April 1892.] THE BOTANICAL SOCIETY OF EDINBURGH. 379 



gidse; along this strand the antipodal nucleus approaches 

 the micropylar (fig. 29^ 29*^). The nuclei are represented 

 in three stages of approximation in figs. 29^ 29^ 29^ both are 

 evidently lying in the micropylar half of the embryo-sac, 

 close to the synergidse or ovum ; in 29*^ the nuclei have not as 

 yet come in contact ; in 29° they are markedly flattened off 

 against one another, and in 29*^ the two nucleoli have come 

 into contact. In fig. 30 the embryo-sac is seen to be lined by 

 a thick layer of protoplasm, sending strands and columns to 

 the conjugating nuclei ; while in fig. 31, after the fusion of 

 the nuclei, the main bulk of the protoplasm is aggregated 

 round the newly formed primary endosperm-nucleus, com- 

 paratively little protoplasm remaining in contact with the 

 wall of the sac. 



Let us study next the details of this conjugation of nuclei 

 resulting in the formation of the primary endosperm-nucleus ; 

 figs. 30-43 illustrate the process as it occurs in Myosurus, 

 and figs. 45-47 show a few steps of the conjugation in Scilla 

 nutans. The latter plant, being a ]\Ionocotyledon, has large 

 nuclei, the details of whose structure are more readily made 

 out than in the comparatively small nuclei of Dicotyledons. 

 The preparations of Scilla were made by me in the Laboratory 

 of the Eoyal Botanic Garden by my picro-corrosive method 

 for Professor Bayley Balfour while acting as his Assistant. 



After the micropylar and the antipodal nuclei have come 

 in contact with one another, we see, in hsematoxylin-prepara- 

 tions (fig. 32), each nucleus enclosed by a very delicate, 

 feebly-stained envelope, the nuclear membrane {n. m.), the 

 inner surface of which lies in direct contact with the chro- 

 matin-granules (c/w) of the nucleus proper. Within the 

 nucleus lies a large deeply-stained nucleolus (71II.) enclosed 

 by a very faintly-stained nucleolar membrane (nil. ??i.).* The 

 protoplasm of the nucleus seems to be divided into a peri- 

 pheral darker (1), and a perinucleolar fainter (2), portion. 

 This appearance of the nucleo-hyaloplasm may be really 

 indicative of the normal structure of a nucleus, as Frommann 

 points out, or may be in this instance simply an artificial 

 product, brought about by a separation and retraction of the 

 fibrils of the nucleo-hyaloplasm from the nucleolar membrane, 

 due to my imperfect method of fixing. But even if this 



* The nulcleolar membrane has not been represented in fig. 32. 



