318 Scientific Proceedings, Royal Dublin Society. 



is shed. This last division must follow the previous one with great rapidity, 

 because, not even once in two years' gatherings was the division into stalk 

 and body cell seen, or any stage of it. Only in typical grains did it occur. 

 Even if the first division cut off one-third — as also in the cases of equal divi- 

 sion — the final stage, unless degeneration, as was common, intervened, was 

 the tube cell and generative cell condition. See especially PI. XVII, figs. 10 

 and 13, and PI. XVIII, fig. 3. Picea excelsa (19), however, goes on to the next 

 division, while Larix europea agrees with Larix leptolepis. 



Again, a comparison between PI. XVII, fig. 12, and PI. XVIII, fig. 3, is 

 of interest. The "horns" characteristic of the single grain are visible in 

 PL XVIII, fig. 3. PI. XVII, fig. 12, can be interpreted without strain as 

 due to a complete formation of such a wall, the three cells on each side being 

 tube, generative, and " second " prothallial. 



Finally, a comparison between the grain of Larix and that of Dacrydium 

 brings out an interesting point. On the authority of Miss Young's paper (31), 

 it can be stated that in Microcachrys each of the two prothallial cells is bounded 

 by a cellulose wall ; in Saxegothca only the first prothallial cell is so bounded, 

 the second being marked off only by a distinct plasmatic membrane : while 

 in Dacrydium, the first as well as the second prothallial cell is bounded by 

 such a layer only. Do we strain a point in considering these plasmatic layers 

 as reduced cellulose walls ? Again, the generative cell in Dacrydium is marked 

 off from the tube cell by just such a distinct plasmatic membrane. Can we 

 not compare the ephemeral wall round the generative nucleus of Larix — for 

 so we earlier interpreted it — with this distinct membrane in Dacrydium, both 

 representing a primitive cellulose wall bounding the generative nucleus, with 

 this difference, that it remains complete as a plasmatic membrane in Dacry- 

 dium; while in Larix, though fading and rudimentary, it has retained its 

 cellulose-wall nature ? If this be so, it seems that the claim put forward (12), 

 that the grain type as seen in, say, Pinus, is ancestral for all conifers, means 

 that Larix and Dacrydium must both have evolved anew a wall round the 

 generative nucleus and lost it again. 



II.— The Female Strobilus and G-ametophyte. 



It is mainly on this side of the gametophyte history of Larix that the 

 gaps already indicated exist, and for these reasons : — The work was begun in 

 the spring of 1915, in which year, as little time was available, it was decided 

 to make only collections sufficient to show the approximate periods of the 

 year in which critical stages should be sought. This was, perhaps, necessary 

 in any case, as the monthly stages of no species of Larix were known, and, 



