And here, curiously enough, Mr. Wallace comes forward with an additional suggestion under this head, which, however, he regards as an alternative hypothesis.' This additional suggestion is that there may be a connection between sexual compatibility and external colouring, such that any variation in the latter may be accompanied by a correlated variation in the former, leading to sterility with the unmodified, or differently modified, type of colour. So that when colour is changed, for protective or other purposes, by natural selection, an indirect or incidental change may be wrought in the reproductive system, such that the modified individuals are fertile only amongst themselves. Now, for reasons mentioned below, this hypothesis does not recommend itself to my mind as at all likely to be 'the true solution of the problem of the sterility of hybrids.' 12 It may possibly be a true explanation of some cases; but to regard it as probably the true explanation of all appears to me absurd. However, the point with which I am concerned is not as to the validity of this suggestion in itself, but merely with the astonishing misapprehension of my theory which leads Mr. Wallace to regard the suggestion as an 'alternative hypothesis.' Far from being in any way opposed to my theory, his suggestion runs directly on its lines; he merely seeks to add another to the many causes of the indirect class on which I myself rely. As clearly explained in my paper, it makes no difference

induced by any cause acting directly on a specific type, in that measure is the indirect action of natural selection superseded by the independent principles of what Mr. Spencer calls direct equilibration.' Of course these principles may co-operate with that of natural selection; but none the less they are quite distinct. In short, my objection to natural selection on the score of free intercrossing only applies to cases of accidental variations,' relatively few in number; where 'collective variations' are supplied to natural selection by other causes the objection, of course, is satisfied.

12 1. Many species which are mutually sterile differ very little in colour. 2. Most species which are mutually fertile differ considerably in colour.

3. Our domestic varieties, both of plants and animals, are largely reared more or less expressly for the purpose of obtaining extreme differences of colour; yet nearly all the resulting varieties are notoriously fertile.

4. In the case of natural species, it often happens that a great difference in respect of fertility occurs according to which has acted as the male and which as the female ; yet in both these crosses the colour of each species is, of course, the same.

5. Similar remarks apply to the case of dimorphic and trimorphic plants. 6. In the case of fertile hybrids, it may be regarded as a general rule that the more nearly they resemble either parent form in colour, the greater is their sterility. 7. Even apart from all these opposing facts, on merely antecedent grounds it is highly improbable that, to use Mr. Wallace's own words, 'so widespread a phenomenon as that of some degree of sterility between species' should be due to any merely accidental correlation between external colour and reproductive function, extending throughout the whole range of organic nature.

8. The suggestion supposes natural selection to be the cause of the colourchange. If so, in most cases the unchanged individuals must die out. How, then, does it come to pass that there continues to be an unchanged type of colour with which the changed type is now found to be infertile ?

to the theory of physiological selection what the particular causes may be which induce the sexual change in any particular case; and I expressly insist that natural selection may well be regarded as one among the sundry other causes of the indirect class which do induce this variation. These causes, both direct and indirect, I believe to be numerous, varied, sometimes complex, generally subtle, and therefore often obscure. But to take on the present occasion a merely bird's-eye view of the matter, when we consider the extraordinary sensitiveness of the reproductive system to slight changes in the conditions of life, we cannot fail to conclude that in the long lifehistories of species-furnishing great vicissitudes to large populations spread over wide areas-many and diverse causes must often be encountered, leading to collective variations on the part of this system, and that these variations must sometimes be of the kind which my theory requires.

6

And here it may be remarked that it was such cases as these of collective variability (or where the physiological variation required by my theory affects a number of individuals simultaneously) which I had in view while writing that such a variation must always be preserved whenever it occurs,' and this with even more certainty than are the useful variations which furnish material to the working of natural selection.' Mr. Wallace calls this a most extraordinary statement,' and no doubt it must have appeared so to him, seeing that he only waited to consider the case of physiological variations arising fortuitously-where, as he needlessly argues a self-evident fact, there must be many chances against even the first mating of the physiological complements. But of course the extraordinary' nature of my statement altogether disappears when its meaning is understood; for it is surely sufficiently evident that if the variation does not merely occur sporadically in an individual here and there, but affects simultaneously a large number of the inhabitants of a district, it is more certain to be perpetuated than any accidental' (even though useful) variation could be; seeing, on the one hand, that it cannot be obliterated by intercrossing, and, on the other hand, that the fitness' of the individuals affected is guaranteed by the fact of their having reached the breeding age. This latter point is important, because Mr. Wallace accuses me of having lost sight of the consideration that my physiological variations must conform to the law of natural selection. He says, 'Mr. Romanes' argument almost everywhere tacitly assumes that his physiological variations are the fittest, and that they always survive! With such an assumption it would not be difficult to prove any theory of the origin of species.' Now, I hold that this tacit assumption is justified by the consideration that if these physiological varieties ever occur at all, ex hypothesi they must have so far passed muster with respect

to general fitness as to be allowed to propagate their kind. It was for the sake of emphasising this feature of my theory that I gave to the latter the alternative title of Segregation of the Fit.'

