456 Ganong. — The Comparative Morphology of 
other forms of this genus, show the earlier formed spines 
feathered with fine hairs, which are prolongations of epidermal 
cells of the spines. This feathering does not occur in Opuntia , 
and not commonly in Cereus , or in any of the groups closely 
related to them, though it occurs also in Mamillaria and its 
relatives. In E. Beguin , in the young seedlings, it is par- 
ticularly well marked, and it is this character intensified 
which gives us the plumose spines of Mamillaria Bocasana. 
Possibly when well developed it has an ecological use, but 
otherwise I think it may represent one of those lines of 
spontaneous variation which have no regard to utility. The 
subsequent and adult development of the species in this 
genus is simple, and very closely like that of the columnar Cerei. 
Schumann, in his monograph, has placed E. Scheerii in 
this division of Ancistroc actus , but its embryology places 
it with the division Thelocactus , under which I shall consider 
it, and where indeed Schumann, in his Cactaceae in Engler 
and Prantl, originally placed it. 
Another division of much importance is Schumann’s Cephalo- 
cactus , in which I have studied E. ingens (Fig. 37) and 
horizontalonius (Fig. 38). These two species show embryos 
very like one another, but very different from those of the 
division last studied. In both the hypocotyl is very stout 
and somewhat elongated, and in both it shows a slight 
flattening, which is perhaps the result of their position in the 
seed. But the chief peculiarity is in the cotyledons, which, 
so far from being pointed, are rounded and scarcely raised 
out of the hypocotyl, and this character is more extreme 
in E. horizontalonius than in E. ingens. In fact the latter 
species, except for its large size, suggests Anhalonium , 
presently to be considered. The early development of the 
epicotyl in E. horizontalonius is shown in Fig. 38 d, where 
the characteristic rounded tubercles are of very unusual form, 
and the early development of E. ingens is much like it, 
though the tubercles are less rounded and pass at once 
to ribs. An excellent figure of a young plant at a much 
later stage is given by Engelmann (PL XXXII). 
