AND THE Embryo-sac in Bignonia venusta 97 



thread with many small granules, and a few larger ones. At tliis 

 stage I have found no indication of distinct rows of granules. 

 They are evidently rapidly formed, however, for somewhat later 

 there is a general splitting of the ribbon with its numerous gran- 

 ules (as in Fig. 21), the parts cf which may be adjacent for some 

 distance, and separating widely in other places. The divided 

 thread becomes greatly looped and twisted, apparently making 

 definite bends upon itself in certain regions, but there does not 

 seem to be segmentation in the sense that the entire ribbon falls 

 into distinct segments, the whole extent of each segment entering 

 into a single chromosome (see Fig. 22). Nevertheless, each bent 

 or loop-like portion resulting is undoubtedly the basis of a chro- 

 mosome, and in a certain view the loop formation is very evident. 

 The figures indicate that we may have a process closely parallel to 

 w^hat is described in the Hepaticae."^^ ' The resulting chromosomes 

 often appear spherical, with a few projecting edges. They may 

 seem to be made up of two or three chromatic masses closely 

 fused ; and again, or from another view, there is an evident con- 

 cavity on one- surface (Fig. 23). 



Throughout this period, also, minute linin attachments connect 

 the chromosomes and the nucleolus, and the nucleolus is often 

 drawn out into a fusiform condition. Some abnormal nuclei occur 

 in w^hicli the nucleolus is actually drawn out into the form of a 

 ribbon. 



In several anthers I have found two nuclei in many pollen 

 mother-cells during the spirem stage. They are usually abnormal 

 in form, but there has been no indication of how the division has 

 been effected. 



TJic First Division 



On the disappearance of the nuclear membrane the kinoplasm 

 of which it is composed opens into the nuclear hollow, the kino- 

 plasmic threads of the membrane being apparently the first to be 

 attached to the chromosomes. Soon, however, the kinoplasm , 

 enters from all directions, and it is drawai into the form of a truly 

 multipol'ar spindle, as often described (see Fig. 24). The general 



^Farmer, J. 1]. On Spore-Fornialion and Nuclear Division in the Ilepaticae. 

 Ann. Bot. 9 : 469-523. 1S95. 



