the Embryos and Seedlings of the Cactaceae. 453 
from Cereus , while the spiral arrangement in the adults is 
a re-development of what is really a primitive character in the 
family Cactaceae. Goebel shows that several other species of 
this genus form the beginnings of four ribs : some continue 
to form them for a time, but R. commune , which when adult 
is triangular, immediately drops one and becomes triangular ; 
while in R. rhombea and R. pachyptera, which in the adults 
have very flat, thin phyllocladia, two of the ribs are re- 
presented by single polsters only, and the epicotyl from the 
first is flat in the plane of the cotyledons in precisely the 
same way as has been described and figured above for 
Phyllocactus Phyllanthus. The exactly parallel character of 
the mode of formation of the epicotyl in Phyllocactus and 
Rhipsalis accompanies a parallelism in the adults in these 
two genera. This does not in the least mean that these 
groups have had any direct genetic connexion ; probably 
both have come off from climbing Cerei but quite indepen- 
dently, Phyllocactus from triangularis- like, but Rhipsalis from 
very slender flagelliformis- like forms. The parallelism, then, 
has come about through modifications due to similar habit 
working upon a similar morphological basis. In both genera 
the ribbed character is primitive, and the flattened condition 
newly acquired, and the working back of the latter into the 
seedling until finally it quite obliterates the ribbed condition 
affords some of the best examples and illustrations of the 
true meaning of repetition of phylogeny in ontogeny yet 
adduced among plants. 
11. Genus Hariota. 
I have no data, nor does the literature contain any, upon 
the embryology of this group. 
12. Genus Pfeiffera. 
P. cereiformis , Salm-Dyck. Figure, PI. IX of Pfeiffer and Otto ; also in 
Goebel, p. 86. 
I have not myself studied this genus. The figure given 
by Goebel shows a short stout epicotyl with stout pointed 
