140 A. H. Graves, 



to go into a lengthy description. For the sake of completeness, 

 however, I will review the more important points, noting a few minor 

 differences from Murbeck's results, and adding a few observations 

 of my own. 



We may regard the history of the male gametophyte as beginning 

 with the appearance of the synapsis stage in the pollen mother-cells, 

 which precedes the first reduction division and signalizes the special 

 preparations for that process. Seven different flowers chanced to be 

 fixed during this stage, and the appearance of the pollen mother- 

 cells in all was typical (PI. XII, fig. 87). The nuclei are very large, 

 but no knoblike processes are appended to them, such as Wiegand 

 (1899) figures for Potamogeton foliosus. Although some such appear- 

 ance was occasionally found, it was not sufficiently general to be 

 called typical of this stage. 



In the first reduction division I succeeded in finding several cases 

 of multipolar spindles, as Murbeck (1902, p. 7) has also reported. The 

 two reduction divisions follow one another in rapid succession, form- 

 ing a tetrad, whose members are oriented to each other after the 

 manner of the four quadrants of a somewhat elongated sphere (see 

 Murbeck, fig. 16). A similar arrangement is figured by Bornet (1864) 

 in Cymodocea, and is not uncommon in the monocotyledons in general. 

 During the two reduction divisions I was able in several cases to 

 count eight chromosomes, which is, therefore, the reduction number, 

 as Murbeck (1902) also found in Ruppia rostellata. 



In the study of the development of the pollen-grain, as in other 

 structures, I found it of advantage to use a certain definite method 

 of external measurement as an index to the stage of internal develop- 

 ment. Thus, in the case of the pollen-grain, I chose the length of 

 the grain, which, from its first formation in the tetrad to the mature 

 condition, increases from about 175 (.i to about 560 ii. Although 

 these dimensions are subject to some variation, even in the same 

 pollen-sac, yet they are fairly constant for the same period of growth. 



At the time of the formation of the tetrads the nucleus is in an 

 approximately central position. Very soon after, or while the grain 

 is still not much more than 175 n in length, the nucleus shows a 

 position nearer to one end of the grain (PL XII, fig. 88), and a central 

 zone of small starch grains has appeared. Almost immediately there- 

 after one finds the nucleus at the end of the grain, the starch grains 

 having become considerably larger and uniformly distributed (PI. XII, 

 fig. 89). A nuclear division now ensues (Fig. 90), and as a result, 

 the small lenticular cell at the end of the grain, the generative cell, 

 is formed (Fig. 92). 



