1895.] Present Problems in the Cactacee. 137 
illary sprout is less extreme than in those which are exposed, 
but it is a question to just what this is due. Can other cases 
be found in which tubercles are partially protected or dark- 
ened, and what is then their behavior? In flat Rhipsalis spe- 
cies, and some Echinocacti, the backs of the tubercles grow 
face. In Cereus rostratus, these wings are bent backward so 
that they form hooks enabling the plant to’climb, and C. Mc- 
Donaldig shows an intermediate condition. Do the tubercles, 
or the wings from them, serve any other purposes in the 
family? 
In some Echinocacti and Mamillarie, a deep hair-filled 
groove unites the spine-bearing and the flower-bearing parts 
of the split axillary point. This groove is simply the greatly 
drawn-out sunken area in which the vegetative point is al- 
ways protected. In the highest Mamzllarie however this 
groove is absent, which is because the splitting of the point 
occurs before the sunken area, or rather its raised walls, are 
formed, and each part of the point forms its own wall. 
macrothele seems to form an intermediate condition, for the 
groove is sometimes present, and sometimes not, but in real- 
ity it has a groove of which the edges growtogether. Are 
there cases in which the groove persists as a sunken tube? 
Or are there other transitional forms? The genus Anhalo- 
nium though so small has both grooved (A. fissuratum) and 
ungrooved - (A. prismaticum) forms, a curious case of ‘‘paral- 
lel-bildung” with Mamillaria, for we cannot suppose that the 
former came off from the grooved and the latter from groove- 
less Mamillariz. 14 
15B 
gelmann and (following ia) Coulter h of t 
tubercles in Anhalonium, and especially in that section ar Tethoscans whi ob 
the latter has elevated to generic rank under the name of Lophophora Sa fo 
pont Revision of the North pyre ned species of Cactus, Anhalonium and 
Phora. ited 
g 
, he has 
sonthalonius— ey with the young “py? of the 
oro pelt plate 32). These species sy Sis berth therefore we “Eci. 
