the lids connected by tissue. This seemed to afford favorable conditions for the exchange of eyes, as they would be kept in place by the tight closure of the eyelids. Having severed the lids, I interchanged the eyes and, as expected, the eyelids shut again tightly and kept the eyeballs in place. But when the eyelids opened again at the normal time, the eyes had grown on, although they were not functional, and totally disap- peared in time. Disappointed at this failure, the experi- ments were discontinued. It is now pretty certain that this poor result was due to the unfavorable conditions obtaining in very young mammals, for we are now able to demonstrate the correctness of my original supposi- tion. Theodor Koppányi, a young Hungarian student, working under my direction in the ‘‘ Biologische Ver- suchsanstalt ’? in Vienna, has succeeded in making the autophorie transplantation of the eye in a variety of species, extending from fish to mammal. The work of Pardo on Triton was confirmed, and older rats yielded excellent results. It seems that in the young stages of rats there were difficulties in the way of the eye obtaining a sufficient supply of blood, since also in Koppanyi’s ex- periments it was far easier to get the eyes to become reattached and functional in older specimens. Indeed, it is probable that the pressure of the eyelids exerted on the replaced eyeball in the new-born rats is a hindrance. Grown rats do not close the eyelids tightly upon the eyeballs, so that it is even advisable to pin the lids or sew them together for a day or two, lest the ani- mal whisk out the implanted eyes or seratch at them be- fore they are attached sufficiently firmly to withstand such treatment. We have been able to show that these replanted eyes are funetional, all possible tests yielding positive results and being in striking contrast to those in blinded ani- mals. Microscopical examination of sections through 1For details of these experiments I must refer to our previous short . 554 THE AMERICAN NATURALIST [Vor. LVI replanted eyes, which had again regained their function, has been made by Professor Walter Kolmer, of the Physi- ologieal Institute, University of Vienna, and the re-in- growth of the severed optic nerve-fibers into the optic thalamus is beyond doubt. Professor Kolmer, as all other authorities, to whom the animals with functioning replanted eyes were shown, stated that they would scarcely have believed the fact, without having them- selves seen and tested it. Some oculists even refused to believe what they saw, taking refuge in far-fetched ex- planations for the absolutely normal behavior of the rats and for the connection of retina and brain in anatomical and microscopical preparations. But is the restoration of function in the vertebrate eye really in contradiction to the facts known to us concerning the regeneration in this animal type? If we resort to our theory of regen- eration as accelerated growth, moving on the same lines as normal differentiation, and waning with higher spe- cialization, it is necessary to inquire into the normal de- velopment of the eye and optic nerve, before answering this question. The vertebrate eye grows from multiple origins, the nervous elements being derived from a fold of the central nervous system (brain). It is generally believed that the nerves of the brain grow in centrifugal direction and are incapable of regeneration, as one does not observe regeneration-cones at the peripheral end of sectioned central nerves as a rule. Ramón y Cajal, on the other hand, thinks that this inability to regenerate is only a consequence of secondary difficulties, regeneration at least commencing when the right nurture is given: this may be accomplished by inserting degenerating nerve-pieces into the pathway of the sectioned nerve. At any rate there would be but little chance of quick and suf- ficient regeneration, if the eye depended on the nerve growing into it from the brain. Fortunately, as is well known, the fibers of the optie nerve in ontogeny grow communications in the Akademischer Anzeiger, Wien; they will be followed by publieation in extenso in the Archiv für Entwicklungsmechanik, 1922. No. 647] AUTOPHORIC TRANSPLANTATION 555 centripetally from the retina towards the thalamus op- ticus. In regeneration this same process need only be repeated. Edward Uhlenhuth, while working at our ** Biologische Versuchsanstalt,” proved in 1912 that the optic nerve of salamander eyes implanted on the back of the same species grows centripetally towards the spinal cord and even in several instances united with the next spinal ganglion. These transplanted eyes were of course devoid of function, as the nerve had not reached its proper center, but it was of greatest interest to note that the eye, although severed and removed from its natural connection, had totally regenerated after a short period of partial degeneration. Bearing these two points in view, the centripetal growth in ontogeny and the same process in transplanted eyes, we see our theoretical de- mands for the reattachment of replanted eyes fulfilled: the nerve fibers will grow backwards through the orbit, continuing on their usual path and probably finding good conditions there in the degenerating central stump. The usual assumption that function of a sensitive organ can not be restored after severing the nerve is based on false presumptions, especially the idea that the proper central nerve center is responsible for regeneration. We have in several instances proved that it is not necessary for a body part to be connected with its normal nervous center for regeneration to set in and proceed till completion. I may call attention to Oskar Kurz’s transplantations of knees taken from developed tritons and placed on the side of the same animal. Out of this bit of leg all distal parts were regenerated, tibia, fibula, foot and toes, al- though connection of the nerve-stump remaining in the graft with the normal nervous center in the lumbar region can not have taken