BOT.-VOL. I.] CAMPBELL— XAIAS AND ZAN.VICHELLLA. 



47 



angiospermous type and the details were not studied. After 

 the nuclear divisions are completed, the ^gg apparatus and 

 antipodal cells are formed in the usual manner, the latter as 

 in Naias, having definite cellulose membranes. As in the 

 latter, the two polar nuclei are much larger than an}^ of the 

 others (fig. io6), the nucleus of the ^gg cell being the next 

 in size, and both the antipodals and synergidas showing small 

 nuclei. After these structures are formed there is a very 

 decided enlargement of the embryo-sac before fertihzation 

 is effected, and during this growth it encroaches a good deal 

 upon the surrounding tissues of the nucellus, but it does not 

 occupy so large a space relatively as in Naias. Both 

 the antipodal cells and those of the ^gg apparatus increase 

 very much in size, and this is accompanied by a correspond- 

 ing enlargement of their nuclei, especially that of the ^gg 

 (fig. io8). The egg apparatus does not differ essentially 

 from that of Naias; but the cells are all somewhat larger 

 and not attached by so broad a base to the upper end of the 

 embryo-sac. 



The process of fertihzation was not followed out in detail; 

 but in one instance (fig. io8) the pollen-tube was seen en- 

 tering the embryo-sac. So far as could be judged, the pro- 

 cess was identical with that of Naias, but only one generative 

 nucleus was observed in the tube. As in Naias, the pollen- 

 tube pushes its way between the terminal cells of the nucellus 

 and through the end of the embryo-sac, being much com- 

 pressed above the point where it enters the sac. 



Several cases were noticed where there was a departure 

 from the normal type of development. The most common 

 was an increase in the number of antipodal cells from three 

 to four or five. This is obviously due to a secondary divi- 

 sion of one or two of the primary antipodals. 



A very much more marked departure from the type is that 

 shown in fig. no. Here the number of nuclei was much 

 larger than usual, although it was difficult to determine just 

 how many were present, as some of them were poorly de- 

 fined. Two, which seemed to represent the polar nuclei, 

 were distinct, and the three antipodal nuclei were pretty 



