Box.— Vol. I.] CAMPBELL— NAIAS AND ZANNICLIELLIA. 



45 



stage just described a somewhat older one (fig. 95), it is 

 seen that the growth on the side of the apex which is turned 

 away from the leaf is stronger than it is on the inner side, 

 so that the apex itself is beginning to be pushed over toward 

 the inner face of the carpellary leaf. This continues until 

 there is a ridge formed on the outer side of the shoot-apex, 

 which is continuous with the margin of the growing carpel, 

 and the two together constitute a tubular envelope extend- 

 ing above the shoot-apex and completely enclosing it (figs. 

 96, 99). The latter has been now forced over to a hori- 

 zontal position by the strong growth on one side and forms 

 a hemispherical projection, the young ovule, which has been 

 erroneously supposed to be an outgrowth of the carpel itself. 

 It is perfectly plain, however, from a study of its develop- 

 ment, that the ovule is really the apex of the floral shoot, in 

 which respect it may be compared directly to that of 

 Naias. It is quite probable that a similar condition may be 

 found in other low Monocotyledons. Thus in S^arganium 

 the description and figures given by Dietz (1887, P- 45> PI- 

 III, fig. 15) would point to a similar origin for the ovule, 

 although he states that the ovule arises from the wall of the 

 circular carpellary leaf, which, however, is described as 

 being very much higher on the side opposite the ovule. 



The development of the ovule itself follows the normal 

 type. The archesporial cell is very early evident as a large 

 hypodermal cell (fig. 97), and can be detected before there is 

 any appearance of the ovular integument, and while the ovule 

 still has the form of a very broad but shallow prominence. 

 From the primary archesporial cell there is cut off a single 

 tapetal cell (fig. 98), and at about the same time the first 

 integument becomes visible as a low ridge extending around 

 the base of the ovule. The nucellus continues to grow, 

 but remains relatively broad, and the apex is almost flat. 

 The tapetal cell divides into two (figs. 100, loi), and the 

 inner archesporial cell rapidly grows and divides so as to 

 form, in all the specimens seen, a row of three cells, dis- 

 tinguished as usual from the neighboring cells by their size 

 and contents. The lower cell is smaller than the others, 



