8 W. J. BAUMGARTNER. 



species of Gryllns show the same conditions. Many of the 

 other cells show the loops more crowded, especially the younger 

 (smaller) ovocytes while the older (larger) ones show them looser. 



With the idea in mind that the chromatin thread is in loops I 

 succeeded in finding Fig. 5 in the early spermatocytes. This 

 figure shows the loops as much massed leaving as much of the 

 nucleus clear as any case I have observed. In most cells the 

 loops fill the nucleus completely ; besides the large majority of 

 the cells are so placed in the field that they show nothing of the 

 looping of the thread. This I consider the characteristic of the 

 "synapsis stage" in the testis of Gryllns and not the massing to 

 one part of the nucleus as Montgomery did in his work on 

 Pentatoma. 



This stage appears quite early in the growth period and lasts a 

 comparatively long time. All the cysts from which Figs. 5, 6, 7 

 and 8 are drawn show some cells where the loops of the chro- 

 mosomes are quite evident, but most cells show little or nothing 

 of such loops owing to the plane of cutting, as indicated in the 

 figures. 



After this stage the chromatin becomes a little more diffuse 

 again before forming the definite chromosomes of the first sper- 

 matocyte division (Fig. 10). We shall now pass to the meta- 

 phase where the chromosomes are completely formed and lie in 

 the equatorial plate region and return to the late prophases sub- 

 sequently. 



A glance at Fig. 18 will give the reader an idea of the various 

 forms of chromosomes seen in a cyst full of cells in this stage. 

 Drawings of all the forms that could be found showed the possi- 

 bility of classifying the shapes into rings, crosses and rods. The 

 rod may be straight, or bent so as to form a bracket, parenthesis 

 or an S. The crosses may have the ends bent so as form an / 

 or an e where one arm is very short or absent. All these shapes 

 can be seen in Figs. 13-19. It was observed that many cells 

 showed two rings (Figs. 13 and 18, a) and one or two crosses and 

 several rods some of which were straight. If there were many 

 straight rods fewer other shapes appeared. Some showed no 

 rings others no crosses. Naturally the question arose : is there 

 any constancy in the number of chromosomes in a cell that 

 assume a certain shape ? 