place. It is quite another question, how far the presence of nerve is necessary for restoration of normal form; a question often confounded with the inability of reestablishing function after severing of nerves. I will not enter into these problems here, as they are being investigated by several of my fellow-workers 556 THE AMERICAN NATURALIST [Vor. LVI and definite statements can not yet be made. The foun- dation for the statement that eyes severed from their connection with the brain are not able to regain sight seems to lie in the fact that the optic nerve in mammals, when the eyes are left movable by their proper muscles, can not find its way to a connection with any nerve center, and then degenerates with the other parts of the eye. It seems that the regenerating ends of the optic nerve fibers coming from the retina are carried to and fro by each rolling of the eye and thus fail to connect with the central stump of the nerve. In contrast to this sheering of the fibers in eyes left attached to the orbit after severing of the nerve, the nerve fibers in autophorie replantation reach their goal before the muscles have grown together and become movable again. It must be emphasized that our method involves no injury to the nerve besides a clean cut, and also that Boeke in Amsterdam has been able to obtain results in nerve regeneration far exceeding those of previous experimenters by avoiding suturing or other- wise ill-treating the nerves. A second opportunity for autophoric replantation is afforded in the vertebrate eye by the lens. It is well known that this part of the eye is derived ontogenetically from an invagination pinching off from the outer layer of ectoderm. The lens of cold-blooded vertebrates, espe- cially urodeles, is capable of regeneration and is easily extracted as a whole, and when it is replanted again into its former place, it fits well into the lens-sac. At my sug- gestion Berthold Wiesner has applied the method of autophorie replantation to the lens of fish and amphibia; the results show that replanted lenses can clear up again and restore normal eyesight to their bearer. In mam- mals analogous experiments have not yet succeeded, perhaps because in the rat, the only available mammals for the present, conditions are unfavorable in respect to the relative size of lens, cornea and eyeball. In other forms, as in man, where the lens relative to the size of the eye is much smaller, replantation should succeed, as No. 647] AUTOPHORIC TRANSPLANTATION 557 the retraction often practised by the oculist is easy, and even regeneration of the lens has been occasionally re- corded (see Literature, Przibram, Regeneration, 1909). Unlike the eye of vertebrates, arthropod eyes are not suitable for our method of transplantation. They usu- ally protrude much too far from their socket to be kept in place after their replantation solely by the friction or other forces exerted by the host. A discovery of Walter Finkler has nevertheless put us in position to avail our- selves of the autophorie method for furnishing insects with a new pair of eyes. This young student, having had the opportunity of seeing the results in vertebrates, sev- ered the head of several types of hexapodes from the thorax and, replanting it on its own body or on that of another decapitated individual, observed its retention by the friction and blood clot. There can be no doubt that also in these cases function is restored, all reactions of the normal animal reappearing after,a few days or weeks, and the tissues joining quickly. Finkler has worked on the larval, pupal and imaginal state. Perhaps the most astonishing fact is the ready response of the imago to such operations in spite of its lack of regenerative power. But also in this case, as in the higher vertebrates, we shall have to take into account that in our experiments no other processes of reparation are called into play than those of slow physiological regeneration, which still persist in adult organisms. At any rate, in all the tissues of adult insects severed connections are quickly restored, when the organs are left in place, as Finkler could prove. His experiments on autophorie transplantation in insects will be.extended to appendages, whilst P. Weiss, Koppányi, Finkler and Wiesner are also occupied with autophorie replantation in parts of the vertebrate body other than the eyes. SUMMARY 1. Well-defined parts of the animal body that may be easily detached at the same connecting points can be re- placed by similar new organs under following conditions: 558 THE AMERICAN NATURALIST [VoueLVi . (a) equal size and orientation; (b) simple exchange with- out exertion of pressure or additional injury to the nerve beyond a clean eut; (c) prevention of loss by the natural means of the animal itself (friction, clasp, blood clot). 2. By this method of ‘‘ autophoric’’ or *' self-retain- ing "' transplantation, the graft taken from an adult in- dividual and replanted into another may be restored to function, even the nerves of the head reuniting, and the bearer being repaired in every respect. 3. These achievements are in accord with the theory stating regeneration to be nothing else than the accelera- tion of physiological processes going on all the time in the body of organisms, for it can be demonstrated that the reattachment proceeds in the same sense as the first growth of the nerve. They contradict, however, the gen- eral assumption that the maintenance and functional re- generation of organs are dependent on their uninter- rupted connection with their special nervous center. 4. Till now we have been able to obtain autophorie ‘replantation with restoration of function in the visceral sae of Antedon (Echinoderms— Preibram, 1901), in the eyes of fish, amphibia and mammals (Vertebrates— K op- pányi, 1921), in the lens of the two former classes ( Wies- ner, 1921), in the heads of insects, walking sticks, water bugs, water beetles (Insects—Finkler, 1921) and in other cases not yet ready for publication. 