650 Journal of Agricultural Research voi. xii, no. 10 



thickens and a series of scales are formed over its surface. The ex- 

 ternal appearance of the extine is shown in Plate C, figures 15 and 16; 

 Plate D, figures 6 and 15. Soon after the microspore division the extine 

 development is practically complete. 



Division of the microspore nucleus. — ^With the growth of the 

 microspore there is an apparent decrease in chromatin content; for, 

 in the nuclei which are just about to divide, the chromatin is distributed 

 in small particles about the periphery of the nucleus and appears very 

 scant. Transition stages between that seen in Plate D, figure 8, and 

 the completion of the spireme have not been observed. The spireme 

 (PI. D, 9, 10) is a continuous, heavy, dark-staining thread. It follows, 

 more or less, the periphery of the nucleus and surrounds the large 

 nucleolus which has at this time begun to break down. The nucleolus 

 takes on a very irregular outline, which is in some way related to the nu- 

 merous threadlike processes which extend from it to the spireme thread 

 (PI. D, 10). The nucleolus is now very light-staining and seems for the 

 most part homogeneous, but it may contain one or more small vacuoles. 

 Soon after the segementation of the spireme into its 26 constituent parts 

 the nucleolus completely disappears. 



Following segmentation of the spireme the nuclear membrane dis- 

 appears, the chromosomes are drawn to one plane, and the spindle 

 is formed (PI. D, 11). The most usual position of the spindle in the 

 microspores of many plants is near the wall and perpendicular to it; 

 the pole which is to form the generative nucleus being nearly in con- 

 tact with it. As a result of this position the generative cell usually lies 

 against the wall and the vegetative nucleus in the cell cavity (PI. D, 15). 

 This arrangement of the spindle is also found commonly in Fragaria 

 spp., but is not universal. Often spindles are found which lie parallel 

 to the wall (PI. D, 12) and result in the arrangement of the nuclei shown 

 in Plate D, figure 14, both of which lie against the wall. 



The spindle is always broad at the poles and short.' The chromo- 

 somes on the equatorial plate are small and oval in shape and may 

 readily be counted if a section can be obtained in which the heavy wall 

 of the microspore is removed both above and below the chromosomes. 

 In these sections 26 chromosomes are plainly visible. The daughter 

 chromosomes are drawn to the poles simultaneously (Pi. D, 12), no 

 instances having been seen in which chromosomes lagged behind on 

 the spindle or in which there was an extrusion of chromosomes. 



Directly after the rounding up of the daughter nuclei and the dis- 

 appearance of the spindle fibers there is visible no sign of a cell wall or 

 limiting membrane (PI. D, 13) between the two nuclei, which is even- 

 tually to set off the generative nucleus in a separate cell (Pi. D, 14, 15). 

 Soon, however, the wall appears and the generative cell begins to round 

 up (PI. D, 14), eventually to lie free in the cytoplasm of the pollen grain 

 (PI. C, I). 



