Gates, Tetraploid Mutants and Chromosome Mechanisms. 



131 



somes in this plant is being made from preparations made by Miss 

 Nest a Thomas. I have thus far been unable to find more than 

 25 or 26 chromosomes, but we are continuing the work on this 

 and other forms, of which a full account will be published later. 



Table VI. 

 Percentages of pollen types among "good" and "bad" grains. 



In addition to the types of pollen grain already mentioned, 

 I not infrequently found in O. gigas*a more nearly globular type 

 in which only two lobes were visible. No account was kept of 

 these, as they varied much and were probably non-functional. The 

 actual percentage of sterility is very difficult to determine, as 

 shrivelled grains of all sizes occur, and I have long known from 

 sections that many of the pollen mother cells also break down 

 before or during the reduction divisions. Another peculiarity worthy 

 of mention, is the way the various types of pollen grains appear 

 to be grouped. Thus in the plant having about 25 per cent, of 

 triangular grains, the latter were often found in groups on the 

 slides, frequently several together, rather than equally distributed 

 through the rest of the pollen. This would suggest that possibly, 

 their production was due to irregularities in the reduction divisions 

 of certain mother cells, by which only the haploid number of chro- 

 mosomes reached the daughter nuclei of the homotypic division. 

 The same phenomenon was observed in normal O. uiyus Italy 

 (No. I. 6). In the examination of pollen from one anther 8 trian- 

 gular grains were found together, and in another anther three 

 triangular grains were observed quite close together on the slide. 



9* 



