No. 472] POLLEX Ch'MX VARIATION 



275 



In a recent short review of Lopriori's paper, C. J. Chamberlain 

 also rejects that writer's conclusion that the numerous nuclei are 

 male nuclei, but he gives no reasons therefor except "judging 

 by his figures." 



The structure in Fig. 16 (PI. 1) requires little discussion. It 

 represents one of the somewhat irregular variations in a double 

 grain, but the portion marked a seems to consist of a mass of pro- 

 toplasm completely cut off from the rest, and contains no trace of 

 a nucleus. In order to determine the frequency with wliicli the 

 variation of the double pollen grain occurred, a count was under- 

 taken. Fresh drops of clove oil containing the jjollen grains were 

 mounted and all the grains on the slide were systematically counted. 

 A separate list was made of all the grains that were double, both 

 those like Fig. 10 and like Fig. 8 and also the ones intermediate 

 between these two. The total number counted was 1120. Of 

 these there were 27, or 2.4 % which showed thv doubU" character 

 more or less plainly. This was a much higlit-r pcrcciitaixc than 

 was expected, but there is no reason to think a mistake was made 

 in the counting. Indeed it is more probable that some of the 

 double grains may have been overlooked. If any of them lay in 

 the position with the dorsal or ventral side toward the observer 

 they would almost certainly have been overlooked since in that 

 position their double nature is very difficult to recognize. The 

 proportion among those that were counted did not impress the 

 observer as being any greater than the proportion among the 

 hundreds that were not counted but were looked over in making 

 the study of the variations found. 



As a result of the study of the type of variation shown in Figs. 

 10-14, the following ad(iitional conclusions may he <hawn:- 



1. The three or four cells lying along the dorsal side of the pol- 



time thought a possibility. 



rated by a division wall into two very u!i(>(|ual ();)i-ti.)ns. eadi of 

 which may form a more or less typical antheri.linni. 



The cells of the larger part of Mich a pi)llcn -rain are similar 

 to the cells in a ty]>ical single pollen grain, and caM'. were oh.erved 

 in which the number of protliailial cells visible was 2. 1, anil 0. 



