378 EDMUND B. WILSON 



now very clearly seen to be separate, constituting a segmented 

 spireme. All the threads still stain deeply and are very much 

 thicker than in the leptotene-stage; hence these nuclei may be 

 called the pachytene-nuclei. In the earlier part of this stage it 

 is uncertain whether the threads are longitudinally split or not; 

 in many cases the closest study fails to reveal such a split in sec- 

 tions (figs. 62, 63), though in smears (fig. 65, photo. 10) the split 

 is very clearly seen in nuclei that seem to belong to this period. 

 In the later part of this period the threads become still thicker 

 and shorter and very often show a conspicuous longitudinal cleft. 

 This is less readily seen in Oncopeltus and Lygaeus (figs. 64, 83, 

 84) in which, indeed, the threads sometimes do not show a trace 

 of such a cleft at this time (which I attribute to defective fixa- 

 tion) . In Largus, on the other hand, the cleft appears in the most 

 conspicuous way, especially in sections fixed with Hermann's 

 fluid (figs. 81, 82), where the threads are often seen to consist of 

 double rows of granules often showing a distinctly paired arrange- 

 ment. 



In Oncopeltus and Largus the sex-chromosomes are at this time 

 hardly changed, still having the form of undivided, rounded chro- 

 mosome-nucleoli. In Lygaeus, the F-chromosome is still of this 

 type, but the X-chromosome (usually near the nuclear membrane) 

 is now very clearly split lengthwise (fig. 84), in which condition it 

 persists from this time throughout the whole growth-period. The 

 plasmasome is considerably larger than before although not yet 

 at its maximum size. 



The number of chromosomes (separate chromatin-masses) is 

 now obviously approximately half that of the diploid groups. In 

 Oncopeltus and Lygaeus this can be determined only approxi- 

 mately; but it is certain that the number is not far from the 

 reduced or haploid number that is to say, there are in Oncopel- 

 tus, in addition to the two chromosome-nucleoli, about seven 

 separate diplotene-threads, in Lygaeus about six. In Largus 

 cinctus (where the spermatogonial number is eleven) nuclei may 

 readily be found in which the number of double threads may be 

 exactly counted. Such nuclei show, in addition to the single 

 chromosome-nucleolus, five double threads (figs. 81, 82) of which 



