MULTIPLE CHROMOSOMES 559 



Upon turning to Mermiria bivittata we encounter these con- 

 ditions: a spermatogonia! complex of twenty-two free elements 

 among which are two atelomitic Vs. If it were assumed that 

 the latter are always multiples, then the male diploid complex 

 would be twenty-four. There are two arguments against this 

 conclusion however. First there is the circumstance that the 

 genus — a well marked one — has species with the normal twenty- 

 three free chromosomes. The value of such evidence has already 

 been considered. A second piece of evidence is the actual 

 morphological structure of the Vs. One of them shows by all its 

 history that it is directly comparable to the similar structure in 

 Hesperotettix — it is a multiple composed of the accessory chromo- 

 some and a rod-shaped euchromosome. The other V is found 

 in the first spermatocyte attached to this multiple element. 

 Unless this combination be of greater valence than a hexad, the 

 second V can be only a dyad. Strong evidence that this V is 

 indeed only a simple dyad appears from a study of certain in- 

 dividuals which I have called the 'green' bivittata. In these 

 specimens the multiple chromosome is almost the shape of the 

 one found in Hesperotettix, the fiber attaching near the free 

 end instead of toward the middle (figs. 58, 59, 60, pi. 7). The 

 valence of the part is very clearly indicated in this way. Owing 

 to mistaken enumeration in the spermatogonium, because of in- 

 sufficient material, I regarded the multiple as a decad instead 

 of a hexad in my earlier work. In this opinion I believed my- 

 self confirmed by the persistent type of chromosomes in the 

 family. It was only when new material revealed to Miss Caroth- 

 ers that, in a full complex of twenty-three chromosomes, there 

 might be numerous double Vs, and that one such an element 

 might be united to a rod, producing a J-shaped tetrad, that I 

 found it possible to account for the form of the Mermiria mul- 

 tiple. With plenty of favorable material I was also able to de- 

 termine accurately the diploid number of chromosomes in the 

 male to be twenty-two. Having all these facts it was not diffi- 

 cult to solve the problem. 



It is apparent, therefore, that form alone is not sufficient to 

 estabhsh the nature of a chromosome. The double V may or 



