BOT— Vol. I.] CAMPBELL-NAIAS AND ZANNICHELLIA. 49 



constituting the embryo proper is more elongated in the early 

 phases of development. 



A more marked difference is seen, however, when we 

 study the relation of the members of the embryo to the 

 primary segments. In Mains, as in Alisma, the whole of 

 the terminal segment develops into the cotyledon, while the 

 stem-apex originates from the second segment. In Zcm- 

 nichellia, on the other hand, both stem-apex and cotyledon 

 arise from the terminal segment, the separation of the two 

 being determined by the first vertical wall in that segment; 

 Zannichellia, therefore, forms another exception to the rule 

 in Monocotyledons, that the stem-apex is of lateral origin. 

 Solms-Laubach (1878) has shown a similar terminal origin 

 for the stem in several members of the Commelynaceae and 

 Dioscoreacese; and Hegelmaier's (1874, ^&- 33) figures of 

 Sfarganiiim and Pistia indicate that in these forms also the 

 stem-apex is derived from the terminal segment of the em- 

 bryo; but he makes no statement as to this. Solms' figures 

 of Heterachtia, one of the Commelynaceae, are strikingly 

 similar to corresponding stages in ZannicheUia, except for 

 the almost complete absence of the suspensor. 



The whole of the central region of the embryo of Zan- 

 nicheUia goes to form the root, the lower ones contributing 

 to the root-cap, which is here well developed, and the sec- 

 ondary suspensor cells, of which there may be as many as 

 four between the root-cap and the primary enlarged sus- 

 pensor cell, which finally attains a very large size. As in 

 Naias, the suspensor cell has densely granular contents and 

 a very large nucleus. 



The separation of the tissue-systems of the embryo is not 

 determined so early in ZannicheUia as in Naias, and the 

 young embryo (fig. 117) shows a less symmetrical arrange- 

 ment of the cells, the limits between the segments not being 

 always perfectly definite. The formation of the dermatogen 

 is not determined by the first pericHnal walls in the seg- 

 ments, but is secondary; the first longitudinal walls cut 

 off the plerome, and this is then followed by a separation 

 of periblem and dermatogen in the peripheral cells. 



