April 1892.] THE BOTANICAL SOCIETY OF EDINBURGH. 357 



walls of the cells x and ij (fig. 10 h). Xot at all iinfrequently, 

 especially in vigorous ovules, all the walls, with exception of 

 those lying in direct contact with the plerome and the 

 dermatogen (fig. 9'\ 10 c, 10 d), show the characteristic change. 

 During the early stages of the gelatinisation the cell-walls 

 stain deeply with Kleinenberg's hsematoxylin, and on 

 surface section (fig. 10 c, y) the whole cell-wall may appear 

 as a dark violet plate. 



This gelatinous swelling of the cell-walls occurring in the 

 three or four cells derived from the archesporium is in every 

 respect analogous to the swelling which takes place in the 

 .sporocyte-walls of other sporangia, c.fj., in the pollen-sacs of 

 Angiosperms, in Selagindla, &c. As further, this change in 

 the walls is, at least in Myosurus, not restricted to those cells 

 not undergoing further development, but as it also occurs in 

 the young embryo-sac-cell, all the cells derived from the 

 archesporium and showing this gelatinisation must be re- 

 garded as sporocytes. Of the number of sporocytes formed 

 in Angiosperms one only normally undergoes further develop- 

 ment, and ultimately gives rise to the embryo-sac with its 

 eight nuclei. 



That the embryo-sac of Angiosperms must correspond to 

 one sporocyte was a conclusion I had arrived at in my last 

 paper * ; but, notwithstanding this, I reason on the very next 

 page that it must be the equivalent of two sporocytes. This 

 mistake was brought about thus : Trying to disprove the 

 view that the archesporium might be regarded as a sporocyte, 

 I stated that it did not always give rise to four cells, but 

 that it might give rise to a number of cells, varying from two 

 to seven ; as further, my mind had been greatly impressed by 

 the constancy with which each sporocyte in the higher 

 plants gives rise to four spores, and, as finally, I was aware of 

 two groups of four nuclei always occurring in the embryo- 

 sac, I was led to believe each of the two groups of four nuclei 

 to be derived from one sporocyte, which naturally would 

 make the embryo-sac the equivalent of two sporocytes. This 

 latter conclusion being at variance with the conclusion I had 

 arrived at at first, an explanation was sought, which resulted in 

 the following : — " The division of the embryo-sac makes the 



* Trans. Bot. Soc. Ediii. (June 1891), pp. 136-148. The conclusion will be 

 found on p. 140. 



