elimination and segregation of nearly allied formssuch as varieties, sub-species, and closely-related or representative species-also in a general way for their geographical association and present range, is comparatively easy, is apparently within the bounds of possi bility. Could we stop here we should be fairly contented. But, to complete the system, to carry out the principles to their ultimate conclusion, and to explain by them many facts in geographical distribution which would still remain anomalous, Mr. Darwin is equally bound to account for the formation of genera, families, orders, and even classes, by natural selection. IIe does "not doubt that the theory of descent with modification embraces all the members of the same class,” and he concedes that analogy would press the conclusion still further; while he, admits that "the more distinct the forms are, the more the arguments fall away in force." To command assent we naturally require decreasing probability to be overbalanced by an increased weight of evidence. An opponent might plausibly, and perhaps quite fairly, urge that the links in the chain of argument are weakest just where the greatest stress falls upon them.

To which Mr. Darwin's answer is, that the best parts of the testimony have been lost. He is confident that intermediate forms must have existed; that in the olden times when the genera, the families, and the orders, diverged from their parent stocks, gradations existed as fine as those which now connect closely related species with varieties. But they have passed and left no sign. The geological record, even if all displayed to view, is a book from which not only many

pages, but oven whole alternate chapters, have been lost out, or rather which were never printed from tho autographs of Nature. The record was actually made in fossil lithograpliy only at certain times and under certain conditions (i. c., at periods of slow subsidence and places of abundant sediment); and of these records all but the last volume is out of print; and of its pages only local glimpses have been obtained. Geologists, except Lyell, will object to this-some of them moderately, others with vehemence. Mr. Darwin himself admits, with a candor rarely displayed on such occasions, that he should have expected more geological evidence of transition than he finds, and that all the most eminent paleontologists maintain the immutability of species.

The general fact, however, that the fossil fauna of each period as a whole is nearly intermediato in charneter between the preceding and the succeeding faunas, is much relied on. We are brought one step nearer to the desired inference by the similar "fact, insisted on by all paleontologists, that fossils from two consecutive formations are far more closely related to each other than are the fossils of two remote formations. Pictet gives a well-known instance-the general rosemblance of the organic remains from the several stages of the chalk formation, though the species aro distinct at each stage. This fact alone, from its generality, seems to have shaken Prof. Pictet in his firm belief in the immutability of species" (p. 335). What Mr. Darwin now particularly wants to completo his inferential evidence is a proof that the same grada may be traced in later periods, say in the Tertiary,

tion

and between that period and the present; also that the later gradations are finer, so as to leave it doubtful whether the succession is one of species-believed on the one theory to be independent, on the other, derivative-or of varieties, which are confessedly derivative. The proof of the finer gradation appears to be forthcoming. Des Hayes and Lyell have concluded that many of the middle Tertiary and a large proportion of the later Tertiary mollusca are specifically identical with living species; and this is still the almost universally prevalent view. But Mr. Agassiz states that, "in every instance where he had suflicient materials, he had found that the species of the two. epochs supposed to be identical by Des Hayes and Lyell were in reality distinct, although closely allied species." Moreover, he is now satisfied, as we understand, that the same gradation is traceable not merely in each great division of the Tertiary, but in particular deposits or successive beds, each answering to a great number of years; where what have passed unques tioned as members of one species, upon closer examination of numerous specimens exhibit differences which in his opinion entitle them to be distinguished into two, three, or more species. It is plain, therefore, that whatever conclusions can be fairly drawn from the present animal and vegetable kingdoms in favor of a gradation of varieties into species, or into what may be regarded as such, the same may be extended to the Tertiary period. In both cases, what some call species others call varieties; and in the later Tertiary shells

"Proceedings of the American Academy of Arts and Sciences," vol. iv., p. 178.

this differenco in judgment affects almost half of the species!

We pass to a second difficulty in the way of Mr. Darwin's theory; to a case where we are perhaps entitled to demand of him evidence of gradation like that which connects the present with the Tertiary mollusca. Wide, very wide is the gap, anatomically and physiologically (wo do not speak of the intellectual) between the highest quadrumana and man; and com paratively recent, if ever, must the line have bifur cated. But where is there the slightest evidenco of a common progenitor? Perhaps Mr. Darwin would reply by another question: where are the fossil romains of the men who made the flint knives and arrowheads of the Somme Valley?

We have a third objection, one, fortunately, which has nothing to do with geology. We can only state it here in brief terms. The chapter on hybridism is most ingenious, able, and instructive. If sterility of crosses is a special original arrangement to prevent the confusion of species by mingling, as is generally as sumed, then, since varieties cross readily and their offspring is fertile inter se, there is a fundamental dis tinction between varieties and species. Mr. Darwin therefore labors to show that it is not a special endowment, but an incidental acquirement. He does show that the sterility of crosses is of all degrees; upon which we have only to say, Natura non facit saltum, here any more than elsewhere. But, upon his theory he is bound to show how sterility might be acquired, through natural selection or through something else. And the difficulty is, that, whereas individuals of the

vory sano blood tend to be storilo, and somewhat romoter unions diminish this tendency, and when they havo diverged into two varieties the cross-breeds between the two are more fertile than either pure stock -yet when they have diverged only one degree more the whole tendency is reversed, and the mongrel is sterile, either absolutely or relatively. IIe who explains the genesis of species through purely natural agencies should assign a natural cause for this remarkable result; and this Mr. Darwin has not done. Whether original or derived, however, this arrangement to keep apart those forms which have, or have acquired (as the case may be), a certain moderate amount of difference, looks to us as much designed for the purpose, as does a ratchet to prevent reverse motion in a wheel. If species have originated by divergence, this keeps them apart.

Here let us suggest a possibly attainable test of the theory of derivation, a kind of instance which Mr. Darwin may be fairly asked to produce-viz., an instance of two varieties, or what may be assumed as such, which have diverged enough to reverse the movement, to bring out some sterility in the crosses. The best marked human races might offer the most likely case. If mulattoes are sterile or tend to sterility, as some naturalists confidently assert, they afford Mr. Darwin a case in point. If, as others think, no such tendency is made out, the required evidence is wanting.

A fourth and the most formidable difficulty is that of the production and specialization of organs.

It is well said that all organic beings have been formed on two great laws: unity of type, and adap