3 14 LESTER W. SHARP 



culum to single or double lines. His « corkscrew " spirem (his fig. 64) 

 probably represents the zigzag stage. According to Gates (12) the chromo- 

 somes of Oenothera lata are formed by the fusion of the threads of the 

 reticulum. Karpoff (04) describes and figures " stellate bands ^ in the 

 prophases of Vicia, which apparently correspond to the reticulate bands of 

 the présent account. Franck's (ii) figures 4 and 5 of }^icia are also sug- 

 gestive. 



Grégoire and Wygaerts (o3), Grégoire (06), Nëmec (10) and Digby 

 (lo) were led to conclude from their researches that the formation of the 

 chromosomes may proceed in the same or différent plants by two methods : 

 first, through the production of thin zigzag threads as above described, 

 and secondly by direct concentration without the zigzag stages. It is évi- 

 dent in Grégoire's préparations, which the writer bas had opportunity to 

 see, that Trillium shows a somewhat variable behaviour in this respect. In 

 Vicia, on the other hand, appearances which might lead to a similar inter- 

 prétation seem to be due to the action of the fixing agents. In the cells of 

 the central région of the root, where fixation is very poor in Flemming's 

 stronger solution, it often appears that the net is passing over directly to 

 fairly thick threads. After the solution of Benda, which fixes niuch better 

 the interior cells, such an altération of the structures is comparatively 

 slight. In the peripheral cells, which such coinparisons show are well fixed 

 in either fluid, the nuclei as far as can be determined always pass through 

 the zigzag thread stage. The newly formed threads, however, are by no 

 means always of the same thickness (compare fig. 6-9 with fig. 10). 



It follows from this description that the interprétation proposed by 

 WiLSON (12(3, b), largely on the basis of Bonnevie's studies on Al/iiiin, is 

 not applicable to Vicia. We hâve hère to do with an irregular condensation 

 of ail the material to form a zigzag thread, and not with the unravelling of 

 a coiled thread formed within the old chromosome. Thèse facts are of im- 

 portance in connection with Roux's theory of the split. 



It is in the thin homogeneous threads that the true longitudinal split- 

 ting of the chromosomes begins, fig. 10 et seq. As soon as the thread 

 becomes sufficiently equalized small vacuoles appear along the axis and 

 rapidly develop into a more or less continuons split. Not ail the threads, 

 nor even ail portions of the same one, undergo the change at the same 

 time. In fig. 6, 9 and 10 it is clear that if we consider ail the nucleus at 

 once the processes of condensation, straightening, equalization and .split- 



