Mutations and Evolution. 
85 
small chromosome fragment, others may have lost a fragment; and 
this renders probable another type of germinal change, in which the 
redistribution of fragments of chromatin may take place. It was 
formerly pointed out (Gates and Thomas 1914) that all these 
“irregularities” will become the basis of germinal changes if they 
are perpetuated by mitosis. Some of the differences between 15- 
chromosome types may yet be explained on this basis. 
This brings us to another subject closely connected with the 
extra chromosome, but which concerns also the question of 
chromosomes constancy in (Enothera . Hance (1918), in the paper 
already referred to, has studied variations in the number of 
somatic chromosomes in CE. scintillans. The writer 1 first pointed 
out such variations in the somatic cells of CE . lata , the number 
ranging from 12 to 20 or 21. Hance finds that the number of 
separate bodies in metaphase ranges from 15 to 21, the higher 
numbers being produced by transverse segmentation of certain of 
the chromosomes. This matter is of minor importance and will 
be referred to later. A fact of more interest is that more or less 
constant differences are found between the lengths of the different 
chromosome pairs in the group of 15. The pairs of chromosomes 
form a graded series, each somatic pair differing in length from 
the next longer by about 9%. This conclusion is based on careful 
measurements of enlarged drawings of the metaphase chromo¬ 
some group in 114 somatic cells from different individuals, selected 
on account of their special clearness and flatness. Of these cells 
51 or 45% contained 15 chromosomes, and in each case the sum of 
the lengths of all the chromosomes was approximately the same, 
showing that increase in number was due to transverse fragment¬ 
ation. This demonstration of constant differences in the length 
of the somatic chromosomes is an important advance in our 
knowledge of CEnotheya chromosomes. While differences in size 
have been previously observed, they have been assumed to be due 
merely to fluctuation, not being great enough to demonstrate 
except by comparison of measurements from many cells. 
Hance assumes that the odd or extra chromosome in 
CE. scintillans is the shortest of the 15. It might equally well be 
the longest so far as his data show. His only direct evidence is 
the difference in total length of the chromosomes in the two 
telophase groups of 7 and 8 chromosomes respectively in the 
heterotypic mitosis of the pollen mother cells. This difference is 
1 Gates 1912b. 
