54 BOTANICAL GAZETTE [JULY 
Perhaps the most striking results of any to be found here 
from the systematic point of view will be obtained by a compar- 
ison of the first three forms. Cereus Fendleri and Echinocactus 
Wislizent are entirely dissimilar as regards character of medullary 
tissue, mucilaginous contents, firmness of interior structure, and 
especially in respect to arrangement and branching of bundles. 
The third form, Cereus giganteus, corresponds in all the above 
mentioned peculiarities, not with the other Cereus, as would be 
expected, but with the Echinocactus. So far as I can tell from 
plates,® this closer resemblance to Echinocactus is carried out 
equally well in external characters, especially in those of the fruit, 
but to some degree at least in those of the spines and areolae 
(the last comparisons taken from living material). I have too 
little systematic knowledge of this group to contest the position 
of the species in classification, and can simply point to the 
fact that the two Cerei here mentioned belong at least to two 
different evolutionary branches within the genus, from one of 
which branches Echinocactus seems to have sprung. 
The genus Cereus seems to me an especially fruitful field for 
the study of internal anatomy on a comparative basis. It is 
rather evident, from a hasty glance at the great variety of 
forms, that many different lines of development begin in this 
vicinity, lines which, though at present somewhat unraveled from 
external character, could at least be well subjected to con- 
firmatory tests upon structural grounds. This would be of 
especial value in searching for the origin of the various other 
genera, the supposed offspring of this group. 
It must not be thought, however, that I wish to give too 
great importance in classification to the internal anatomy, pre- 
ferring it to the external characters. Its principal use is con- 
firmatory. Too many internal variations spring up independently 
in several forms to make them alone sufficiently reliable. There 
must be some gradual internal metamorphosis, however, to cor- 
respond with the external. In what deep-seated organs this is 
shown, and what parts are more easily influenced by environmental 
® Engelmann’s Cact. Mex. Boundary. 
