NA TURE 



49 



D' 



ROMANES ON WEISMAXN. 



An Examinatio7t of Weismannism. By G. J. Romanes, 



M.A., LL.D., F.R.S. (London : Longmans, Green. 



and Co., 1893.) 



R. ROMANES is a most competent hurler of hard 

 words, and in this volume is concerned at least as 

 ' much to convince the reader that Weismann is an un- 

 certain guide as to be to him himself a certain guide. In 

 the preface he states his intention to publish his criticisms 

 "in separate form and in comparatively small editions, 

 so that further chapters may be added with as much 

 celerity as Prof. Weismann may hereafter produce his 

 successive works." In the text, writing of the relations 

 between the views of Galton and Weismann, he talks of 

 those immense reaches of deductive speculation, which, 

 in his opinion, merely " disfigure the republication of stirp 

 under the name of germ-plasm " (1) The mention of 

 certain occurrences which are believed in by Dr. Romanes, 

 but the admission of which he considers illogical on the 

 part of Weismann, "seemed attributable to mere care- 

 lessness on the part of their author." Another consider- 

 ation is "made by Weismann for the sole purpose of 

 saving as much as he can of his previous theory of 

 variation." Another is " an obvious equivoque." The 

 mechanism of heredity is planned out (in Weismann's 

 latest volume) " in such minuteness of detail and assur- 

 ance of accuracy that one is reminded of that which is 

 given by Dante of the topography of the Inferno." 



Of the actual criticism the last chapter and the two 

 appendices alone require special treatment, as they alone 

 were written after the publication of " Amphimixis" and. 

 " The Germ-plasm." 



It does not seem useful to insist with Dr. Romanes 

 that the continuity of the germ-plasm is the inverse of 

 the basis of the theory of pangenesis. The most im- 

 portant part of the continuity theory has no parallel in 

 Darwin's provisional hypothesis. It is the attractive 

 suggestion of a material basis of heredity which can be 

 identified with structures visible under the microscope ; 

 which can be seen, in some cases, to separate immedi- 

 ately from the fertilised ovum to form the foundation of 

 the germ-cells of the new individual, or, in other cases, 

 to move along "germ-tracts " to the foundation of the 

 germ-cells of the new individual. What is directly com- 

 parable in the two theories is the picture each gives of the 

 phenomena of inheritance viewed in pangenesis as a roll- 

 ing up of gemmules from an existing body to form germ- 

 cells ; in the germ-plasm as the unrolling of germ- 

 plasm to form a developing body. In this, as Dr. 

 Romanes points out, the one theory is the inverse of 

 the other, and very naturally similar groups of facts 

 may appear in the one as stages of rolling up, in the 

 other as stages of disintegration. But here again Weis- 

 mann, aided no doubt by the vast advance in micro- 

 scopical science, constantly is more in touch with 

 observed facts of microscopical detail than was Darwin. 



Dr. Romanes uses a good deal of space for a minute 

 and interesting comparison of Weismann's germ-plasm 

 with the "stirp" of Galton. He urges that natural 

 NO. 1255, VOL. 49] 



selection, so potent in the organic world, probably does 

 not cease in the separate parts of a body during develop- 

 ment, and therefore supports Galton's view of a competi- 

 tion among many gemmules of the same order as to which 

 shall actually cause development. But natural selection 

 is not a force : it is merely an aspect of certain occur- 

 rences, and while there may be many (as Galton thinks) 

 or few (as Weismann thinks) units of germ-plasm each 

 capable of causing development, and only one of which 

 does cause development, the aspect of the occurrences 

 on which Weismann wishes to direct attention is that 

 the process of development goes on by an orderly dis- 

 integration of the germ-plasm through various stages of 

 units, and that the order is determined by the " historic 

 architecture " of the germ-plasm. This "historic archi- 

 tecture" is the material representation (on Weismann's 

 theory) of the observed fact that ontogeny does to some 

 extent repeat phylogeny. A continual struggle among 

 innumerable units would account for too much variation, 

 and would leave unrepresented the habitual fixedness of 

 heredity. 



In his criticism of Weismann's view of evolution Dr. 

 Romanes first states how recent further investigations 

 (those of Maupas and others) into the conjugation of Pro- 

 tozoa have led to an identification of conj ugation with 

 sexual reproduction, so far as they both result in a mingling 

 of germ-plasm, but he quarrels with Weismann for not 

 abandoning the potential immortality of the Protozoa. 

 But whether Protozoa conjugate or not, on the broad 

 average they divide by fission. That means that 

 Protozoa alive to-day have come down in a con- 

 tinuous chain of cell-life from primeval Protozoa, unless 

 indeed there have been continual re-creations of Protozoa. 

 Even if it were proved that spore-formation invariably 

 interrupted at long or short intervals chains of simple 

 fission, still practical immortality may be held by regard- 

 ing spore-formation as merely multiple fission. 



In a more important criticism Dr. Romanes seems to 

 me to misinterpret Weismann's position. When the 

 continuity of germ-plasm first presented itself to Weis- 

 mann's mind, and brought with it the idea that the somas 

 of each generation were mere pendants of the chain of 

 germ-plasm, it became difficult to see how the impression 

 of outside nature on the soma could be impressed in turn 

 on the germ-plasm. This led to an examination of a 

 new kind into acquired characters, and the result of that 

 examination satisfied Weismann and many others that 

 there was no sufficient reason for supposing that charac- 

 ters acquired by the individual were transmitted to the 

 progeny. Of course this is still a matter of argument, and 

 as Dr. Romanes in this book refers to a full treatment of 

 the question, the publication of which has been delayed 

 by his regrettable illness, it may well be that he will ad- 

 duce fresh and important considerations. But the fact 

 remains that Weismann, driven back from acquired 

 characters as a cause of phylogenetic variation, 

 came to regard the mingling of germ-characters in 

 amphimixis (traceable back to the direct influence 

 of the environment upon organisms antecedent to 

 amphimixis) as the source of all variation. The 

 germ-plasm lived as a parasite within the soma, 

 and was related to it only by the fact that it got food 

 from the soma. In the more developed doctrine 



D 



