84 VARIATION the view of evolution thus reached is that of definite variation, with progress essentially through the subordination of individual struggle and development to species-maintaining ends. Arrest of Let us commence with the origin of the flower, which all botanists vegeta- agree in regarding as a shortened branch. The apparent explanation lion by of natural selection (from two other alternatives, one lengthened, repro- the other unshortened), although morphologically reasonable, is at ductiou. once excluded by the physiological one of inevitable shortening, since the expense of the reproductive functions necessarily checks the vegetative ones, and since we cannot speak of selection where the imaginable alternatives are physically impossible. Similarly, the shortening of the inflorescence from raceme to spike or flower- head, or its hollowing into a fig, with the corresponding reduction in the size of the flowers, receives the same explanation growth of axis and of appendages checked by reproduction superseding that current, viz., by mere convenient adaptation to fertilization by insects. The internal structure of the flower is similarly modified, as may be shown in detail in the passage from hypogyny through perigyny to epigyny (these being simply stages of the progressive arrest of the growth of the axis), in the reduction of the floral envelopes and stamens and the number of carpels and ovules, and even in the transition from perispermic to endospermic and finally exalbuminous seeds. But these are the most important of floral variations, and furnish the essentially distinguishing characters of the natural orders ; hence, when these are seen, instead of being indefinite, to be parallel and definite (i.e., determined through the continuous checking of vegetation by reproduction along what is thus a single and definite groove of progressive change), the import ance of natural selection changes wholly from that of selecting and accumulating supposed indefinite variations to that of retarding definite ones after the stage of maximum utility has been in dependently reached. The same simple conception unlocks in numerable problems of floral morphology, large and small alike : e.g. , it interprets with equal ease the inevitable development of gymnosperm into angiosperm (by continuous subordination of the reproductive carpellary leaf) and the origin of refined minor adapta tions, like the splitting fruit of the geranium or the cupped stigma of the pansy, not as achievements of natural selection from among fortuitous variations, but as naturally traceable to the checked vege tation of their respective types of leaf organ, just as in the familiar case of the pinnately-lobed outer sepals of the rose bud. The origin of floral colour as primarily an inevitable consequence of the same principle of vegetative subordination through reproductive pre ponderance was long ago pointed out by Spencer, who unfortunately, however, also accepted without scrutiny the orthodox hypothesis that this incipient floral colour would revert to the ordinary green vegetation but for the accident of its attractiveness to insects, and consequently must have owed its accumulation simply to the repeated preservation of the offspring of the accidentally brightest coloured ; whereas we now see that the arrest of vegetative greenness in the flower and the exclusive appearance of those colour ing matters associated with the imperfect vegetation of spring and autumn are a continuous process of its subordination as a vegeta tive, and its development as a reproductive, apparatus. Adapta tion to insects thus takes a quite secondary place ; similarly with the leafy arrest and floral coloration of bracts ; while a detailed examination of thorny plants practically excludes the hypothesis of mammalian selection altogether, and shows spines to arise as an expression of the diminishing vegetativeness in fact the ebbing vitality of a shoot or even of an entire species. Evergreens tend to arise in all orders among the forms of more vegetative habit, and in such forms the amount of flowering is usually diminished rela tively to the deciduous members of the same group, whilst such con stitutional vegetativeness checks the progress of floral evolution. Hence it is that evergreens, and even orders in which they largely occur, are usually less differentiated than their deciduous congeners. The apparently indefinite variations presented by domesticated plants, e.g. , the cabbage tribe, are at once classified and interpreted as so many stages along the same course (from leafy kale, through Brussels sprouts with many leaf buds, to cabbage with apical flower bud, and finally to cauliflower). The importance of this line of evi dence for the Darwinian view of variation and selection becomes still further reduced, although also broadly settled, when we note that, since the subordination of vegetative to reproductive pre ponderance gave us the origin of ordinal characters, the converse improvement of individual characters at the expense of reproduc tive ones (on which domestication usually insists and artificial selec tion depends) can only give us minor ones, for the most part varietal or specific, if so much. The Lyellian analogy for the ultimate accumulation of varietal differences into ordinal characters has thus been misleading. The preceding transition of both organography and physiology from empirical to rational carries, with it a similar transition from empirical to rational taxonomy as well : the antagon ism of vegetative and reproductive habit is seen to be also general and constitutional. Just as the liliaceous t3 pe ranges on one side towards the characteristically