the Embryos and Seedlings of the Cactaceae . 433 
The development of the earliest stages of the epicotyl is 
much the same throughout the family. The cotyledons, how- 
ever, behave differently in the different genera. In Pereskia 
they apparently fall off; in Opuntia and some of the broad- 
cotyledoned Cerei, and in Phyllocactus , they wither up without 
falling ; in the other genera they are persistent, but in the 
further growth in diameter of the hypocotyl they broaden 
at the base and gradually become merged into the stem, 
a curious case of ontogenetic metamorphosis. In the develop- 
ment of the epicotyl, first of all two leaves appear from the 
growing-point at right angles to the cotyledons ; then another 
pair nearly at right angles to these ; then there comes either 
a fifth in position to begin a -J spiral, which later may become 
a I, as is common in Opuntia and Mamillaria ; or else the 
second pair is at right angles to the first and the third pair 
over the first, thus forming four rows, the foundation for four 
ribs, as is common in the ribbed forms. These ribs may 
then diminish in number by dropping one or more, or may 
increase by the splitting of one or more. In the further 
development of the epicotyl, the peculiar characteristics of 
the adults appear at diverse stages, a subject which will be 
treated later in the special part of this paper. 
The roots in both embryos and seedlings are very simple, 
slender, and sharply distinct from the hypocotyl. They rarely 
become succulent. In some species, as Echinocactus Wisli- 
zeni , the epicotyl develops very rapidly and the root very 
slowly, so that the seedling has a curious three-storied appear- 
ance — a swollen globular epicotyl above, a cylindrical smaller 
hypocotyl below that, and the very slender root below all. 
Cellular Anatomy. 
The cellular anatomy of this entire family is of great interest 
and is still imperfectly known. I have naturally made some 
observations upon this subject in the embryos and seedlings, 
but shall not attempt to treat of it here, for the reason that 
it is being thoroughly studied by one of my students whose 
results will before very long appear, and whose paper on the 
