de Fraine .— The Seedling Structure of certain Cactaceae. 165 
Group II. 
The transition in the majority of the species examined in this group 
was of Van Tieghem’s Type 3; many variations on it occurred, of which 
the following were the most important:— 
(a) In Phyllocactns , Cerens , Pilocereus , Echinocereus , sp. of Echinopsis , 
and certain species of Echinocactus , the root is not of the typical diarch 
nature, but possesses four phloem- and two xylem-bundles ; this is due 
to the fact that the opposing groups of bast remain apart in the hypocotyl 
instead of fusing. 
(< b ) In many seedlings a cotyledonary tubercle-bundle is developed ; 
it usually merely fuses laterally with a portion of the cotyledon-strand, but 
in Echinocactus Ottonis , Echinopsis multiplex Series 2, Pilocereus exerens 
Series 1, Mamillaria multiceps , M. rhodantha (one seedling), and M. 
pentacantha } it played an essential part in the root formation, forming half 
of each of the { double ’ bundles. 
(c) In Echinocactus hexaedrophorus and Mamillaria missouriensis 
a tetrarch root is formed in a manner which recalls the similar feature in 
Opuntia maculacantha . 
In Group II, as in Group I, the transition-phenomena afford little 
assistance in the delimiting of genera within the group, for the details 
of the seedling anatomy are not always constant in a single species, and 
much less in a particular genus ; this is well seen in Echinopsis multiplex , 
Pilocereus exerens , Mamillaria rhodantha , and M. pentacantha . Further, 
although in some cases, as for example in the genus Cereus , the seedling 
anatomy of the species is a fairly constant feature, yet individuals of other 
genera bear so close a resemblance to them that it is impossible by this 
means to separate one genus from the other; in other words, it is impossible 
to define a genus in terms of its seedling structure. 
On the other hand, just as the seedlings of Group I are distinctly 
marked off by their external appearance from those of Group II, 1 so the 
transition-phenomena of the members of Group I are fairly constant and 
are sharply marked off from the type of transition characteristic of Group II. 
From this it would seem that the physiological factors which have 
ultimately resulted in the specialized seedlings of Group II have reacted 
on the seedling anatomy, producing a transition very different from that of 
the less reduced members of Group I. 
Relation of the Seedling Anatomy to some Phylogenetic 
Problems. 
One of the most interesting features in the seedling anatomy of the 
Cactaceae is the occurrence in some genera, e. g. Opuntia and Nopalea , 
of the type of transition shown by Miss Sargant to be characteristic of 
1 Compare Figs. I, 2, and 6 with Figs. 7, 8, 9, 16, and 17. 
