384 Edmund E. Wilson. 



for the spermatogonial division is most abundant in the case of 

 Lygaeus, which is much the most favorable form for an accurate 

 count, the chromosomes being well separated and showing with 

 almost schematic clearness. In the preparations of this form 

 numerous spermatogonial plates appear, showing, whenever an 

 accurate count can be made, without exception 14 chromosomes 

 (Fig. 5^, /?). Nezara, of which numerous spermatogonial 

 divisions are also available, shows the same number and with 

 almost equal clearness, though the chromosomes are in this form 

 more crowded. In the other forms the material is less abundant, 

 but the relations are clearly shown in most of them. Coenus 

 (Fig. 5/) Euschistus (Fig. 5y), and Brochymena (Fig. 777) also 

 show 14 spermatogonial chromosomes, while in Podisus (Fig. 5^) 

 the number is 16. * In all these cases, therefore, the spermat- 

 ogonial number is double that of the chromosomes in the second 

 spermatocyte-division, and two less than double the number in 

 the first division. The most striking fact is that in all these 

 forms, with the exception of Nezara, the spermatogonial groups 

 show but one microchromosome (marked i in Figs. 4, 5, 7); 

 in striking contrast to the fact first determined by Paulmier in 

 Anasa and afterward by Montgomery in many other Hemiptera, 

 that two such bodies ("chromatin-nucleoli" of Montgomery) 

 equal in size, are often present. Did this observation rest only on 

 the examination of a few division-figures (as in the case of Coenus, 

 Euschistus, Podisus and Brochymena) I should hardly trust in its 

 general applicability to the species; but in Lygaeus numerous 

 demonstrative cases remove every doubt regarding this point, and 

 the agreement of the other forms in their later history makes it 

 nearly certain that my observation is not at fault with them. 2 



Distinct, though not very great, size-differences may be observed 

 in the larger spermatogonial chromosomes, as has been indicated 

 by Montgomery in several other genera of Hemiptera. Though 

 these differences are not nearly as marked as those recognized 

 by Sutton in Brachystola, it is nevertheless pretty clearly evident 



JMontgomery ('01, i) gives the numbers as follows: Euschistus variolarius 16, E. tristigmus 14 

 Nezara 16, Ccenus 14, Brochymena 16, Podisus 16. 



2 This point is emphasized since Montgomery describes and figures two spermatogonial microchro- 

 mosomes ("chromatin-nucleoli 1 ') in Euschistus variolarius ('01, i, Figs. 2, 3), E. tristigmus (pp. cit., 

 Fig. 20), Coenus delius (Fig. 55), and Brochymena (Fig. 47). One of the figures of the first-named 

 species shows them equal, the other unequal, in size. 



