Doyle — Observations on the Morphology of Larix leptolepis. 315 



folgende Scheidewande entstehen, . . . . ; die Scheidewande welche die 

 grossere Pollenkornzelle von der anderen kleineren abgrenzt, ist immer 

 mehr nach innen gewolbt als die anderen Scheidewande." No figures are 

 given, so the reference is not clear. It may refer to a three-prothallial cell 

 development, but such has not been observed by me in Larix leptolepis. 



2. Variations in the grain. — Practically speaking, PI. XVII, fig. 5, shows 

 a typical mature grain — the prothallial cells a little variable, hut always 

 very distinct. The variations to be mentioned presently were found in only a 

 small percentage of the total grains ; but actually they should not be counted 

 in the bulk, because they never appeared in the top sacs of a cone, only in the 

 lower sacs, and mainly in the last two whorls. Some cones are singularly free 

 from these variations, but in others they comprised at least 50 per cent, of 

 the grains of a sporangium. 



Some of these double-pollen grains are shown in PI. XVII, figs. 9, 10, 

 11, 12, 13. Some are marked by much degeneration, but a perfect double 

 grain appears in fig. 13. 



Irregularities of this type have already been indicated in Picea excelsa, by 

 Pollock (19) ; Picea canadensis and Abies balsamea, by Hutchinson (7, 8) ; and 

 by Coker (2) in Larix europea. Only in Picea canadensis (7) have they been 

 fully described. But even so, we find Coker's brief account of double-pollen 

 development differs from Hutchinson's. With the latter the double grain 

 arises as a result of the twisting through a right angle of the normal first 

 division spindle, the resulting wall running straight across the cell. With 

 Coker the pollen mother-cell divides occasionally into two only; and although 

 here the wording is not quite clear, with the help of his figures we can gather 

 that these do not separate, but that the wall between them remains as the 

 median wall of the double grain. Hutchinson, not recognizing this, even says 

 that his fig. 13 is the same as Coker's fig. 6. Eeally they are absolutely 

 distinct. In Coker's fig. 6 the outline is the wall of the pollen mother-cell; 

 the median line is really the adjacent and distinctly separate walls of the 

 two grains within. Hutchinson's fig. 13 has one outer line — the intine, and 

 one thin median wall continuous with it. In view of this, it was thought 

 useful to follow fully the development of the grain of Larix leptolepis. 



3. Development of Male Gametophyte. — The pollen passes the winter in the 

 mother- cell stage. Its resting period is not long if the season is mild. Thus 

 the first tetrad divisions were completely missed in January, 1916, even 

 though the first collection was made in the middle of the month, every cone 

 examined having passed the stage. The second week of February is the other 

 limit. The development is rapid and, as usual, irregular. All stages can be 

 found during the last week of February, while the tree has shed all the pollen 