5. Experimenís with larval stages of amphibia and in- sects as compared with the imaginal state of the same species show that there is no radical difference as to the restoration of function after excision and replantation of a part, in mammals (rats) grown-up specimens even seeming to be more favorable for autophoric replanta- tion. LIST OF REFERENCES Boeke. Studien zur Nervenregeneration, I. Verhandl. Kon. Akad. van etensch. te Amsterdam, 2. Sekt. Deel XVIII, 1916; II. XIX, 1917. Cajal, Ramón y. Studies sobre la Degeneraeión y Regeneracién del sistema nervioso. Madrid, Hijos de Nicolás. moya, I, 1918, II, 1914. Finkler, W. Kopftransplantation bei Insecten, I. Funktionsfühigkeit No. 641] AUTOPHORIC TRANSPLANTATION 559 replantierter Köpfe. Akademischer Anzeiger, Wien, No. 18, 1921. II. Austausch von ee ee zwischen Männchen und Weibchen, Ak, Anz. , 192 uss des replantierten Kopfes auf das Farbkleid pends Sr aad kademischer : Anzeiger, Wien, 1922. Jelinck, A. Die —X ani von Augen VII. Drwecrvitsnehe an Ratten. prom Anzeiger, Wien, 1922. pii. W. Die Divides von Augen V. ub aeger cur n transplantierten Augen. Akademischer Anz , Wie Eo. Th. Die Re plats tiis von Augen, II, Makjariett dui Tub tionsprüfung bei verschiedenen Wirbelttierklassen. Akademischer An- zeiger, Wien, Nos. 7-8, 1921. III. Die ae ee rot id pen jg Süugeraugen. Akadem licher Anzeiger, Wien IV. Ueber das Waehstum der replantierten Jupo. y poy Wie No. 18, 1921. VI. Wechsel des Augen- und Kórperfarbe bei Anamniern, Akad. Anz., Wien, 1922. "VIII. Hetero- und Dysplastik. Akademischer An- zeiger, iia uem Kurz, O. Uebe Regeneration ganzer Extremitäten aus transplantierten Ex Elo en vollentwiekelter Tiere. Zentralblatt für Physiol- ogie, gp No. 12, 1908. Pardo, R. Enucleazione ed innesto del bulbo oculare nei tritoni. Rendi- nli Accademia Lincei (5), XV, 2. Sem., 744, 1906. pomis H. Experimentelle singe über Regeneration. Archiv fiir Ent- icklungsmechanik, XI, , 1901. Experimentalzoologie 2. Regen- Sen Leipzig & Wien, F. nitida, 1909. Methodik der Experi- mentalzoologie. Abderhaldens Handbuch der biologischen Arbeits- methoden (S. 41), 1921. Die Replantation von Augen, Die autophore Methode. Akademischer Anzeiger, Wien, Nr. T n 8, 1921. pna m Die Transplantation des Amphibienauges, I. hiv für ungsmechanik, XXXIII, 732, 1912. II. Archiv de e wick- lun Mb 839 XXXVI, 211, 1913. Wiesner, B. Replantation der Linse, I. Fische d Amphibien. Akad, Anzeiger, Wien, 1921 SPONTANEOUS METAMORPHOSIS OF THE AMERICAN AXOLOTL PROFESSOR W. W. SWINGLE OSBORN ZOOLOGICAL LABORATORY, YALE UNIVERSITY Tar following experiments on axolotl neoteny and meta- morphosis are published, not because of the conclusive nature of the results obtained, but the reverse—because of their inconclusiveness. A record of the work seems warranted in order that other investigators of this prob- lem may be spared considerable expense, time and effort due to unsuitability of the material for experimentation. The latter part of April, 1922, one hundred and nine axolotl larvæ of Amblystoma tigrinum were received from Albuquerque, New Mexico. These animals were obtained through the courtesy of Mr. J. N. Gladding. They varied in length from four inches to fourteen inches, though the average total length was about seven inches. One animal measured fourteen inches from snout to tail tip, another measured eleven inches. They were the largest individuals of the lot. The animals were in excellent condition on arrival and none showed any indications of metamorphosis. EXPERIMENT 1. Avrtopuastic THYROID TRANSPLANTATION May 5, 1922, the thyroids of seven axolotls, seven inches in length, were removed under chloretone anesthesia and each gland transplanted intraperitoneally into the same individual from which it was taken. The idea was that the acquisition of a new blood and nerve supply by the gland in its new environment might permit the release of the accumulated secretion and so metamorphose the animal. It was shown by the writer (21) that the thy- roid glands of axolotls are highly active metamorphosis- inducing agents providing the hormone escapes into the 560 No. 647] METAMORPHOSIS OF AMERICAN AXOLOTL 561 blood stream. In these forms there appears to be some inhibition of the secretory (excretory) functions of the thyroid, and the hormone is retained within the gland vesicles, The experimental animals and their controls were kept in large aquaria with plenty of water and food. One of the grafted animals had metamorphosed by June 27. Three others transformed by July 1; a fifth animal died without transforming July 6. The two remaining axolotls had not metamorphosed by September 1. Dur- ing the interval between May 21 and September 1, all of the controls spontaneously transformed. The experi- ment is, of course, without significance because of the unstable nature of the control material. It is highly probable that the operated animals would have meta- morphosed just about as rapidly if the thyroid had been left in its normal position. EXPERIMENT 2. Homopuastic THYROID TRANSPLANTATION Five seven-inch axolotls were engrafted intraperitone- ally with the thyroid gland of other animals of similar size and appropriately controlled by at transplanted with pieces of muscle tissue. The transplants were made May 2, 1922. One animal had transformed by June 3, a second by June 6, a third June 11. Two animals remained as larve and were re- engrafted June 11 with axolotl thyroids, and metamor- phosed by July 3. In the meantime the controls also transformed. A large series of transplantation experi- ments were performed, using various endocrine glands,’ but in every case except two experiments the controls metamorphosed along with the operated individuals. Heteropiastic THYROID TRANSPLANTATION Four eight-inch axolotls were engrafted intraperitone- ally with the glandular tissue of adult Necturus macula- tus. Each axolotl received the entire thyroid of a single 562 THE AMERICAN NATURALIST . [Vor. LVI Necturus. The experiment was performed May 15. By June 11 