of the Pollen in certain Sweet-Peas. 149 



Fertile Plants. 



Before the phenomena observed in the spermatogenesis of the 

 sterile plants are described, an account will be given of the process 

 normal in fertile plants. The material used for this purpose was 

 taken from pure bred plants of the race known as " Emily Hender- 

 son " — one of the parent races of the sterile plants* 



In these plants the active division of the archesporial cells, to 

 which the pollen-mother-cells owe their origin, is succeeded by 

 a period of rest and growth. Towards the end of this period, 

 at the time when the nuclear reticulum is about to become 

 transformed into the coiled spireme, the threads connecting the 

 nucleolus with the peripheral reticulum become very prominent, 

 taking the stain more deeply than those which form the network 

 (Fig. 1). In the nucleolus there appear small vacuoles, which 

 increase in size while the reticulum shows an increasing affinity 

 for the chromatin stain. In correspondence with its increased 

 vacuolation, the nucleolus, which hitherto has been very deeply 

 coloured, takes the stain to a degree which decreases with each 

 successive stage. The numerous threads by which it was con- 

 nected with the reticulum disappear, apparently being broken off. 

 In some of the later stages the nucleolus was found to be connected 

 to the reticulum by only one rather prominent thread. This 

 single connexion is, I believe, finally broken off; at any rate 

 no connexion is made visible by the stains which have been 

 usedj". The nucleolus is now a spherical body, which consists 

 of a' very faintly stainable shell enclosing a large vacuole, which 

 may be central or excentric. There can, I think, be no doubt 

 that these changes indicate a transference of chromatin (or of 

 some substance closely allied to chromatin) from the nucleolus to 

 the nuclear thread J. The subsequent fate of the nucleolus is very 

 difficult to observe ; I have been able to find no trace of it at any 

 stage later than that at which the spindle becomes conspicuous. 

 In the sterile plants the nucleolus often fragments, and, shortly 

 after the breaking down of the nuclear wall, I have seen in a few 

 cases two or three small spherical bodies lying in the extra-nuclear 

 protoplasm. Both in colour§ and texture these bodies resemble 

 exactly that portion of the nucleolus which remains in the clear 

 area occupied by the nuclear thread (see Fig. 12). This fact would 

 appear to suggest that the remains of the nucleolus may finally be 

 absorbed by the protoplasm of the cell. 



* i.e. p. 91. 



t Iron-alum haematoxylon, and Flemming's triple stain. 



X See Wager, A muds of Botany, 1904, Vol. 18, p. 29. 



§ These preparations were stained with iron-alum haematoxylon. 



