4 Proceedings of the Royal Irish Academy. 



appearances wliicli led me to adopt this view of the process are briefly, 

 the following : — 



(1.) The frequent loops on the double twisted thread (figs. 2 and 

 10) could scarcely be found if the two twisted portions arose by the 

 longitudinal fission of a single thread. The assumption of the fission 

 origin would involve the assumption of a premature transverse fission 

 also accompanied by the fusion of some of the adjacent ends of the 

 partial threads. Such a transverse fission in this stage has not so far 

 been recorded. 



(2.) The twisted portions of the thread are not in close proximity 

 to one another throughout the whole course of the double thread, 

 but "wide divergences are of frequent occurrence, forming secondary 

 loops (figs. 2 and 9), If the adjacent portions arose by longitudinal 

 fission the divergences would necessitate a longitudinal displacement 

 of the portions past each other throughout the thread. Such a 

 sliding of the viscid portions of the chromatin thread past each other 

 seems highly improbable. Divergences of this nature, involving 

 more or less longitudiaal displacement (if longitudinal fission is 

 assumed), are also figured by Miss Sargant^ and Schaffner.^ 



(3.) The thread of nuclei, which are in a transitional stage between 

 the dolichonema and the strepsinema condition, in places appears 

 double and twisted and in places single. Where it is double, the diameter 

 of each of the portions is apparently equal to the diameter of the 

 single thread. If the two portions of the double thread were really 

 derived by the fission of the single thread, it is evident that the single 

 portion should be nearly double as thick as each of the two twisted 

 portions. The actual proportion of the single to the double thread is 

 shown in fig. 2, and also by SchafPner. ^ 



(4.) If the twisted portions of the strepsinematous thread had 

 arisen by fission, the chromomeres of the adjacent portions should 

 correspond, since the opposite chromomeres of these threads have 

 arisen, by hypothesis, from single chromomeres of the original thread. 

 Frequently, however, this is not the case, as observation will show 

 (fig. 9), and as, indeed, the careful drawings of Schaffner* and 

 Gregoire' demonstrate. The non -correspondence of the chromomeres 

 in the twisted portions is what would be expected if these portions 



1 Ann. of Botany, Sept., 1896, and June, 1897. ^Bot. Gaz., June, 1897,%. 10. 



sSchafiher, Bot. Gaz., June, 1897 : fig. 10a. 



^ Schaffner, loc. cit., figs. 8i and 8c. = La Cellule, t. xvi., 2« fasc, fig. 2. 