all of the engrafted animals had transformed but none of the controls for this particular group, though nor- mal, untreated animals used as checks for other experi- | ments were metamorphosing during this interval. Despite the unstable nature of the control material used, this experiment seems fairly sound and indieates that Necturus thyroids when injected in sufficient quan- tity will metamorphose axolotl. To be absolutely reliable this experiment should have been performed upon thy- roidectomized forms, but unfortunately the unsuitable nature of the controls was not known until too late. THYROID FEEDING EXPERIMENTS Five six-inch axolotls were fed desiccated thyroid tis- sue (Parke, Davis and Company), containing 0.21 per cent. iodine by weight. The feeding was done by means of a pipette May 18. Two animals had transformed by May 27, and all by June 10. None of the controls meta- morphosed during this interval but all transformed by July 25. The experiment seems trustworthy, especially in view of similar results obtained by other investigators on animals of the European strain. HETEROPLASTIC PITUITARY TRANSPLANTATION Five axolotls varying in length from four to seven inches were each grafted with two whole pituitary glands of adult Rana clamata frogs. The grafts were made May .5. June 3 one animal metamorphosed; June 7 a second transformed. June 10 the three remaining animals were reengrafted with frog pituitaries. All metamorphosed by June 25. During the interval between May 5 and June 25 only two of the controls for this particular group transformed, but it must be remembered that control animals of other cultures were metamorphosing. The experiment is re- corded for what it is worth, but the writer believes that No. 647] METAMORPHOSIS OF AMERICAN AXOLOTL 563 injection of fresh pituitary substance does induce axo- lotl metamorphosis possibly by serving to release the thyroid hormone. This experiment should be tried on the Mexiean strain of axolotl which apparently rarely spontaneously metamorphoses and hence can be safely controlled. THYROIDECTOMY AND METAMORPHOSIS Eight axolotls varying from seven to fourteen inches were thyroidectomized and at the present writing, Sep- tember 1, are still larve and show no indications of transforming. Out of the original one hundred and nine animals received from New Mexico these eight are the only ones that have not metamorphosed. It is a fairly safe assumption that these axolotls will remain perma- nently as larva now that the thyroid gland is lacking. The thyroids of several animals were removed after the onset of metamorphosis, i.e., after the tail fin and gills were undergoing reduction, but in all cases the removal of the thyroid failed to prevent the completion of meta- morphosis. Discussion The conclusion to be drawn from these experiments is that the New Mexican strain of axolotl is entirely too unstable to work with on any problem involving the methods of feeding, injection or transplantation, where the results require a lapse of several weeks to obtain. The animals can not be controlled when the thyroid ap- paratus is left intaet. It is evident that conclusive ex- periments of the above kinds on the New Mexiean strain of axolotl (where the animals themselves are used as 1 The thyroidectomized animals were kept for five months and then in- jected with iodotyrosine and iodoserumglobulin. Metamorphosis resulted within a period of twenty days following injections of either substance. metamorphosed by injection of iodoserumglobulin. Injections of tyrosin dibromtyrosin and globulin had no effect upon metamorphosis. Uhlenhuth’s conclusion that only thyroid iodine (iodine which has undergone transforma- tion within the thyroid gland) is tena in metamorphosing urodele larve is invalid. 564 THE AMERICAN NATURALIST [Vor. LVI experimental material) ean only be obtained by using thyroideetomized animals. Professor Henry Laurens, of the Department of Physi- ology, informs me that several years ago he had a similar experience with axolotls from New Mexico. He received a shipment of several dozen in the spring, but was unable to prevent them from transforming shortly after arrival in New Haven. Only one animal of the lot failed to metamorphose and was kept two years in the laboratory, attaining a length of 14.25 inches. "This individual was used by the writer for thyroid transplantation work. The marked tendency of the New Mexican and other American axolotls to metamorphose spontaneously when moved from one locality to another prevents their being used for aquarium purposes. It is an odd fact that prac- tically the only axolotls used as aquarium material in the United States are those that have been shipped from Europe. : The European strain seems to differ from the New Mexiean form in regard to spontaneous metamorphosis, because these animals are handled by practically all aquarium dealers in Germany and ean be obtained for a few cents apiece. Apparently they rarely spontane- ously transform according to Jensen (’20), who has worked extensively with this strain. The curious thing about the New Mexican strain is that in their native habitat they too may remain for considerable periods as larva, yet when shipped from New Mexico to New Haven promptly metamorphose regardless of size or age. One large animal of this strain obtained by Professor Lau- rens failed to transform and was kept in the laboratory for two years; at the end of this time it showed no indica- tions of metamorphosis and was killed for thyroid trans- - plantation work. According to Gadow (708) the strain of axolotls estab- lished in Europe came originally from the vicinity of Mexico City. The first axolotls were brought to France by Marshal Forey in 1863, and the present strain is de- No. 647] METAMORPHOSIS OF AMERICAN AXOLOTL 565 scended from these animals. Gadow also states that the axolotls of Lake Xochimileo have never been known to metamorphose in their native habitat. However, several of the descendants of the animals taken to