living specimens. When the animal was cut into two parts just behind the cerebral ganglion and above the beginning of the digestive tube, so as not to leave the slightest part of this last organ in the front piece, the hind part was nevertheless regenerated in full, including the whole of the digestive tube. The regeneration in full of such an important organ was thus accomplished by another organ, from which the former is quite independent, both anatomically and genetically. The new organ was grown in consequence of a re-differentiation of the elements of a neighbouring, quite different organ.' "1 31 Another fact, still more interesting for all theories of heredity, was established by the same experiments. It was the important part taken in regeneration by the so-called ' wandering cells.' As soon as the regeneration process had begun, these cells became active, and they acted as phagocytes; they absorbed particles of reserve materials lodged in the tissues, as also the grains of pigment, and then themselves were eaten up by the cells of the regenerated parts, and thus aided the growth of the latter. From the few facts just mentioned it is already obvious that the researches on regeneration do not yield support to those hypotheses of heredity which make no allowance for the inheritance of modifications acquired by the body-cells. In fact, when Weismann framed his hypothesis in the years 1883-1885, he paid little attention to regeneration. But after the work of Götte and Fraisse he had to take it into account, and so he did in his Germ-Plasm, published in 1892. It was shown by Götte that when the amputated leg of a Triton is regenerated, all the parts of the leg-its bones, its muscles, the connective tissue, the mucous glands, the nerves, and the blood-vessels-are regenerated in full. And the question arose : Wherefrom come the determinants which 'marshal' the regeneration? Weismann answered this question by adding a new hypothesis to his previous ones. The cells which are capable of producing regeneration-he wrote in Germ-Plasm-must contain, besides the principal idioplasm (Haupt-Idioplasm), additional idioplasm (Neben-Idioplasm), which consists of the determinants of those parts which it may have to regenerate. Thus all the cells of the bone of the [amputated] front leg must contain, beside the determinant No. 2, also the determinants 3 to 35 of the additional idioplasm, as they have to reconstitute all the succession of bones "Jósef Nusbaum und Mieczyslaw Oxner, 'Studien über die Regeneration der Nemertinen,' i., in Roux's Archiv für Entwicklungsmechanik der Organismen, Bd. xxx. (Festband), Th. I., 1910, p. 115. These researches-the authors remark-are a further development of the observations made in 1909 by Prof. Davydoff, and a similar process was also described by MM. Salensky and Lebedinsky in 1897. in the front leg.""2 Besides, there is a sufficient number of 'lieutenant' or 'reserve-determinants' (Ersatz-Determinanten) lodged along special germ-tracks' running from the reproductive cells to different parts of the body. There are even two different reserve-determinants in the middle portions of the Lumbriculus : one for regenerating, in case of need, the front part of the animal, and another for regenerating its hind part.33 One may be tempted to treat all these suggestions as a mere work of imagination; but, let us not forget what Tyndall said in one of his brilliant lectures about the part of imagination as a scout in scientific research. Let us suppose that the imagined picture corresponds to some extent to reality. What would be the result? It would be that the impossibility' for the germplasm to be influenced in the proper way by modifications in the body-cells the inconceivability' of the process would be irretrievably gone! Once more we are thus compelled to recognise the existence of a close intercourse between the germ-cells and the body-cells. The fact is more and more firmly established. There remains only to see what sort of intercourse that may be. Are not the germ-cells influenced by the changes going on in the body-cells in the same sense as the latter, so as to be capable of reproducing the changes in the offspring? VII One of thechief results of the discussion which took place in the years 1880-1893, and in which Herbert Spencer took a prominent part," was to define more accurately the proper rôle of natural selection in the evolution of new species. It was shown that natural selection cannot be the origin of the so-called 'determinate' or 'cumulative' variation, unless there is at work some cause affecting many individuals at the same time, in the same direction, and for a succession of generations. A great number of biologists sought, therefore, the origin of variation—as Darwin had done-in the direct action of the surroundings; while those for whom the main thing was to repudiate the hateful 'Lamarckian factor' followed their spokesman, Weismann, who 32 Das Keimplasma, p. 137. 33 Das Keimplasma, p. 169. Contemporary Review, February to May 1893. Spencer's articles were reprinted separately as a pamphlet under the title The Inadequacy of Natural Selection (Williams & Norgate, 1893). Prof. Marcus Hartog contributed also an important paper to the same Review, 'The Spencer-Weismann Controversy. July 1893. was maintaining at that time the Allmacht-the all-sufficiency' of natural selection. Weismann, however, soon abandoned this position, thinking that he had found the true origin of variation in sexual reproduction, i.e. in the mixing together of the germ-plasms of the two parents during the process of fertilisation, to which he gave the appropriate name of Amphimixis. Microscopical investigation had shown, a few years before, that in sexual reproduction an actual mixing together of the two parent plasms takes place in the fertilised egg, whereupon one-half of the coalesced two plasms is thrown out-the remaining half only going for the development of the embryo. To take a familiar example, things go on as if two packs of cards, one of which represents the characters inherited from the father, and the other represents those of the mother, were shuffled together; whereupon the double pack is divided into two equal parts, one of which is put aside and the other retained. The number of determinants corresponding to the immense variety of characters of each parent being extremely great, the possible combinations of them are countless, and this is why two children of the same parents, or even two puppies of the same litter, are never quite alike. This was the idea developed by Weismann in 1886 and 1891.35 He was so much taken by it that he saw in Amphimixis the true cause of all inheritable individual variation—the keystone of the whole structure' of the theory of heredity.36 Therefore he still more emphatically denied the possibility of the germ-plasm being influenced by external agencies in the same direction as that taken by the somatogenic changes which follow the same. cause. 