INADEQUACY OF NATURAL SELECTION, ETC. 675 



lateral controversy, intending to resume it presently, when the 

 original issues have been dealt with.] 



No one will suspect me of thinking that the inconceivability 

 of the negation is not a valid criterion, since, in " The Universal 

 Postulate," published in the Westminster Review in 1852 and 

 afterwards in The Principles of Psychology^ I contended that it is 

 the ultimate test of truth. But then in every case there has to 

 be determined the question — Is the negation inconceivable ; and 

 in assuming that it is so in the case named, lies the fallacy of the 

 above-quoted passage. The three separate ways in which I dealt 

 with this position of Professor Weismann are as follows : — 



If we admit the assumption that the form of the soldier-ant 

 has been developed since the establishment of the organized ant- 

 community in which it exists, Professor Weismann's assertion 

 that no other process than that which he alleges is conceivable, 

 is true. But I pointed out that this assumption is inadmissible ; 

 and that no valid conclusion respecting the genesis of the soldier- 

 ant can be drawn without postulating either the ascertained, or 

 the probable, structure of those pre-social, or semi-social, ants 

 from which the oro^anized social ants have descended. I went on 

 to contend that the pre-social type must have been a conquering 

 type, and that therefore in all probability the soldier-ants repre- 

 sent most nearly the structures of those ancestral ants which ex- 

 isted when the society had perfect males and females and could 

 transmit acquired characters, while the other members of the 

 existing communities are degraded forms of the type. 



No reply. 



A further argument I used was that where there exist differ- 

 ent castes among the neuter-ants, as those seen in the soldiers and 

 workers of the Driver ants of West Africa, " they graduate in- 

 sensibly into each other " alike in their sizes and in their struc- 

 tures ; and that Professor Weismann's hypothesis implies a spe- 

 cial set of " determinants " for each intermediate form. Or if he 

 should say that the intermediate forms result from mixtures of 

 the determinants of the two extreme forms, there still remains 

 the further difficulty that natural selection has maintained, for 

 innumerable generations, these intermediate forms which are in- 

 jurious deviations from the useful extreme forms. 



No reply. 



One further reason — fatal it seems to me — was urged in bar 

 of his interpretation. No physical cause has been, or can be, 

 assigned, why in the germ-plasm of any particular queen-ant, the 

 "determinants" initiating these various co-operative organs, all 

 simultaneously vary in fitting ways and degrees, and still less 

 why there occur such co-ordinated variations generation after 

 generation, until by their accumulated results these efficient co- 



