BoT— Vol. I.] CANNON— A VENA FA TU A. 345 



the basal proembryo cell, and is therefore not homologous 

 to the cells above. For the sake of clearness in tracing its 

 history in the developing embryo this last is designated as 

 segment three. 



The growth of the embryo is such as to obliterate these 

 primary divisions, but a comparative study shows that prob- 

 ably the cotyledon and stem-apex are derived from the first 

 segment, the hypocotyl and the initials which give rise to 

 the root-cap, the cortex, and the dermatogen of the root 

 from the second, and the coleorhiza from segment III. 

 This is practically the conclusion of Norner (1. c). As 

 the boundary between the first and second segments par- 

 ticularly, in the advanced embryo, however, is not at all 

 clearly defined, it was not possible for me to determine 

 whether the stem-apex arose from the first segment, as 

 in Lilcea (Campbell) and others, or from the second, as in 

 Alisma (Schaffner, 1897). 



In connection with the origin of the members of the 

 embryo it is interesting to note that the position of the root 

 and of the stem-apex in the embryo are the same with 

 relation to the vertical as in the mature plant (which of 

 course is the case in anatropous ovules in general). The 

 reaction of the embryo to the influence of gravitation can 

 here be traced to the one-celled stage. Experiments might 

 be devised which would show not only the immediate 

 influence of gravity on the embryo, but also which would 

 give some grounds on which to base a plausible hypothesis 

 to account for the anatropous condition of the ovule in this 

 and in other plants where this condition occurs, and also its 

 relation to the embryo. 



Continuing the history of the development of the embryo, 

 the next wall to be formed after the suspensor is cut off, 

 the fifth wall in the embryo, is a vertical one in the second 

 segment. Up to this time in the growth of the embryo the 

 succession in cell formation is quite regular, but after this 

 it varies considerably. The next wall to be formed is 

 generally a vertical one in segment III, making a seven- 

 celled embryo (fig. 33). This segment does not, however, 



( 2 ) April 17, 1900. 



