106 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 
the same pole. Using this as a basis, he was able to find satisfactory 
explanation of the variation in number, and one which served to sup- 
port the theory of the individuality of the chromosomes. Stevens 
(?12b) found in Diabrotica supernumerary chromosomes varying in 
number from 1 to 5, and believed that they had their origin in trans- 
verse and longitudinal divisions of the X-chromosome, which normally 
divides only longitudinally. These anomalies can therefore be 
explained on the basis of some unusual method of distribution of the 
chromosomes in mitosis; the fact that such extra chromosomes persist 
in all the cells of the animal in which they are found is a striking bit 
of evidence in favor of the belief that they maintain their individuality. 
Della Valle (’09, ’12) has attacked the theory of individuality, 
declaring that the chromosomes are temporary and variable structures, 
which form in the prophase and dissolve in the telophase. He thinks 
their number is the quotient of the quantity of chromatin divided by 
the average size of the chromosomes. The quantity of chromatin 
is said to vary with conditions of nutrition, and the number of chromo- 
somes with variations in external conditions. He made counts of 
chromosomes from cells of the peritoneum of salamander larvae and 
obtained numbers varying from 19 to 27. Montgomery (’10) points 
to the following grounds for doubting the accuracy of Della Valle’s 
conclusions: — “1. The chromosomes counted are long, sinuous 
ribbons, that overlap and interlace, the most difficult kind to count 
with accuracy. 2. He included in the counts some cells in prophases, 
where one cannot be certain that all the chromosomes have fully 
separated. 3. The total number of the chromosomes is so large, 
about 24, that the chance of error in enumeration is great. It is but 
fair to conclude that while his technique was excellent, his choice of 
material was bad, consequently a degree of scepticism might well be 
maintained toward his results.” Della Valle in his latest paper (’12) 
argues that the chromosomes are variable structures, because he has 
been able to find transition stages between mitotic and amitotic 
methods of cell-division in the erythrocytes of young salamanders. 
It is a well-known fact that amitosis frequently accompanies degenera- 
tion, and the figures of Della Valle present strong indications of being 
those of degenerating cells. It is precisely in degenerating cells that 
one would look for inconstancy in the behavior of the chromosomes. 
2. Constancy in size and shape.— It will be convenient to consider 
the subjects of size and shape together. As to shape, we may dis- 
tinguish spheres, rods, and V-shaped elements. Spheres are invariably 
small and may be regarded as short rods. It will be convenient to 
