NA TURE 



[June i i, 1896 



principle as directed to one definite character alone. 

 Individual selection must, however, step in from time to 

 time, to check the other process when this latter exceeds 

 the demands of utility. This is how so many different 

 modifications can be set going at the same time. It is 

 to be observed that qualitative no less than quantitative 

 modification must be under the influence of the same 

 principle. Selection must affect the " biophors " as much 

 as the " determinants " which they compose. A (Quanti- 

 tative alteration of constituent biophors appears to us as 

 a qualitative modification of the corresponding deter- 

 minants, and this enables us to understand how " units 

 of variation" may play their part by either simultaneous 

 or independent modification, as on the under side of a 

 butterfly's hind wing. The phenomena of mimicry can- 

 not be accounted for by accidental variations only, but 

 must depend on variation definitely directed by utility. 

 It is to be observed that the determinants and groups of 

 determinants here postulated have nothing to do with 

 Bonnet's preformation theory. The determination of the 

 character of a developing ovum by its own constitution, 

 instead of by the action of external forces, must be 

 admitted by all those who do not, like O.. Hertwig, 

 confound the conditions of development with its causes. 



The assumption seems inevitable that every heritable 

 and independent variation in the soma depends on the 

 variation of a definite part of the germ ; not, as Spencer 

 thinks, on that of all the units of the germ. The latter 

 theory leads to needless complication. It is no valid 

 objection to the determinant theory that it deals with 

 invisible elements. The same is true of the chemical 

 assumption of atoms and molecules. The theory justifies 

 itself as such in that it can be used as a formula, to 

 express, for example, the conditions of di- and poly-mor- 

 phism. The " Hotspurs of biology " forget that all our 

 knowledge is provisional. The assumption of biophors 

 and determinants is parallel to such conceptions as 

 "force," "atoms," and "ether-waves" in the domain of 

 physics. 



Epigenesis does not, as has been held, allow a simpler 

 structure for the germ than the counter-theory, and 

 germinal selection explains entirely the direction of 

 variation by utility. It also disposes of the objection 

 that selection cannot cause the variations with which it 

 works. Given the numerical fluctuation of the units, 

 selection will do the rest. Hence both the exactitude and 

 simultaneity of useful variations, a simultaneity which 

 may affect like parts, as in the development of eyes and 

 limbs ; or unlike, as in the production of complex 

 mimetic patterns. The principle of selection reaches 

 just so far as utility reaches, and translates, as we have 

 seen, quantitative into qualitative modifications. Utility 

 undoubtedly goes hand in hand with modification, but 

 the dwindling of disused parts shows that the inheritance 

 of characters actively acquired does not cover the whole 

 ground, as the selection theory docs ; for how can the 

 disuse of an organ affect the germ ? The Lamarckians 

 are right in pronouncing individual selection inadequate to 

 account for the facts, and also in denying that panmixia 

 could bring about the entire disappearance of a disused 

 organ ; they err, however, in attributing the results of 

 Roux's " struggle of parts " to heredity. 

 Thus then the three stages of selection are (i) personal 

 NO. 1389, VOL. 54] 



or individual (that of Darwin and Wallace) ; (2) histo- 

 logical, as maintained by Roux ; and (3) germinal, as 

 pointed out by Weismann. There is indeed another 

 stage, that namely between races or stocks. Here 

 individuals play the same part as organs in individual 

 selection ; the analogy, however, is not in all respects 

 complete. 



Everything in nature, says the author in conclusion, is 

 purposeful ; and this fact can only be accounted for by 

 the theory of selection. What is obscure in the process 

 is so from the imperfection of our methods, not of the 

 principle. All kinds of knowledge ultimately resolve 

 themselves into the hypothetical and unknowable. But 

 doubt is the parent of progress ; the veil is raised little 

 by little ; and what still remains dark in the explanation 

 points, like the magic wand in the hands of the water- 

 finder, to the hidden springs of truth, ready to yield 

 themselves up to the persevering seeker. 



The preface contains a forcible and dignified vindica- 

 tion of the use of hypothesis in scientific investigation ; 

 both generally and with special reference to the author's 

 own theory of heredity. Appendices are added, in which 

 several points raised by the paper receive more de- 

 tailed treatment. The controversies that have centred 

 round Prof Weismann's former works are not likely to 

 be hushed by the present treatise. We may safely ven- 

 ture to predict that the olive-branch held out to the neo- 

 Lamarckians (p. 59) will not be accepted, though the 

 admissions as to the inadequacy of individual selection 

 will be welcomed by many as evidence of a change of 

 view. The absence of all reference to amphimixis no 

 doubt simplifies the argument greatly ; it will, however, 

 be probably used in some quarters to point the moral 

 of the author's inconsistency. But, whatever the amount 

 of acceptance which this latest development of the selec- 

 tion-hypothesis is destined to achieve, there can be no 

 question that the present «'ill rank among the most 

 interesting and suggesti\e of the Freiburg professor's 

 contributions to biological theory. F. A. DiXEY. 



[Addendum. — Since the above notice was written, an 

 Enghsh translation of Prof Weismann's treatise has 

 been issued by the Open Court Publishing Company, 

 Chicago. — F. .\. D.] 



RIVERSIDE LETTERS. 

 Riverside Letters; a continuation of " Letters to Marco." 

 By George D. Leslie, R.A., author of "Our River." 

 Pp. xvi+251. (London: Macmillan and Co., Ltd., 

 1896.) 



ALTHOUGH in his preface Mr. Leslie is careful to 

 state that he cannot assert in the case of these 

 Letters, as he did in the previous volume of his " Letters 

 to Marco," that they were written wholly without view to 

 publication, yet is there little or no change in subject- 

 matter or in style. They are, like the former collection, 

 genuine letters sent to his friend Mr. MarkSjoR.A., and 

 the topics on which he writes are of mutual interest to 

 the two friends who both, as he says, " love nature for 

 her own sake, untrammelled by the ijrcpossessions that 

 not unfrequently accompany that lo\e among the votaries 

 of science or sport," and in publishing them he doubtless 