Europe did metamorphose, so that spontaneous transformation in the Mexican strain does sometimes occur, though rarely. In an earlier paper (722) the writer showed that the thyroid mechanism of axolotls is filled with physiologi- eally aetive hormone capable of inducing metamorphosis but that the secretion is apparently not liberated into the blood stream, hence the retention of the larval characters despite the possession of a large well-formed gland. The thyroid of a fourteen-inch axolotl several years of age was extirpated and eut into small pieces, each piece then transplanted into an immature Anuran larva. The single axolotl thyroid promptly metamorphosed five such tad- poles within fourteen days, whereas left intact within the axolotl's body it was quite ineapable of inducing trans- formation. This same experiment was repeated upon thyroidecto- mized and hypophysectomized Rana sylvatica tadpoles with similar results. Small pieces of axolotl thyroid when engrafted into thyroidless and pituitaryless larve promptly induce metamorphosis within ten or twelve days. It is quite clear from these experiments that axolotl neoteny is due to retention of the thyroid hormone within the gland vesicles. Under normal conditions and in its native habitat, the releasing mechanism apparently fails to act, but when the animals are shipped from one place to another and subjected to new environmental conditions metamorphosis promptly ensues. In the New Mexican strain slight stimulation is sufficient to initiate meta- morphosis, but in the European and Mexican forms very powerful stimulation is needed to overcome the thyroid inhibition and release the secretion. In the European strain the following agents have been used successfully for inducing metamorphosis: thyroid feeding (Laufber- 566 THE AMERICAN NATURALIST [Vor. LVI ger 713), salicylic acid injections (Kaufman 718), iodine and iodoform injections (Hrischler ’18-’19), organic io- dine feeding —iodothyrosine, also injections of iodocasein, iodoserumglobulin and iodoserumalbumin (Jensen ’21) ; and of course Marie von Chauvin’s experiments are well known. It is evident that the peculiar thyroid inhibition caus- ing neoteny in axolotl is due to genetic factors and that the condition is hereditarily transmitted. It is interest- ing to note that in axolotl we have one of the best ex- amples of hereditary transmission of an endocrine defect known. Attempts to explain neoteny by assuming that environmental agencies such as cold, altitude and the like are the chief causative factors are too crude to be seri- ously considered and for this reason—the aquarium deal- ers of Europe breed their animals as larve and the young grow up as axolotls, the matter of cold or altitude not en- tering into the question. As was previously mentioned, the European strain arose from a few animals taken to France in 1863. Then, too, both Professor Laurens and myself received our animals from Albuquerque, New Mexico, where they breed. The animals were old when captured. The tem- perature of the pools in the vicinity of the city can not be very low even in winter—not nearly so cold as those of the middle western states, northern New York, Ohio, or Wisconsin—and axolotls have never been reported as oecurring in these states so far as the writer is aware. The Amblystoma tigrinum resulting from the meta- morphosis of the axolotls during my experiments were placed in certain pools in the vicinity of New Haven where other species of Amblystoma are known to breed. The animals are full grown and should breed next spring (1923). By following the life history of the larve it is hoped that some new light may be shed upon the obseure and much debated problem of the relation of neoteny to environment. No. 647] METAMORPHOSIS OF AMERICAN AXOLOTL 567 LITERATURE CITED Chauvin, ai 1875. ber die Verwandlung des mexikanischen S e rai Zeitschr. f. wissensch. Zool., Bd. 25, Suppl., olot und ies 27, 1876. Gadow, H. 1908. Through Southern Mexico. Charles Scribner’s Sons, ve k. ame G. 1918-19. Sur la metamorphose didit: chez l'axolotle à e d'iode et des expérience apparentées. Extrait de Ko wl. A 3 Soc. Polonaise d. Naturalistes à Leopol. (Cited by Kopee, Biol. ull., Vol. XLII, No. 6, 1922. Jensen, C. O. 1921. Métamorphose provoquée par l'injeetion de prépara- tons deca et de thyroxine a des Axolotls ayant subi la eité d'animaux t Vid hake ompt. rend. soc. de bio!., 85, Kaufman, L. Researches on the emu Metamorphosis of Axolotls. Bull. de VAc C. a Laufberger, V. 1913. O SEDAN apnd mt Mdh krmenim zlazou stitnou. cep ber Lysty. (Cited by Adler, L., 716, Arch. f. d. ges. Physiol., bg Shufeldt, R. id Mexican Axolot] and its Susceptibility to Trans- sient > satin Vol. 6. Swingle, W. W. 1921. A. ees me the Thyroid Glands of Necturus and Azolotl. Anat. Record, 1 IND. 4. De 100.. 1922. B. Thyroid Glands of -the cni ipei Amphibians, Anat. Record, Vol. 23, No. 1, page 106. 1922, C. Experiments on the Meta- morphosis of Neotenous Amphibians. Jcur. Exp. Zool., 1922, Vol. 36. SHORTER ARTICLES AND DISCUSSION MORE EYELESS CLADOCERA Just before a note appeared in Science (Vol. 53, pp. 462- 463, May 13, 1921) concerning an eyeless cladoceran individual (a Simocephalus exspinosus), two additional eyeless daphnids occurred in another species of the experimental stock at the Station for Experimental Evolution. These were among off- spring of some Moina rectirostris which were being subjected to crowding in a sex-control experiment (10 mothers in each 130 e.c. wide-mouthed bottle containing about 75 c.c. of culture medium). While these two eyeless young were released on sue- eessive days and possibly in separate bottles, they were in bottles whieh belonged to the same series and received the same treat- ment. The precise identity of the mother of neither eyeless young could be determined (since there were 10 mothers producing parthenogenetic young in each bottle), but it is certain that the mothers were normal-eyed and were sisters, or came from mothers which were sisters. All of the mothers’ collaterals, which were examined, approximately 250, had normal eyes. 302 other young, produced by the 10 mothers in the bottle in which the second of these eyeless appeared, were normal. In all about 5,953 young were microscopically examined—a few of which were presumably sisters of the eyeless individuals and the others of which were young from sisters of the mothers of the eyeless individuals. All were normal-eyed. One of these eyeless individuals produced 5 broods, contain- ing in all 66 young, all normals. The other produced 4 broods, containing 38 individuals, all normals. 