522 Edmund B. Wilson. 



2. RELATION OF THE CHROMOSOME-NUCLEOLUS TO THE SPER- 

 MATOGONIAL CHROMOSOMES. 



In view of the foregoing conclusion it will readily be admitted 

 that a derivation of the chromosome-nucleolus from the two 

 spermatogonial microchromosomes is a priori highly improbable; 

 and in point of fact, all the actual observations not only of myself, 

 but also, I believe, of Paulmier and Montgomery, are opposed 

 to such a conclusion. 



This question has been complicated in a most unfortunate way 

 by errors in counting the spermatogonial chromosomes. It was 

 natural that the earlier observers should have expected to find an 

 even number of chromosomes in the spermatogonial divisions; 

 and the number is in point of fact an even one in all the forms 

 that possess the idiochromosomes, as I have shown in the first 

 of these studies. Regarding the forms that possess an accessory 

 or heterotropic chromosome the existing accounts are, however, 

 in conflict in giving sometimes an even number (Anasa, t. Paul- 

 mier and Montgomery, Syromastes, /. Gross, Alydus pilosulus, t. 

 Montgomery), and sometimes an odd one (Protenor, Harmostes, 

 (Edancala, Alydus eurinus, t. Montgomery). A similar difference 

 occurs in the existing accounts of the spermatogonia in Orthoptera, 

 some of which are described as showing an even number and some 

 an odd. This contradiction has enormously increased the com- 

 plication of the subject; for it has necessarily involved the view 

 that in cases showing an even number the heterotropic chromo- 

 some is a bivalent body, formed by the synapsis of two of the 

 spermatogonial chromosomes; and this, in turn, very naturally 

 led Montgomery ('04, '05, etc.) to the further conclusion that in 

 cases showing an odd number one of the chromosomes (presum- 

 ably the "accessory") is already bivalent in the spermatogonia. 



I myself had at first no doubt of the correctness of both these 

 interpretations, and my faith was not shaken even after the dis- 

 covery that the number is 13 in Alydus pilosulus (Fig. I, A), 

 15 in Archimerus (Fig. 3, /), and 21 in "Chariesterus." 1 When, 

 however, demonstrative evidence was obtained that even in Anasa 

 in opposition to the concordant results of Paulmier and Mont- 

 gomery on Anasa and those of Gross on the related form Syro- 



J The indentification of this form (from Paulmier's material) is doubtful. 



