Doyle — Observations on the Morphology of Larix leptolepis. 317 



cub off approximately one- third of the whole, the second prothallial cell is 

 still typical and normally placed (PL XVII, fig. 11, and PI. XVIII, fig. 2.) 

 The first large persistent " prothallial" cell can itself divide (PI. XVII, fig. 11). 

 If the first division results in two equal cells, the second prothallial cell is 

 usually small, and may be normally placed relative to the long axis of the 

 grain in both halves of the double grain, as in PI. XVII, fig. 13, or normally 

 placed in one against the middle membrane, as in PI. XVIII, fig. 3, while the 

 other is laterally placed. Compare also PI. XVII, figs. 9 and 12. In fig. 9— 

 a giant grain — there was even sign of rapid degeneration, the nuclei being 

 most uncertain. In PI. XVII, fig. 12, any structure in the left half com- 

 parable in succession to a second prothallial is large and massive, much more 

 so even than in the laterally placed one in PI. XVIII, fig. 3. The position 

 in PL XVII, fig. 10, was only found once, while, owing to the loss of a few 

 sections from the slide at this point, the actual nature of the lower cells 

 was not made clear ; presumably the empty space is a large prothallial cell 

 in this portion. Lastly, no matter how perfect the double grain may be, 

 never more than this single second prothallial cell is cut off. One can gather 

 that the same is characteristic of Picca canadensis, though it is not explicitly 

 stated. Pollock seems to describe the presence of two prothallial cells in 

 one half of a double grain of Picea esecelsa, but it is worthy of note that 

 even he never found two prothallial cells in both halves ; if there were two 

 at one side of the middle wall, there was never found more than one at the 

 other. Of such a case he figures one, and even without his own confession 

 that the course of the central wall was not clear, it is just as easy to interpret 

 his figure as having one prothallial only at each side as in our figure, the 

 apparent extra prothallial really being a small "stalk" cell— and this while 

 of course allowing the possibility of his interpretation being correct, since the 

 prothallial number in Picca cxcelsa is variable. It would appear, then, that 

 as far as Larix leptolepis is concerned, the number of divisions is normally 

 definite. The primary nucleus divides twice before advancing to the forma- 

 tion of what we may term the " antheridial " cells. If the first division 

 results in a very large cell, the fixed appearance of only one evanescent cell 

 in each of the first two sister cells forces us to consider the large half cell as 

 the homologue of the typical first and normally degenerating prothallial cell. 



After returning again to the resting stage, the division into tube and 

 generative cell follows. PL XVIII, fig. 4, shows the nearest approach to a 

 resting stage following this division. As far as could be seen, the nuclear 

 material became rapidly compact, even if no further division took place 

 (PL XVII, fig. 6). Most often the generative cell again divided into the stalk 

 cell and body cell, as before, with compact nuclei. In this stage the grain 



SCIENT. PROC. K.D.S., VOL. XV., NO. XXVIII, 3 F 



