390 NATURAL SCIENCE. Nov.. 



Weismann has expanded those hints in previous writings, that 

 Dr. Romanes called illogical admissions, and in his most recent 

 book, by allowing that unequal nutrition affecting the germ-plasm 

 may cause variation of the germ-plasm, he "admits" the action 

 of the soma upon the germ-plasm to an extent which will probably 

 be found to cover the residuum of facts now that they have been 

 sifted by controversy. For this source of variation, in addition 

 to amphimixis, and to the inherited results of the original action 

 of the environment upon primitive organisms, brings Weismann's 

 theory of evolution so nearly into line with what Dr. Romanes 

 agrees to, that at first he proposed to cancel a large part of this 

 book — but he has left it, showing how very prettily he would have 

 fought had there been anything to fight against ! But he still insists, 

 against Weismann, that Weismann's theory requires the absolute 

 stability of the germ-plasm. However, his examination of this 

 question cannot be complete without the promised sections upon 

 acquired characters, and everyone will hope that Dr. Romanes' health 

 will allow him to publish the new volume soon. 



In other matters, however, there is quite enough left tor 

 controversy. 



Dr. Romanes A^ery acutely divides Weismannism into the theory 

 of heredity which deals chiefl)' with the germ-plasm, and the theory 

 of evolution which deals chiefly with variation. 



He objects to the vast elaboration of the germ-plasm with its 

 molecules, biophors, determinants, ids, idants, and so forth, chiefly 

 because he regards it as a work of " artistic imagination " rather than 

 of " scientific generalisation." So far as concerns "determinants" — 

 the particles of germ-plasm which correspond to groups of cells that 

 vary similarly and simultaneously — he admits the great value of the 

 conception, and he accepts the " ids " as a logical consequence of 

 "determinants"; but he sees no reason whatever for identifying 

 idants with chromosomes. In this, Dr. Romanes makes the same 

 error as he did in the matter of the continuity of the germ-plasm. In 

 both cases, Weismann started with the observed fact ; the visible 

 changes in the Hydrozoa led him to the continuity, and the actual 

 divisions and marshallings of idants in sexual cells led him to the 

 particulate theor)'^ of germ-plasm. And it is because his theories 

 start from observed facts of this kind that they appeal so strongly to 

 " laboratory naturalists." 



He gives a very interesting comparison between Galton's 

 " Stirp," and Weismann's germ-plasm, pointing out how the recent 

 extensions (Romanes would call them " revolutions ") in the latter 

 theory bring it more into harmony with the former. Both allow that 

 the soma arises from the germinal cells, and that in each ontogeny 

 only a part of the germinal material is used up — ^the greater part being 

 handed on. According to Galton, occasionally, but rarely, there are 

 contributions from the soma to the stirp, and so acquired characters 

 may be inherited occasionally : in Weismann there are never such 

 centripetal contributions, and so acquired characters, as such, are 

 never inherited. In the mechanism of ontogeny, Weismann insists 

 on a peaceful disintegration of the architecture of the germ-plasm ; 

 Galton on a constant struggle between competing carriers of heredity. 

 Weismann, in his dealing with " xenia " — the influence in plants of 

 fertilisation, not on the fertilised cell, but on the tissues of the female, 

 allows the possibility of the direct action of germ-plasm upon 

 the somatic tissues, even though those tissues may belong to 