841 offspring of daugh- ters of the one eyeless, and 412 offspring of daughters of the other eyeless were found to have normal eyes. All examined among the collaterals of the eyeless individuals, 5,953 in all, and 1,357 direct first and second generation descendants of the eyeless mothers themselves—a total of 7,310—were normal. Hence despite the fact that there were two eyeless individuals produced by sisters (or by individuals whose mothers were sisters), while among many thousands of Cladocera previously seen under the microscope only a single similar individual had 568 No. 647] SHORTER ARTICLES AND DISCUSSION 569 been found, eyelessness in these individuals was clearly not in- herited. The lack of inheritance in these Moina rectirostris would have been anticipated if due regard had earlier been given to a peculiar feature of the head of these eyeless indi- viduals. This will be discussed in a later paragraph. The next occurrence of eyeless Cladocera was in February, 1922, when seven eyeless Moina macrocopa were found among 147 young of the third brood from 10 mothers in a crowded bottle. The culture water in this bottle seemed rather cloudy, an appearance known frequently to be associated with unfavor- able conditions which sometimes result in death to part or all of the Cladocera in such a bottle.' In the present ease in addition to one eyeless male and 6 eyeless females among the 67 females and 80 males in the bottle, there were other abnormals—6 or 8 with abnormal eyes (pigment reduced or eye not completely formed) and perhaps an equal number with abnormal antenne (certain segments missing, aborted or fused with others) and one male with an abnormal eye and an abnormal antennule. Some of the eyeless individuals and some with abnormal eyes had abnormal antenne also. Others showed abnormality in only one feature. Sinee these abnormals appeared in a erowded bottle (10 mothers) it is impossible to know, but they probably did not come from a single mother. Among the next brood of young from the same mothers were a few with abnormal an- tenne and slightly abnormal eyes. Subsequent young were normal. Early attention to an interesting feature of the heads of these eyeless individuals removed any temptation to anticipate in- heritanee of eyelessness in these cases; and, as expected, all the numerous young examined from tha eyeless individuals (and from the other abnormals as well) were normal. Since in these eases eyelessness was not hereditary some developmental aeci- dent would seem probably responsible for its oeeurrenee. In- deed, it seems fairly evident, in view of the occurrence of other abnormalities in the same and other similar eulture bottles, that these abnormalities were related to some unfavorable fac- 1In other cases such conditions hey the culture medium were associated with pigmentless eyes in some of the newly released young. However, the pigment develops to its full nea in from one to five days after the young animals are released from the mother’s brood chamber. Newly released young from the formerly pigmentless-eyed individuals have nor- mally pigmented eyes from the first. 570 THE AMERICAN NATURALIST [Vor LVI tor or factors in the environment, although nothing definite is known as to what these factors were. A peculiar structural feature of the heads of the young eye- less individuals suggested the possible manner in which eyeless- ness came about in these cases. When young, the seven eyeless Moina macrocopa had on the anterior head margin a small nodule or excrescence which, though not so conspicuous at later stages, yet in most cases persisted through several moults. In each of these eyeless individuals the optic ganglion was reduced or lacking, and the margin of the head was readjusted to com- pensate for the reduced and missing organs. Substantially the same structural conditions were found with the two eyeless Moina rectirostris, absence or reduction of optie ganglia, the shortening of the head margin and the occurrence of a small bit of apparently necrotic material attached to the front of the head.’ It seems possible that this apparent exudate on the heads of the eyeless individuals really represented an aborted or necrotic portion of the embryo which included the primordium of the missing parts.? The fourth occurrence of eyeless Cladocera (the eleventh eye- less individual seen) was June 26 in a crowded bottle of Moina macrocopa. In addition to the lack of eye and of optie ganglion, the brain proper was reduced in size. This animal was not ex- amined until mature and an excrescence on the head, if present in the young animal, had by that time disappeared. This indi- vidual swam in small circles, although its swimming organs ap- peared entirely normal. It died after producing two broods (10 females and 12 males) of normal young. The occurrences of eyeless Cladocera have included three species, eleven individuals and four different time periods. The last three occurrences, and probably the first one, were in erowded bottles, suggesting environmental factors as causative 2 That this material was intimately associated with the head structures and really a part of the animal is attested by the fact that it persisted through eedysis, whereas any material merely esie to the external surfaee of the exoskeleton would be eliminated by eedys 3 A somewhat similar appearance in larve arising om pE eggs of Ambystoma punctatum was presumably correlated with failure of development of the anterior part of the head. (Banta, A. M., and Gortner, A., ‘‘Aceessory Appendages and Other Abnormalities Produced in Amphibian Larve through the Action of Centrifugal Force,’’ Jour. Exp. Zool., 18: 433-446, pls. 1-3. 