vegetative grass or reproductive orchid, so is it with the main variations of every natural alliance, be this order, genus, or even species. Thus the Ranunculaccsz have their grassy and their orchid-like types in meadow-rue and larkspur, yet the species of these very genera show examples of the opposite swing of variation ; again, the two species or varieties of British oak ( Q. pcdunculata and Q. scssiliflora] are thus, so to speak, the incipient grass and orchid forms of their common ancestor (Q. Robur). "What we call higher or lower species or orders are thus the leaders or the laggards ftlong one or other of these two lines of variation, the repre sentatives of some stage of the predominance on one side or other of that oscillating balance between vegetative and reproductive pro cesses which have long been known as the essential functions of organic life. The results of a similar survey of the animal world, despite the Workii greater intricacy of the problem, are scarcely less definite or com- of the prehensive, and at once lead to a deeper interpretation of the whole, law in Commencing with an ascending survey, we recognize such essential animal types of Protozoa as the rhizopods, gregarines, and infusors, not as world, the empirically selected products of spontaneous variation among indefinite possibilities, but simply as the predominatingly amrtboid, encysted, and motile phases of the primeval cell-cycle (see MORPHO LOGY), these three forms of which are fixed by the properties of proto plasm itself (see SEX, vol. xxi. pp. 720-724), each particular phase being fixed by the constitutional bias (diathesis) of its type towards anabolism or katabolism (see PHYSIOLOGY, PROTOPLASM). This rationale of variation in ultimate terms, i.e., both cellular and proto plasmic, both morphological and physiological, may be continued through all fields, e.g., embryological or pathological, for some diseases are coming to be interpreted as the progressive variation of some func tion and organ which only disturbs the general balance in its progress towards a new and higher equilibrium. 1 This conception of physio logical as well as merely structural life-history rationalizes both animal and vegetable taxonomy. Thus the greatest of all steps in morphological progress, that from the Protozoa to the Mctazoa, is not due to the selection of the more individuated and highly adapted forms, but to the union of relatively unindividuated cells into an aggregate in which each becomes diminishingly competitive and increasingly subordinated to the social whole. Passing to trifling variations, such as the internal deposition of spicules, we interpret these neither as mere accidents, nor as pure advantages for support or defence, but as arising like plant-crystals or gouty deposits as unremoved waste products, accumulating with local or constitu tional passivity. How the inevitable constitutional preponderance of anabolism or katabolism excludes spontaneous variation and subordinates external selection in the formation of types would become clear could we pass in review the animal kingdom, and note how its various alliances all range from passive to more active forms. We should compare the coral or sponge at one end of the scale with the similarly passive (and consequently also skeletal) tortoise and glyptodon at the other. We should read in their anabolic diathesis the secret of accumulating size and of inevitable extinction (nay, even unravel the apparent contradictions with re spect to the influence of the environment, as being specially strong upon the passive types, which hence are far richer in species than the corresponding less readily influenced active forms). Finally, the preponderating importance of the species-maintaining over the individual functions should again be noted in animals as in plants, since the conviction thus becomes inevitable that, in stat ing the process of evolution essentially in terms of the survival of the fittest in competition, the centre of gravity of the subject has been misplaced. The constant primary insistence upon individual competition for food, and the very subordinate recognition of the importance of sexual and social co-operation for well-being, which are characteristic of the prevalent theory (witness their proportion in the preceding summary), are traceable at once to the confusion of putting the nutritive factors "in the first place" because they precede the reproductive in time ; whereas the organism inevitably enters upon reproduction and so cedes the preponderance "the first place" to the other-regarding functions. That increase of reproductive sacrifice which at once makes the mammal and marks its essential stages of further progress (monotreme, marsupial, placental), that increase of parental care, that frequent appearance of sociality and co-operation which even in its rudest forms so surely secures the success of the species attaining it, be it mammal or bird, insect or even worm, all these phenomena of survival of the truly fittest, through love, sacrifice, and co-operation need far other promi nence than they could possibly receive on the hypothesis of the essential progress of the species through internecine struggle of its individuals at the margin of subsistence. Yet these cases are only the supreme outcomes of that continuous and definite process of variation through individual subordination and species-maintain ing preponderance which we have seen modelling even the vege table world. In short, while no more denying the existence of com petition in nature than the fact of organic progress, we deny the assumed relation of these as cause and effect.,
1 Cf. Button, General Pathology, London, 1886.