If I have succeeded in making myself intelligible, it will have been seen that Mr. Wallace's objection to my theory admits of a twofold answer. In the first place, it is impossible for him to 'show' that the origin of a species is any more frequent than it ought to be, even upon the assumption which he has imputed to me-namely, that such origin is always due to the chance mating of more or less extremely rare varieties. And, in the next place, this assumption on his part is wholly gratuitous-or rather, I should say, directly opposed both to my own statements and to all the probabilities of the case.

From which it is easy to perceive the inevitable inference, or, if not, by stating it I will furnish a cue to future critics. The real difficulty against my theory is precisely the opposite of that which Mr. Wallace has advanced. This real difficulty is that the differentiation of specific types has not been of nearly so frequent occurrence as upon the theory of physiological selection we should have antecedently expected. Looking to the great sensitiveness of the reproductive system, to the many and the varied causes which affect it, to the frequency with which these causes must have been encountered under Nature, to the fact that whenever a collective variation occurs of the kind which induces physiological selection it must almost certainly leave a new species to record the fact-looking to all these things, the only real difficulty is to explain why, if physiological selection has ever acted at all, it should only have done so at such comparatively rare intervals, and therefore have produced such a comparatively small measure of result. If my critics had adopted this line of argument, I should have experienced more difficulty in meeting them. But, as the case now stands, it seems enough to remark that I do not know of any way in which an adverse criticism admits of being more thoroughly exploded, than by showing that the difficulty which it undertakes to present is the precise opposite of the one with which an author is in his own mind, and at that very time, contending.

Seeing how remarkable has been the misunderstanding displayed by such competent readers as Mr. Wallace and Mr. Seebohm-a misunderstanding on which they both found their only objection to my theory-I should have been compelled to suppose that my paper failed in clearness of expression, were it not that (as above shown) they have disregarded the literal construction of my sentences. Nevertheless, it is probable enough that I may not have sufficiently guarded against a misunderstanding which it never occurred to me that any one was likely to make. For I supposed that all readers would have perceived at least that the main feature of the theory is

what my paper states it to be-namely, that sterility with parent forms is one of the conditions, and not always one of the results, of specific differentiation. But, if so, is it not evident that all causes which induce sterility with parent forms are comprised by the theory, whether these causes happen to affect a few individuals sporadically, a number of individuals simultaneously, or even the majority of an entire species?

GEORGE J. ROMANES.

THOMAS DEKKER.

Of all English poets, if not of all poets on record, Dekker is perhaps the most difficult to classify. The grace and delicacy, the sweetness and spontaneity of his genius are not more obvious and undeniable than the many defects which impair and the crowning deficiency which degrades it. As long, but so long only, as a man retains some due degree of self-respect and respect for the art he serves or the business he follows, it matters less for his fame in the future than for his prosperity in the present whether he retains or discards any vestige of respect for any other obligation in the world. François Villon, compared with whom all other reckless and disreputable men of genius seem patterns of austere decency and elevated regularity of life, was as conscientious and self-respectful an artist as a Virgil or a Tennyson: he is not a great poet only, but one of the most blameless, the most perfect, the most faultless among his fellows in the first class of writers for all time. If not in that class, yet high in the class immediately beneath it, the world would long since have agreed to enrol the name of Thomas Dekker, had he not wanted that one gift which next to genius is the most indispensable for all aspirants to a station among the masters of creative literature. For he was by nature at once a singer and a maker: he had the gift of native music and the birthright of inborn invention. His song was often sweet as honey; his fancy sometimes as rich and subtle, his imagination as delicate and strong, as that of the very greatest among dramatists or poets. For gentle grace of inspiration and vivid force of realism he is eclipsed at his very best by Shakespeare's self alone. No such combination or alternation of such admirable powers is discernible in any of his otherwise more splendid or sublime compeers. And in one gift, the divine gift of tenderness, he comes nearer to Shakespeare and stands higher above others than in any other quality of kindred genius.

And with all these gifts, if the vulgar verdict of his own day and of later days be not less valid than vulgar, he was a failure. There is a pathetic undertone of patience and resignation not unqualified by manly though submissive regret, which recurs now and then, or seems to recur, in the personal accent of his subdued and dignified appeal to the casual reader, suggestive of a sense that the higher triumphs of art, the brighter prosperities of achievement, were not VOL. XXI.-No. 119.

G