222 BOTANICAL GAZETTE [aren 
over the seedlings of some seventy-five species in this family! 
have seen no other case of it. : 
The plants from which my seeds were taken have been grow : 
ing and flowering luxuriantly for at least four years in the 
Botanic Garden of Smith College. They agree with the char 
acters given in books for that species, but the source from which 
they came into the garden is unknown. They set seed every 
year in great abundance. When the seeds are planted, from 
many, perhaps a half, of them more than one seedling comes up, 
and there is the greatest variation in the number, size, and 
degrees of union with one another of these seedlings. This 
variety is best made evident by the figs. 7 to 5, typical cases 
drawn the natural size, and of course there are all sage” 
between. I have not tried to follow them further, but what | 
have noticed seems to show that the larger of a set crowds ot 
the others. It is now important to ascertain the morphol 
origin of this polyembryony. 7 
The ovule of Opuntia vulgaris is at first amphitropous, # 
t, at the same 
niculus makes 
a complete turn around it, so that finally = 
pylotropous condition (fig. 6). Its development If pee? 
cies of this genus, together with the development Of ©" 
species of other genera in this family, has been ies sted 
d’Hubert, and his account fits this species fairly well. e 
the funiculus are the integuments, made up ” ree 
layers of cells, and inside of these is a nucellus, which a 
absorbed, except for a small portion (at 4, AS 6) PY eee 
as the seed ripens. Finally there is the distinct a 5 
which there forms after fertilization an abundant ‘the who! 
which consists of protoplasts without cellulose walls, eas 
of course absorbed by the growing embryo. re 
In ripe seeds 8g ude Geuaill a large embryo neat 
it, with others much smaller and pressed to a = ¥ 
ripe seeds one finds such a condition as is shown ey i 
there is one larger embryo with several smaller oney 