1915.) No. 647] SHORTER ARTICLES AND DISCUSSION 571 agents. Those which lived to produce young gave rise exclu- sively to normal young, indicating that genetic changes were not responsible for the abnormal heads. However, in view of the known inheritance of eyelessness in cave arthropods and vertebrates and in Drosophila melanogaster, it seems of interest to examine each case of profound eye modification in erusta- eeans and elsewhere to gain information on the origin and in- heritanee of any possible mutation of this character.* ARTHUR M. Banta . A. Brown STATION FOR EXPERIMENTAL EVOLUTION COLD SPRING HARBOR, L, I. CROSSING-OVER INVOLVING THREE SEX-LINKED GENES IN CHICKENS IN the course of the last year several crosses of chickens earried out at the genetics station at Anikovo (near Moscow) have made it possible to observe erossing-over in this form. The genes ‘‘suke,’’ ‘‘tuge’’ and ‘‘trage’’ were studied. The first, suke, retards the development of feathering in the chicks, so that at the age of 1 to 1.5 months they have very small tails. The development of the wings, too, is very slow. The genes trage and tuge together eause the well-known Plymouth Rock mark- ings, trage eausing the erossbarring, and tuge (not very visible in Plymouth Rocks, where it causes the contrasts in the mark- ings) is the same gene as silver coloring, which was first re- ported by Hagedoorn in the Assendelver chickens. Later (1912) Davenport observed it in the cross of Dark Brahma X Brown Leghorn, where, however, on account of the absence of several other genes, tuge has very little expression—only as a whitish edge on the feathers. The genes suke, tuge and trage are all present together in the Plymouth Rceks. The Russian Orloff chickens have none of these genes, a condition which may be expressed as asuke-atuge- atrage. All these genes are sex-linked, and therefore are trans- mitted with complete linkage from mother to son. The cross 4Since this manuscript went to the printer two more eyeless Moina macrocopa were found in a crowded bottle. These two with the last one mentioned above were the only eyeless occurring among approximateiy 33,000 individuals microscopically examined (in sex-control experiments) during three months. The facts, that of these three two occurred in the same bottle and that the character is not inherited, again indicate clearly enough that external, not internal, factors are responsible. 572 THE AMERICAN NATURALIST [VoL. LVI Orloff male Plymouth Rock female gives cocks closely re- sembling the true Plymouth Rock, that is, erossbarred with slow feathering development. All the hens, however, are black (since in the Plymouth Rock there is also a gene for melanism, ‘‘tifa,’’ which is not sex-linked), and they develop feathers quickly. d: asuke atuge atrage atife X 9: suke tuge trage tifa F, d: suke tuge trage tifa 9: asuke atuge attage tifa In F, the coupling between suke, tuge and trage becomes broken, and different new combinations are to be observed in rather large numbers. More often the forms asuke-tuge-trage are obtained, colored like Plymouth Rock, but with quick de- velopment of feathering (among these there are also cocks), and conversely suke-atuge-atrage, with slow feathering, but black (when tifa is present). In one ease a suke-tuge-atrage chick appeared, with slow feathering and silvery, but not crossbarred. In the light of the Morgan theory these facts can be explained by regarding the genes suke, tuge and trage as being in a sex chromosome which cannot give crossing-over in the heterozygous sex (female). But when the same chromosome is transmitted to the F, male, it undergoes crossing-over with its partner, which occurs most often in the space between suke on the one side and tuge-trage on the other. Crossing-over between suke-tuge on the one side and trage on the other oceurs less often, wherefore the arrangement of the genes in the F, may be represented as follows: trage However, the counts of chicks which have so far been obtained in F, are not yet large enough to ascertain definitely the order of the genes, and therefore still less the exact distances, S. SEREBROVSKY INSTITUTE OF — Brioroav, 41 SrvTSEV VRAZ Moscow, pps 21, 1922 [Crossing-over between ''suke"' (barring) and ‘‘tuge’’ (sil- very) has also been announeed by Goodale (1917) and by Hal- dane (1921), in the papers listed below, which were not available to the above author. i: Goodale, H. D. 1917. Crossing-over in the Sex Chromosome of the Male Fowl Science, N. S., Vol. 46, p. 213. Haldane, J. B. S. 1921. Linkage in Poultry. Science, N. S., Vol. 54, p. 663. Note of Transmitter, H. J. Muller.] No. 647] SHORTER ARTICLES AND DISCUSSION 573 A FOURTH ALLELOMORPH IN THE ALBINO SERIES IN MICE? In recording the occurrence of a new mutant gene in the house mouse, allelomorphie to color and albinism, Detlefsen (721)? described a very dilute, wild form in which the hair showed traces of a light brownish tinge with a suggestion of sootiness, and the eyes were somewhat less heavily pigmented than in the wild type. This general form of pigment reduc- tion is also eharaeteristie of other eolor allelomorphs; for in the ease of the ruby-eyed rat, the ruby-eyed guinea-pig and the chinchilla rabbit (Castle 721), the yellow pigment is very greatly reduced or even obliterated, while the darker pigments (black or brown) are at least slightly modified. The mutant mouse, however, showed a far greater pigment reduetion than either the rat, guinea-pig or rabbit mutants. Breeding tests demonstrated that this dilute mouse mutant was a color-albino allelomorph, and in this respeet resembled the ruby-eyed rat and guinea pig genetically (the chinchilla rabbit had not been recorded at that time), but Dr. Detlefsen