554 
K. EFGGLES GATES AND xNESTA THOMAS. 
from 22 to 27. The increase in number is brought 
about in nearly all cases by duplication of the sir. all 
idiochromosomes. Tliis is shown not only by the sizes^ but 
particularly by the behaviour of the supernumeraries during 
the reduction divisions. That is, they retain their condensed 
form during the growth stages of spermatogeuesis, forming a 
compound body with the idiochromosomes. Like the latter 
they divide as separate univalent chromosomes on the hetero- 
typic spindle, on which they usually occupy a characteristic 
position, while the other chromosomes undergo a reduction 
division. In the homotypic division they are somewhat 
erratically distributed, and this results in variations in the 
numbers of supernumeraries in different individuals of the 
next generation. By repetition of this process the number of 
supernumeraries may sometimes become as great as six. 
As regards the origin of the supernumeraries, Wilson has 
observed cases in the spermatogenesis of 22-chromosome indi- 
viduals in which both idiochromosomes passed into the same 
germ-cell in the second, which is for them the reduction, 
division. Such a sperm, if functional, would produce an 
individual having 23 chromosomes, including one super- 
numerary ; the later duplications to give as many as five or 
six supernumeraries, apparently all arise through similar 
irregular distributions of these particular chromosomes in later 
generations. These facts constituted a remarkable confirma- 
tion of the theory of genetic continuity of the chromosomes. 
Thus the supernumeraries in Metapodius owe their origin to 
the same cause as does the extra chromosome in OE . lata, 
namely, to an irregular meiotic distribution of chromosomes. 
But while plants having the extra chromosome are easily 
recognisable by their external characters, it appears that 
insects having even five or six supernumerary chromosomes 
are indistinguishable from the others even in size. This 
indicates that the supernumeraries are inactive, which is m 
harmony with the general view of Wilson that the small 
idiochromosome, or Y-element, of which they are duplicates, 
takes no active part in the determination of sex. 
