STUDIES ON CHROMOSOMES 413 



univalent chromosome (X) of the same size. It is most inter- 

 esting to observe in Protenor the gradual emergence of the biva- 

 lents from the confused nuclear threadwork of Stage g. Early 

 stages of the process are seen in figs. 115 to 117, which clearly demon- 

 strate (1) that the threads do not constitute a continuous spireme, 

 but are separate, (2) that they do not lie side by side in pairs to form 

 a diplotene, but are single and undivided. Figs. 118 and 119 show 

 two nuclei, only a little later than the preceding, in which all the bi- 

 valents are clearly seen and the large one is perfectly evident. It 

 is of course only now and then that a nucleus in this stage can be 

 found in which all the chromosomes are thus clearly distinguish- 

 able; the bivalents are still so extended and irregular as to present 

 a hopelessly confused picture in sections, and very frequently also 

 in smears. It is however my firm belief that the bivalent-figures, 

 intricately entangled though they are, are already quite distinct 

 at least as early as fig. 115, and I do not hesitate to accept this 

 as probable for the still earlier and more confused nuclei that 

 precede. 



From the latter part of Stage h (i.e., figs. 118, 119) every step 

 may readily be followed as the chromosomes continue to con- 

 dense, contract and increase in staining capacity. Of the innu- 

 merable nuclei showing these stages in my smear-preparations a 

 few are shown in figs. 120 to 131, and in photos. 40 to 51. As 

 these figures show, the typical number of separate chromatin- 

 bodies is eight, which may however be reduced to seven by the 

 coupling of the two smallest (m-chromosomes, figs. 124, 129, 

 130, photos. 45, 46, 48), or in certain rare abnormalities may be 

 increased to more than eight (fig. 128, photo. 43). Of these 

 eight, three are univalent, namely, the two smallest (m-chromo- 

 somes) and the large X-chromosomes, the latter always distin- 

 guishable by its more compact consistency, greater staining capac- 

 ity, rod-like form, and simple longitudinal split. 



The remaining five are the bivalent autosomes, which from the 

 beginning have the same forms as in Oncopeltus — i.e., double 

 crosses, double F's, or longitudinally split rods which sooner or 

 later develop a transverse suture at their middle points and thus 

 are plainly seen to be of quadripartite nature. 