37 The only source from which inheritable individual differences could be derived was sexual reproduction-Amphimixis. 'It was only in this way that hereditary individual differences could arise and persist'; 3—only in this way new species could originate with the aid of natural selection. There is no need to say that such a position could not be maintained. The believers in the sufficiency of Amphimixis as a cause of variation were shown that if variation were limited to a redistribution of already existing characters, no progressive evolution would have been possible, and Weismann himself had to recognise the force of this remark, so that in 1904 he 35 The Significance of the Sexual Reproduction in the Theory of Natural Selection,' in Essays, i. 257-342, and 'Amphimixis; or, the Essential Meaning of Conjugation and Sexual Reproduction,' in Essays, ii. 99-222. abandoned the Amphimixis hypothesis in his Vorträge über die Descendenztheorie (The Evolution Theory in the English version). Even now [he writes in this work] I still consider Amphimixis as the process by means of which new re-combinations of variations are produced— a process without which the building up of an organic world infinitely rich in forms and incomprehensively complicated could not have taken place. But I do not consider it as the true root of the variations themselves, because the latter cannot possibly depend upon a mere exchange (Austausch) of ids: they must rather depend upon a modification of the ids. . . . Amphimixis, i.e. the coalescence of two plasms, certainly does not modify the determinants: it only produces new and new combinations of the ids (the ancestral plasms). If the appearance of variations were limited to this cause, the transmutation of species and genuses would have been possible only within a very limited scope.39 But this abandonment of the Amphimixis hypothesis was not sufficient. Every hypothesis of heredity is bound to indicate the source of definite,' or ' cumulative' variation which accumulates from generation to generation certain changes in a given direction. In Weismann's terminology, it had to indicate the possible cause of a continued modification of the determinants of the germ-plasm in a given direction. Let us take, as an instance, the classical example, worked out by Professor Marsh, of the horse's hoof evolved out of the median toe. It so happens that we have a practically complete chain of the ancestors of the present horse since the Eocene period; and we see how, these ancestors having dwelt in regions with a hard ground, where increased rapidity of locomotion was an advantage, the median toe, which was becoming the chief support of the foot, gradually developed more and more, and its nail became a hoof; while the other digits, touching the ground no more, ceased to be used, and were atrophied, so that now they are only splint-bones. This is what the determinants' hypothesis had to explain, without recognising the inheritance of acquired characters which it repudiated. Weismann tried to do so by means of a new hypothesis-cf ' germinal selection,' or of struggle between the determinants in the germ-plasm. W. Roux had just published at that time a remarkable work in which he described the sometimes conflicting claims of the different organs upon the disponible stock of nutritive stuffs in the organism as a struggle between them. Weismann applied the same conception to the determinants within the germ-plasm, and described the effects of that competition as ' germinal selection.' 3o Vorträge - (2nd edition), ii. 163. However, the new hypothesis did not help to solve the difficulties. We certainly may use the word 'struggle ' as a metaphoric expression for processes far more complex in reality than a competition for nutritive stuffs; but we must not forget the physiological facts covered by our vague expression-those facts which Roux and his co-workers precisely are studying now at his laboratory for the study of the mechanics of evolution,' and of which the Archiv für Entwicklungsmechanik is the organ. Thus, reverting to the causes which may secure to the mediane toe a 'victory' over the other toes, they are of a physiological nature; namely, an increased use of the toe, leading to a stimulation of its tissues, and consequently to an increased nutrition of that toe. But how do these changes of nutrition of the different toes affect the germ-plasm determinants of these toes? Do they affect them at all? and, if they do, is there any parallelism in the changes of nutrition of the toes and the nutrition of the respective determinants? Formerly, Weismann used to assert most positively that there is no such correlation : not even an approach to it. 'I must thus affirm to-day, even more decidedly than I formerly did (Weismann wrote in 1892 in Germ-Plasm) that all lasting, i.e. inheritable, modifications of the body originate in primary modifications of the germs' predispositions (Keimesanlagen), and that neither mutilations, nor functional hypertrophy or atrophy, nor any changes that have originated in the body as the effects of temperature, or food, or any other influences of environment, are transmitted to the germ-cells, so that therefore they might become inheritable.' 940 The transmission of differences of nutrition from the digits of the horse to the determinants of these digits, so as to provoke an increased nutrition of the mediane toe determinants, and a decreased one for the determinants of the other toes, was thus absolutely excluded. Germinal selection having no relation whatever to the struggle' that is going on between the toes, it could even run in the opposite direction and favour the development of all the digits but the mediane one, thus counterbalancing the cumulative increase of the latter. Only in 1904, when Weismann published his Lectures on the Theory of Descent, he seemed to make a concession; he recognised that the effects of a varying nutrition may be such as better to feed and to increase certain determinants more than the others, in which case the organs they determine would also increase; but he expressed no opinion as to the possibility, or the impossibility, of the reverse effect: he did not say that the increase of an organ (of the mediane toe in our case) should result in a 40 Das Keimplasma: eine Theorie der Vererbung (Jena, 1892), p. 518. VOL. LXXI-No. 421 LL |