398 Edmund B. ff^ilson. 



forms to the true one, and it is probable that all of these cases 

 will be found to agree in the essential phenomena with those 1 

 have determined in Lygaeus, Coenus, Nezara and the other forms. 



We may now inquire what is the relation of the idiochromo- 

 somes to the accessory chromosome. The observations suggest 

 so obvious an answer to this question that I wish to mdicate not 

 only the evidence in its favor, but more especially the difficulties 

 it has to encounter. In forms possessing an accessory chromo- 

 some the spermatozoa fall into two equal groups that differ only 

 in respect to one chromosome. The same is true of Lygaeus and 

 other forms that lack the accessory but possess the idiochromo- 

 somes, with the difference that in the former case the distinctive 

 chromosome is present in but one-half the spermatozoa, while in 

 the latter case two such distinctive chromosomes are present, one 

 of which is present in one-half, the other in the other half, of the 

 spermatozoa. It is impossible to overlook the evident analogy 

 between the two cases; and the idiochromosomes may in one sense 

 be considered as two accessory chromosomes that are never allotted 

 to the same spermatozoon since each fails to divide in the second 

 mitosis (precisely as is the case with the single accessory in other 

 Hemiptera)./ The difference between Lygaeus and Ccenus in the 

 size-ratio of the idiochromosomes obviously suggests the view 

 that the single accessory of other forms may have arisen by the 

 disappearance of one of the idiochromosomes; and in Lygrcus the 

 smaller one is already so minute as distinctly to suggest a vestigial 

 structure. We might accordingly assume that in a more primitive 

 type the two idiochromosomes were of equal size (as in Nezara), 

 undergoing synapsis and subsequent reduction in the same way 

 as the other chromosomes; that Coenus and Lygaeus represent 

 successive stages in the reduction of one of these chromosomes; 

 and that by the final disappearance of the smaller one in such 

 forms as Anasa or Pyrrochoris a single accessory chromosome 

 remains. 



This hypothesis at first sight seems to give a clear and intelli- 

 gible view of the origin of the accessory chromosome, and to recon- 

 cile the remarkable mode of spermatogenesis occurring in the 

 insects with forms in which no accessory seems to appear. But 

 further reflection shows that it has to encounter a formidable if 

 not insuperable difficulty in the fact that in some of the forms 

 possessing an accessory chromosome the number of spermato- 



