396 EDMUND B. WILSON. 



associated (as in a and 5) they may be described as forming an 

 actual tetrad, the first division of which is evidently equational. 



An excellent opportunity is afforded by these stages for critical 

 comparison of the relative sizes of X and Y. Such a comparison, 

 made in a large number of cells, shows that a slight but evident 

 inequality almost always exists, the fact being placed beyond 

 doubt by its appearance in both daughter-pairs (Figs. 18, df). 

 The inequality is however not always seen, perhaps because of 

 slight differences of form or foreshortening. In respect to the 

 size-differences between X and F, Pentatoma senilis seems to 

 approach Oncopeltusfasciatus, as described in my eighth "Study," 

 and differs markedly from P. juniperina, where the difference is 

 as evident as in Euschistus. 



Among numerous cases examined I have not found an excep- 

 tion to the rule that X and Y lie near together, often more or 

 less in contact, and divide in this position. This relation is 

 maintained during the anaphases, so that in the final stages X 

 and Y form a pair that is hardly to be distinguished from the 

 small bivalent autosome, now always distinctly double (Figs. 

 II, 12). Each secondary spermatocyte thus receives three pairs 

 of chromosomes, two of which are the longitudinally divided 

 daughter-autosomes and the third the XY-pair. These three 

 double elements enter the second division spindle directly, 

 without forming a "resting-nucleus," and each is separated into 

 its two components (Figs. 1416). This of course involves the 

 disjunction of X and F. In this division the inequality between 

 these chromosomes is almost always in evidence, the appearance 

 of the XY-pair being somewhat similar to that of Nezara hilaris, 

 as described in my seventh "Study." 



It would be hard to find a more convincing demonstration 

 than is here afforded of the relation between univalents and 

 bivalents in the maturation-divisions (cf. Wilson, '12) or of the 

 fact that the bivalent is equivalent to two univalent dyads united 

 together. Were the association of X and Y only a little closer 

 they would in fact form an actual tetrad essentially like the 

 autosome tetrads. Considering the X F-pair as such, it evidently 

 divides longitudinally (equationally) in the first division and in 

 the second undergoes a typical reduction-division (disjunction of 

 X and F). 