pointed out that ‘‘it is hardly safe to insist that these mutations are identical. . . . : We are also unable to prove that they are different, for the genes may be identieal but simply give different somatie ef- feets, sinee the residual inheritanee ean not be the same." He also suggested that the diseovery of a new dilute type of mouse (whieh he was seeking at that time), more like the rat or guinea pig in its somatie appearanees as well as in its genetie behavior, would give us more assurance that his extreme dilute mouse mutant was not the homolog of the ruby-eyed rat or guinea pig. Unusual as it may seem, I had discovered exactly such a new dilute mutant mouse in January, 1919. By com- paring it with Dr. Detlefsen's set of rodent skins and by test- ing it in appropriate matings, I reeognized its genetie signifi- eanee just before his paper appeared in print. The diseovery of this new mutant mouse enables us to say at onee that the extreme dilute mutant was not the homolog of the ruby-eyed rat or guinea pig or the chinchilla rabbit, and supports Dr. Detlefsen's position in hesitating to homologize 1 Paper No. 22 from the Genetics Laboratory, College of Agriculture, University of Illinois 2 Detlefsen, J. A., 1921, AMER. Nat., Vol. 55, p. 469. 3 Castle, W. E., 1921, Science, N. S., Vol. LIII, p. 387. 574 THE AMERICAN NATURALIST [Vou. LVI his mutant with these forms. Dr. Detlefsen’s form is evidently lower in the scale running from color to complete albinism in very much the same way that the Himalayan form is nearer to the albino than is the chinchilla rabbit The new mutant was procured from a fancier who had been breeding it for some time. It resembles the ruby-eyed guinea pig, ruby-eyed rat and the ruby-eyed or chinchilla’ rabbit (Castle ’21)* in the degree of pigment reduction in the hair, but the eyes are apparently darker than those of the rat and guinea pig. I have not had an opportunity to examine the eyes of the chinchilla rabbit. It forms one of a series of quad- ruple color allelomorphs in the mouse and may be designated as c”. In a scale of dominance, the four forms probably fall into the following order: ordinary intense or wild eolor, C; dilute, c" (described in this paper); extreme dilute, c^ (de- scribed by Detlefsen ('21);* end complete albinism, c, Wild eolor (C) is eompletely dominant to the other allelomorphs, but c” and c? are incompletely dominant to albinism. The cross between c" and c^ has not yet been made, but the heterozygote (c"c^) will probably be found to give an intermediate shade. The black agouti type of the homozygous mutant (AABBc'c") possesses black pigment which is reduced to a/very dark dull slate-color, while yellow is greatly reduced and appears about intermediate between white and the normal yellow of the wild type. In the non-agouti type of the homozygous mutant (aaBBe'c"), which ean be distinguished readily from the agouti form, the black pigment is also reduced to a very dark dull slate-eolor, but perhaps darker than in the agouti type. en the blaek agouti type of the mutant is heterozygous for albinism (AABBe'c), black pigment is reduced to a brown- ish shade and yellow is practically reduced to white. In the non-agouti type of the heterozygous mutant (aaBBc'c), black is reduced to a dull brown, a little lighter than the ordinary fancier’s chocolate type. The heterozygous mutants, mated interse, produce the rae ea ns type, the heterozygous type and albinos in the ratio of 1: 2: 1. I have not yet identified the mutant without black pigment that is in the cinnamon or brown class. H. W. FELDMAN 4 Loc. cit. 5 Loc. cit. INDEX NAMES OF CONTRIBUTORS ARE PRINTED IN SMALL CAPITALS Alcohol and White — EDWIN CARLETON MACDOWELL, 289. ALLEN, E. J., Frera of Life in the Sea , 481. Asterias, Tube- feet in, RoBERT H. BowEN Axolotl, Spontaneous Metamorpho- sis of; . SWINGLE, 560. BAN ARTH M., ‘and L.' A indie ade vx es 568. BELLING, JOHN, ALBERT F. BLAKESLEE, Chromosomes in Trip- loid Daturas, 339; A. F. BLAKES- LEE and J. RTHUR HARRIS, Dom- inant ee 458 LBERT FRANCIS, ve d B., Sexual and Sex- init. Characters, tur P MHALL, R5 mily R- mblanees can bibe Men of eg ce, 504. Brow L. A. and ARTH M. Bats, Eyeless Cladocera, "568. CHIDESTER, F. E., Fish Migration, 373. Chiton, Color Variations * (luster -formation J. Crozier, 478. in, adi of Sperm i Growth Factors i d ae ewe Sar CHUR JR. ; URNA OR in n Shelled Tot d Cicada, Fugitive Net- seng in, W. ToM. » 191. Coleoptera Food Habits of, Harry B. Weiss, 159. CROZIER, W. J., Color Variati ns in Chitons 189; * * Cluster-forma- tion? of Sperms of Chiton, 478. Datura, Variations in, ALBERT FRAN- cis BL*KESLEE, 16; Chromosomes in Triploid, Jonn BELLING and Y: PAYNE Gynandromorph in Trosi Melanogaster, 383. Dominant eit gee = Tus BLAKES- LEE, and J. ARTHUR Pise 7 458. Drosophila, ag aset Complete Linkage in, MARIE S. and Jo Willistoni, LANCEFIELD and x I HARLES W. Salamand^rs ‘Phylogeny P eae 418. Du rose: of Buff rnd tin Fowls, 242; Black Pig- ment in Domestice Fowls, 464, and Emerson, R. A., Bud Variations, 64, Environment, Effect of on Animals, 144. Bae, Paleontology of Arrested, F RUEDEMANN, 256. Family Resemblanees among Ameri- ean Men of Seienee, DEAN 0 FASTEN, NATHAN, Tareworm In'ce- . ourth All lo- MM ‘in Albino Boden in Mice, 573 Fishes, Variati HUBBS s, 360; geo hs us TY. E CHIDESTER, 373. FISHER, A., Genes, è FORB . T. M., Fugitive Net- vei Cicada, 191. MN Crosses m ies and Colum- C. Puit in Domani L. C. Du Frogs, Transformation of Sex in, W. W. SwINGLE, 193. Genes, are. Bo: LR 32; Loeat n of, R. A. Fis 406; Cro eb over involving Three Sex linked, in Chickens, S. SERE- BROVSKY, : GOVAERTS, ALBERT, and J. ARTHUR a Assortive ‘Mating in Man, cers Mar S. and Joun W., raiik ge mag 286. Guyer, M. F., Serological Reactions 80; Or rthogenesis of Sero logieal Phenomena, 116 Harris, J. ARTHUR, and ALBERT GOVAERTS, jere i Mating in Man 381; A. F. BLAKESLEE and JOHN BELLING, Dominant Individ- uals, 458, 575 576 HAUSMAN, LEON A., Miero-filter for enn Flagellates, 284. HENDE