1 88 Edmund B. Wilson 



to those seen in other insects that I cannot doubt their essential 

 similarity also in respect to sex-production. 



In Metapodius the facts are still more evidently opposed to the 

 quantitative interpretation. The number of chromosomes has 

 here no relation to sex-production; and, as w^ill be seen from the 

 table at p. 149, in the forms with supernumeraries the relative 

 frequency of high numbers and of low is nearly equal in the two 

 sexes. If my general interpretation of the chromosomes in this 

 genus be correct, a like conclusion applies to the total relative 

 mass of chromatin in the two sexes ; for all individuals ahke possess 

 the type-group of 22 chromosomes (Montgomery's form excepted) 

 while the supernumeraries represent the excess above this amount. 

 I have endeavored to determine whether this appears in direct 

 measurements, independently of my general interpretation; but 

 have found this impracticable for several reasons. Very consider- 

 able differences in the apparent size of the chromosomes are pro- 

 duced by different degrees of extraction; but this will not account 

 for the considerable differences seen in the same slide when the 

 extraction is uniform. It is evident that the actual size of the 

 chromosomes varies with the size of the cells; for example, both 

 in Metapodius and in many other genera, the chromosomes in 

 the larger spermatogonia near the tip of the testis are larger 

 (in many cases much larger) than those of the smaller spermato- 

 gonia of other regions. How great the differences are may be 

 appreciated by a comparison of the figures. For example, in 

 the spermatogonial groups of No. 2 (23 chromosomes. Fig. 7, 

 v-x), the chromatin mass is obviously much greater than in 

 those of No. 21 (24 chromosomes, Fig. 11, /-/). In the 25-chromo- 

 some female groups shown in Fig. 12, i-k (No. 27), the chromatin 

 mass is evidently much less than in the 21-chromosome male group 

 shown in Fig. i, ^, or in the 23-chromosome male groups of Fig. 7, 

 v-x. Conversely, the 22-chromosome female group of No. 44 

 (Fig. 4, s) shows a much greater chromatin mass than in the corre- 

 sponding male group of No. 46 (Fig. 4, o), or the male 24-chromo- 

 some group shown in Fig. 11, /. 



Evidently, therefore, the relative mass of chromatin can only 

 be determined by means of accurate measurements of both the 



