of the Pollen in certain Sweet-Peas. 151 



deeply stained network, and then a further fusion into the large 

 nucleoli found in the mature cells*. This description is not 

 entirely applicable to the process which I have observed in the 

 reduced nuclei of the pollen grains. The nuclei certainly pass 

 through a stage in which there is a single continuous thread 

 (Fig. 23), which is directly formed by the fusion of the chromo- 

 somes end to end. Almost immediately after this stage is reached 

 the coils of the thread are seen to be connected with one another 

 by fine branches. From this stage the reticulate condition is 

 attained. 



Attempts made to trace the origin of the nucleolus met with 

 no more success than those made to trace its ultimate fate in 

 mitosis. One or two nuclei are first observed as small, spherical, 

 faintly-staining bodies lying in close proximity to the chromatin 

 thread with which they are connected by a fine thread or threads. 

 The single large nucleolus of the later stages is formed by the 

 fusion of the smaller ones which, in the intermediate stages, are 

 seen closely applied to one another. The nucleolus is connected 

 with the nuclear reticulum, to which the chromatin thread has now 

 given rise, in the early stages sometimes by only one connecting 

 thread (Fig. 24), but later by more. 



These observations appear to me to indicate that in the 

 process of reconstitution of the daughter nuclei exactly the same 

 changes occur as in the prophases of division, but in the reverse 

 order; and, further, that the nucleolus receives and stores the 

 bulk of the chromatin during the resting periods which intervene 

 between successive mitoses f. 



Sterile Plants. 



The somatic mitoses of these plants (including those which 

 take place during the development of the archesporium) are quite 

 normal, and differ in no way from those of the fertile individuals 

 of the pure race. The rapid growth of the archesporium is, 

 however, followed by an abnormally long period during which no 

 divisions of the pollen-mother-cells take place. This is to be seen, 

 not only from the large number of preparations which are obtained 

 in this condition, but also from the large growth of the anther 

 itself which takes place. Since there is no corresponding increase 

 in the sporogenous cells, the pollen-mother-cells are found lying 

 quite loosely, and in the later stages almost scattered, in a large 

 cavity. A deceptive appearance is thus produced which is sug- 

 gestive of the formation of a smaller number of pollen-mother-cells 

 than is usual, but is in reality due to the fact that a fertile anther 



* Annals of Botany, 1904, Vol. 18, p. 47. 

 t Cf. Wager, I.e. 



