154 Edmund B. Wilson 



ing a very characteristic bipartite body; but in a good many cases 

 they are separate (Fig. 6, c, d, Photo 26). A large and very dis- 

 tinct plasmosome is also present. 



Such a group of 22 chromosomes may be regarded as the type 

 of which all the other forms may be regarded as variants, and 

 probably as derivatives. In forms having more than 22 chromo- 

 somes the increase in number is due to the presence of from one to 

 six supernumeraries. These vary in number and size in different 

 individuals, but both are constant in a given individual. Their 

 maximal size is equal to that of the small idiochromosome (in 

 which case they are indistinguishable from the latter); such forms 

 will be called "large supernumaries. " Their minimal size, 

 ("small supernumeraries") is about the same as that of the m- 

 chromosomes; but from the latter they are always distinguishable, 

 in the male, by a quite different behavior in the maturation pro- 

 cess. When a single supernumerary is present it may be either 

 large or small, its size being (with slight variation) constant in the 

 individual. When more than one is present all may be of the 

 same size (the most usual condition) or they may be of different 

 sizes, the relation being again an individual constant. Whatever 

 their number or size their behavior is essentially the same as that 

 of the idiochromosomes. In the growth-period they have a con- 

 densed form and are typically united with the idiochromosomes 

 to form a compound chromosome-nucleolus, the components of 

 which are often distinctly recognizable and vary in number with 

 the number of the supernumeraries. In the first division they 

 divide as separate univalents, and this division accordingly shows 

 as many chromosomes above 12 as there are supernumeraries 

 i.e., if the spermatogonial number be 22 + n, the number in the 

 first division is typically 12 + n. Their typical position in this 

 division is, like that of the idiochromosomes, outside the ring of 

 larger bivaients, though there are many exceptions. In the sec- 

 ond division they are, as a rule, again associated with the idio- 

 chromosomes to form a compound element, though not infre- 

 quently one or more of them may be free from the others. 



A definite correlation thus appears in each individual between 

 the number and relative sizes of the chromosomes seen in the 



