The Evolution Theory, Vol. 2 of 2

by August Weismann

Transition species of Celebes snails, according to Sarasin—Possible variations in the shell due to nutrition—Natural selection plays a part—Germinal selection—Temporary transitions between species—The fresh-water snails of Steinheim—How do sharply-defined species arise?—Nägeli's Developmental Force—The species a complex of adaptations—Adaptive differences between species—Adaptive nature of specific characters—The case of Cetaceans—Of birds—Additional note: the observations and theories of De Vries.

Our study of the influence which geographical isolation may have in transforming old and giving rise to new forms of life has led us naturally to a much more important problem, that of the origin of species as more or less sharply defined groups of forms, and I wish to make the transition to this problem by discussing another case of species-splitting effected in association with, or, as is usually said, through isolation. The naturalists Paul and Fritz Sarasin, well known through their excellent studies on many components of the tropical fauna, have published in their latest work interesting discoveries in regard to the terrestrial snails of Celebes. These observations show that on this island a great transformation of snails has taken place, even since the later Tertiary period. A large number of new species of snail have arisen on this island since that time, and this, as the authors show to be probable, in association with the receding of the sea, that is, with the elevation of the island further out of the water, and thus with the increase of its surface. The modern terrestrial snails show chains of forms connected in many ways so that a series of species is connected by transition forms, and therefore does not really consist of separate species at all, although the extremes would seem to be separate species if they were studied by themselves without taking the transition forms into account. The state of things is exactly as if a Tertiary snail had spread from any small area over the whole island, and had been transformed slowly and in a definite direction in accordance with its distance from its starting-point. It is thus that we must interpret this discovery; we have here, beside each other in space, and indeed often disposed along geographical lines, the individual stages of a phyletic process of transformation, which has reached different levels at different places. One of the longest of these chains of forms is that of Nanina cincta, which runs across the island from east to west, and, beginning with the smallest and most delicate forms, ascends through many intermediate stages to the giant form N. limbifera. Such chains of forms have been previously recognized; thus Kobelt described one in the case of the Sicilian land-snails of the genus Iberus, and other cases are recorded in literature, but in all instances they refer to areas which must be regarded as isolated for the snails, and which have been colonized from a single starting-point.

We have now to inquire whether and how we can explain the origin of these chains of forms. The cousins Sarasin tell us how they at first attempted to refer the differences between the individual links of such a chain to the diverse influence of the external conditions of life, but in vain; neither the height above sea-level nor the character of the soil was sufficient, and natural selection was no more so; 'for why should a high Obba-form twisted like a beehive be either better or worse equipped for the struggle for existence than a smaller and flatter one?' It is true that we do not understand why, but this does not seem to me any reason to doubt that natural selection should be regarded as one of the causes of the divergence of these species, for we could not answer the same question in regard to any of the other structural differences between two species of snail, for the simple reason that we have far too little knowledge of the biological value of the parts of a snail. Or could any one tell of what use it would be to a snail-species to have the horns slightly longer, the foot somewhat narrower, the radula beset with rather larger or more numerous teeth? We might indeed imagine many ways in which it might be of advantage, but we are not in a position to say definitely why, for instance, longer horns should be better for one species than for another, and yet we do not believe that the structure of snails is less well adapted to the life of each species than that of any other animals. The snail's structure is certainly built up of hundreds and thousands of adaptations, like that of every other animal species, but while in many others we can, at least in part, recognize the adaptations as such, we cannot do so at all in regard to the snail. Simroth has pointed out that the spiral asymmetrical shell bears a relation to the one-sided opening of the genital organs, but that only states the general reason for the coiling of the shell. In studying the differences in the shell one is apt to think of its external appearance alone, of the protection which it affords to the soft internal organs of the easily wounded animal; perhaps also of the distribution of weight, which must be different in a high tower-like structure and a low flat spiral; possibly, too, of the varied obstacles and resistance the snail has to encounter in creeping into clefts and holes, or among a tangle of plants, according to the form of its shell; but is it not also conceivable that the form of the shell has been determined by its contents? As Rudolph Leuckart taught, the snail may be regarded as composed of two parts, one of which is formed by the head and foot, the other by the so-called 'visceral sac': the former may be called the animal half, because it chiefly contains the dominant organs—the nerve-centres, almost the whole mass of muscle, and the sense-organs; the latter the vegetative half, since it contains the main mass of the nutritive and reproductive systems—the stomach and intestine, the large liver, the heart, the kidneys, the reproductive organs, and so on. The vegetative half of the animal is always concealed within the shell; would not therefore any great variation in the size of liver, stomach, intestine, and so on, bring with it a variation in the size and form of the shell, as well as in the expansion or contraction of its coils? And might not such variations become necessary because of some change in the food-supply? It is only a supposition, but it seems to me very probable that becoming accustomed to a new diet, less easily broken up and dissolved and of diminished nutritive value, would cause modification not only of the radula and jaw-plate, but also of the stomach and the liver, the intestine and the kidneys, whose activity is closely associated. The stomach must become more voluminous, the liver which yields the digestive fluid must become more massive, and so forth. I will not follow this hypothetical example further, for I merely wished to recall the fact that the snail shell, to the form of which no biological significance can be commonly attributed, is actually a sort of external cast of the visceral sac, and consequently dependent on the variations to which that is liable in accordance with the conditions of its life. To give precise proofs for such processes is certainly not yet possible, for we do not even know with certainty what the diet of the various species of snail is, much less the difference between the modes of nutrition in two varieties, or the nutritive value of the materials used, or the changes in secretion, absorption, assimilation, and excretion which must be brought about by these differences. But we can at least see that variations in nutrition must be enough in themselves to give rise to new adaptations in the size, constitution, and mutual adaptation of the internal vegetative organs, and we cannot overlook the possibility that the form and size of the vegetative half, and therefore the form and size of its secretion, the shell, may also be caused to vary[29]. The fact that we cannot recognize, for instance, the beehive shape of an Obba as an adaptation, is thus no proof that it is not one. But let us assume for the moment that it is not, and that it cannot be referred to natural selection any more than the other variations in the Celebes chain of forms, and we may further admit that they cannot be referred to sexual selection, still less to some 'inherent principle of perfecting,' not only because there is no question of perfecting in the matter, but because such a mystical principle is outside of the scope of natural history and its principles of interpretation.

