1916] Brandt: Notes on Trillium 55 



reports for the heterotypic mitosis in T. grandiflorum, and as Gregoire 

 (1912) reports for somatic mitoses in the same species and in Galtonia. 

 As the chromosomes migrate to the poles they gradually divide and 

 before the cross-wall is formed in the equatorial region twelve sep- 

 arate chromosomes may be found in each polar group. They are much 

 more slender than the bivalent chromosomes, are irregular in outline, 

 and a single row of granules is plainly discernible. The spindle fibers 

 disappear and a cross-wall is formed before the homotypic spindle 

 fibers appear, but a true resting stage was not found, such as Ernst 

 (1902) reported to intervene between the heterotypic and homotypic 

 divisions in the megaspore mother-cell of Paris. My results seem to 

 confirm those of Atkinson in this matter. The homotypic division 

 takes place so rapidly that I have been unable to obtain preparations 

 showing stages earlier than the polar groups of six chromosomes each. 

 Soon after the homotypic division has been completed cross-walls are 

 laid down to form the tetrad. 



The tetrads soon break up into very angular young pollen grains 

 whose non-vacuolate cytoplasm appears to be very compact. Soon the 

 pollen grains begin to round out, but at no time in their development 

 does their cytoplasm become vacuolated, as Dorsey (1914) reports is 

 the case in Vitis species. As stated above, in approximately three 

 weeks after the homotypic division many of the pollen grains have 

 attained maturity, the exine has its characteristic markings, and the 

 nuclei are dividing. The microspore nuclei do not divide with the 

 uniformity in regard to point of time characteristic of pollen mother- 

 cells, and only a few division figures may be found at any one time in 

 a given locule (pi. 10, fig. 1). Many pollen grains in this species 

 never round up properly and many others are less than the normal size. 



The cells of the tapetum attain their maximum size when the pollen 

 mother-cells are in the tetrad stage. Their nuclei are very large and 

 very rich in chromatin. When they divide, the chromosomes — having 

 apparently the full somatic number — form a very compact division 

 figure. Variations from the binucleate condition have not been found. 

 As the young pollen grains mature the cells of the tapetum gradually 

 lose their contents and become flattened against the outer layers. 

 Pollen sterility is often anticipated in an abnormal condition of the 

 tapetum, which may become separate from adjacent parietal layers or 

 may begin to degenerate before the pollen mother-cells divide. Other 

 tissues of the anthers also degenerate. Most frequent is the degener- 

 ation of the tissue between adjacent locules, which may include the 



