A Comparative Study of the Chromosomes etc. 
251 
the last spermatogonial telophase to the early leptotene stage that this 
phenomenon occurs. 
A comparison of cclls of the last spermatogonial generation (PI. XIX, 
Figs. 17, 18, 19) with those of earlier generations (Plate XIX, Figs. 14, 
15, 16) from the same tubule shows that the degree of stainability 
in the quiescent period increases in the later generations and that the 
chromosomes in division, although more closely massed, are not notice- 
ably smaller than those of earlier generations. The chromatin, therefore, 
increases from generation to generation more rapidly than the other 
constituents of the cell. It is possibly the lack of non-chromatic ele- 
ments, upon which the chromatin should distribute itself, and the diffi- 
ciilty of interchange between nuclear and cytoplasmic substance which 
cause the ceUs to degenerate. The chromatin becomes vacuolated, hyaline 
and clumpcd together in irregulär masses. 
There are twenty chromosomes in the spermatogonia (Plate XIX, 
Figs. 15, 18, 20—23), the most noticeable ones being a large pair, the 
macrochromosomes (M), which are often somewhat curved and occasion- 
ally show a constriction at the middle. Sometimes the second largest 
pair can also be made out, but the others cannot be paired accurately, 
nor is it possible to distinguish any single one of the eighteen sniaUer 
chromosomes which differs from the others in size, in staining reactions, 
or in any other way, sufficiently to allow its being identified with either 
a large or a small idiochromosome. These conelusions are substantiated 
by the fact that in all the gemini formed by the union of the chi’omosomes 
in pairs in the fii’st maturation division the components of each pair are 
of equal size. 
After comparing the relative value of sniears and sections in studying 
the chromosome numbers in spermatogonia, I am convinced that sections 
are far more reliable. Sniears, in fact, are often misleading. Although 
the chromosomes are well separated in smears (Plate XIX, Fig. 24), yet 
there is a tendency for the long macrochromosomes to separate into two, 
thus giving the appearance of twenty-two chromosomes. Since in Anasa 
the X-chromosome is as long as the macrochromosomes in Enchenopa, 
it is possible that we may here have an explanation of the fact that Foot 
and Strobell count 22 chromosomes in smears of this species, whereas 
Wilson (’05a, 05b), Montgojiery (’06), and Lefevre and McGill (’08) 
find only 21. The material at my command in the form of sections was 
very abundant and always showed with remarkable clearness twenty 
well defined chromosomes, never nineteen, which was the number pre- 
viously reported by Boring (’07). My conclusion is based on the study 