[29] That this suggestion was not unjustified is evident from a recent contribution by Simroth ('Ueber die Raublungenschnecken,' Naturwissenschaftliche Wochenschrift, December 8 and 15, 1901). In this paper the author, who is an expert as regards the biology of Gastropods, shows that a change of diet may evoke many kinds of changes in the structure of the food-canal, which may indirectly compel changes in the shell. Thus in a small indigenous snail, Daudebardia, the pharynx has grown so enormously in thickness and length in adaptation to the predatory mode of life, that the head and the anterior part of the body can no longer be retracted within the shelter of the shell. For this reason, and also because of the snail's habit of following earthworms into their burrows, the shell has been shunted far back and obliquely downwards. It has at the same time markedly changed in its shape, as may still be verified by comparing the form of the shell in the young stages with that of the adult.

But that transformations in a definite direction can and must arise from fresh disturbances in the equilibrium of the determinant system, that is from germinal selection, we have already shown.

Even if the changes of form with which we are here dealing had really no biological importance, they might quite well have been brought about by germinal selection, and only one thing remains obscure, that is, why the different stages on the path of distribution of a species are at a different level of evolution and not all at the same level. Why have not all been transformed? Why have some colonies remained near the ancestral form, while others have varied only a little, and others again a great deal? This cannot be explained by any assumption of an internal power of development, and the explanation can only be found in germinal selection associated with isolation, since the internal processes in the germ-plasm can quite well run a different course in different colonies. Nevertheless I am inclined to infer from these differences in the individual colonies of these chains of forms, that natural selection in the accepted sense has also played a part in the evolution of these snail varieties.

Such series of forms are especially interesting, because they show us the process of species-formation in its different stages beside each other in space, and thus simultaneously. They represent, so to speak, a horizontal branch of the genealogical tree of the species, as the Sarasins well express it, that is, a series of species arising from each other, which do not break off, but are all capable of life at the same time, and so exist simultaneously on different areas; they are species adapted to different localities, not to different times. The same is true of the snails of other isolated regions, except that the chains of forms are usually not simple, but split up into several chains of forms arising from one ancestral form, and under certain circumstances each of these may break up into two or more diverging series. The great number of related species in Madeira, or in the Sandwich Islands, compels us to this assumption, although the branching of the genealogical trees can no longer be demonstrated with certainty.

This splitting up of forms into several series on a varied insular region shows us once more that it is germinal selection alone which forms the basis of all transformations, and that there is not, as earlier naturalists, especially the botanists Nägeli and Askenasy, maintained, any peculiar impelling Force of Development innate in organisms. If there were such a force, a species would be obliged to go on continuously in the same direction, exactly like the Sarasins' chains of forms, but no breaking of the species into one or many forms could occur. But this breaking up into series is easy to understand when we take germinal selection into consideration, for the germ-plasm contains many ids and determinants, and each of these can enter upon new variations, so that one colony can vary in this direction, another in that, and a great diversity of forms living in isolation must, or at least may be the result, as we see in the case of the Sandwich Islands.

Let us delay a moment over the Sarasins' case of the Celebes snails. We are dealing here with series of forms in regard to which the ordinary conception of species fails us, for they contain varieties whose extremes are as far apart as distinct species usually are, which are not, however, distinct, since they are connected with one another by one and often by several transition forms. Thus we can only break them up into two or more 'species' by an arbitrary division at one place or other. The phenomenon itself is not new to us; we have seen that even Lamarck and Treviranus made use of similar series of forms, connected by transition stages, in their attack upon the old theory of creation, and sought to prove by means of these that the idea of species is an artificial one, read into nature by man, and not innate in nature, and that the forms of life were only apparently fixed and sharply defined, being in reality in process of slow transformation. Such beautiful and convincing examples as we now possess were not available at that time, but it might even then be said that it was the easier to make a new species the fewer examples one had to deal with, and the more difficult the more numerous these became, because with the number of individuals, especially if they come from a wide area, the number and diversity of the divergences increases also, so that in many cases, as in that of the Celebes snails, it becomes impossible to draw a line between the different species.

There are, however, many animal and plant forms which do not show such marked divergences, but rather exhibit a great harmony of individuals even in detail, and the conception of species is more readily applied to these. It would certainly be foolish to give it up, since we should then lose all possibility of arriving at any sort of orientation among the enormous wealth of forms in nature. But at the same time we must not forget that these 'typical' species only appear so to our short-sighted vision—short-sighted as far as time is concerned—and that they are connected from long-past times with 'species' which lived at an earlier date, by just such transition stages as connect the Celebes species of to-day, which are all living at the same time. The world of life on the earth only presents at any given time a 'cross-section of its genealogical tree,' and according as its branches grow out vertically or horizontally we receive an impression of typical, sharply defined species or of circles or chains of forms. In the first case the evolution of new species was associated with the dying out of the horizontal branches, and the end-twigs of the branch stand beside each other now apparently isolated and sharply defined; in the other case only a portion of the ancestral species has been transmuted, and the other part continues to live alongside of the species derived from it, and perhaps repeats the process of giving off a varied race of descendants.