304 PROCEEDINGS OF THE ACADEMY OF [April, 



exactly correspond at this time to the ring-shaped chromosomes figured 

 by vom Rath (1892) for Gryllotalpa. Blackman (1905) considers the first 

 maturation division of Scolopendra a longitudinal one because "the 

 longitudinal division of the chromatin segments Is the first which occurs 

 in prophase"; but though in promethea the longitudinal division is 

 present from post-synapsis, yet this is clearly the plane of the second 

 maturation division and must be explained as a precocious splitting. 

 Constrictions such as are seen in figs. 22 and 26 can be easily traced 

 to such stages as are seen in figs. 28, 30 and 31. By concentration small 

 thicker rings are formed, until, by continued thickening, the central 

 space becomes more and more reduced and in most cases the longi- 

 tudinal split is entirely or almost entirely concealed. It may be^ re- 

 tained as a narrow slit between the two ends of a dumbbell or a rounded 

 or diamond-shaped space in the center of an occasional tetrad (figs. 

 37 and 38). 



ft r w ■ H 



c 



b <x D ° G '^ c O d ^ yt S * 



A B 



The changes in form which the chromosomes undergo during pro- 

 phase are shown in text figs. A and B. 



In early prophase the chromosomes appear first as granular aggre- 

 gations, as at A and E; these aggregations next open wide to form 

 granular rings B; then by condensation the rings become much thicker 

 and the granular appearance is lost. Rings may be lengthened, C, and 

 show slight constrictions which mark the plane of the first division, or 

 may be clearly rounded, D; F and G shows an approach to tetrads. 

 Complete condensation is rarely seen in smear preparations, a, b, c, 

 are the earliest stages shown in sections. These are made up of 

 more of less condensed granules with no indication of a longitudinal 

 split, but a slight constriction is usually seen as in a and b ; and though 

 no split is seen, yet stages like d and e make it necessary to assume such 

 a stage. The granules are single and the ring of about one-half the 

 thickness of the loop, and the ends which overlap in d become joined in 



