412 EDMUND B. WILSON 



ent light. The history of these chromosomes as contrasted with 

 that of those just considered, nevertheless affords some important 

 evidence bearing on this question. As has been stated, the pro- 

 phase-bivalent is of quadripartite composition (though this may 

 fail to become visible until the later prophases) while thepro- 

 phase-univalent is bipartite. This has already been emphasized 

 in the case of Oncopeltus, but may studied to still greater advan- 

 tage in Protenor, because of the greater size of the chromosomes 

 and of their very marked individual size-differences. In the male 

 of this form, as heretofore described by Montgomery, Morrill 

 and myself, the spermatogonial groups contain thirteen chro- 

 mosomes, and the unpaired ^-chromosome is very nearly twice 

 the size of the largest pair of autosomes (photos. 35, 36). In the 

 female two such JT-chromosomes are present (photos. 37, 38). 

 In the male the X-chromosomes of course remains univalent 

 throughout the entire maturation-process, while the large pair of 

 autosomes produces a bivalent that is of nearly the same size as 

 the univalent X. The latter is at every period distinguishable. 

 In the earlier stages of the growth-period, when the autosomes are 

 in a diffuse and lightly-staining condition, it remains compact, in 

 the form of a somewhat elongate vermiform body, that is closely 

 coiled about or within a plasmasome to form a rounded "chromo- 

 some-nucleolus the true nature of which only appears after con- 

 siderable extraction (cf. Montgomery, '01, fig. 127). In smears 

 (as is the rule among the Hemiptera) the plasmosome usually 

 collapses, setting free the X-chromosome, when its rod-like form 

 becomes clearly apparent (photo. 39). In later stages it shortens, 

 thickens and splits lengthwise, so as to appear in the prophases 

 as a rather short, thick rod, very plainly split (figs. 120-131, photos. 

 40 to 51). The large bivalent, on the other hand, first becomes 

 recognizable in the prophases, as the process of condensation occurs, 

 when it may be studied to the best advantage in smear-preparations, 

 in which all the chromosomes are spread out in one plane. 



In such preparations, of which I have a large number, the large 

 bivalent is invariably distinguishable by its large size nearly 

 twice that of any of the others; and we thus have opportunity to 

 compare it accurately, side by side in the same nucleus, with a 



