Gates, Tetraploid Mutants and Chromosome Mechanisms. 



grains "much in excess", but has not determined the proportions 

 of the various types. Stomps also mentions (1912b, p. 533) that, 

 in the triploid mutant he obtained in the F 2 from 0. biennis cru- 

 ciata X biennis, the pollen grains were frequently quadrangular. 

 In the small-flowered plant of yigas Italy above, the pollen grains 

 (see Table VI) were about 23 per cent. 3-lobed, 75 per cent. 4-lobed, 

 and 2 per cent, with more than 4 lobes. 



As a matter of fact there are different types of 4-lobed grains. 

 The typical yiyas pollen grain is quadrangular with a lobe at each 

 corner of the square , while the typical Lamarckiana grain is 

 triangular with a lobe at each corner of the triangle. But triangular 

 lobes also occur having an extra lobe on one side of the triangle. 

 It is not impossible that such grains will be found to have a 

 chromosome-number between seven and fourteen, for they will 

 contain less cytoplasm than a corresponding quadrangular grain. 

 Even if the above-suggested relation between the shape of the pollen 

 grain and the contained number of chromosomes should prove not 

 to be a constant one, it might still be frequent enough to serve 

 as a valuable indication of probable chromosome number and 

 distribution during meiosis, as a preliminary to making crosses and 

 a cytological study. That this is so will be evident from certain 

 facts here presented, and in any case the association here suggested 

 is at least true to the extent that 3-lobed grains seldom occur in 

 normal giyas and 4-lobed grains seldom in normal Lamarckiana. 



The data from my examination of pollen grains are presented 

 in the following tables. They are incomplete, as I had only a limited 

 time to devote to this work 28 ), but a number of very interesting 

 facts appear. In a more thorough study, the proportion of good 

 to bad grains should be determined in all cases, and also the number 

 of triangular and quadrangular among the bad grains, for it is not 

 impossible that there may be selective elimination of one type as 

 the pollen matures. It is to be hoped that other workers will 

 include a statistical examination of the pollen grains in the descrip- 

 tion of the plants they study. 



In the flowers of yiyas Italy very little pollen was produced, 

 and it was granular, rather than stringy as in other forms. The 

 anthers also usually failed to dehisce properly, so that in making 

 pollinations the pollen had usually to be dragged out with a pin. 

 It should perhaps be added that these pollen examinations were 

 all made within a few days of August 20 th , in the midst of the 

 flowering period. 



28) My only reason for publishing these incomplete data at this time is to 

 enable other workers to combine this method of pollen study with the observation 

 of other characters. 



XXXIII. 9 



