38 MISSOURI BOTANICAL GARDEN, 
of all of Mr. Brandegee’s species. No small profit has been 
derived further from study of small plants collected this 
year by Dr. Rose, as part of his very large gathering of liv- 
ing succulents, and for which I am equally indebted to the 
officers of the National Museum and of the New York Botan- 
ical Garden. The study of all of this material has proved | 
unusually interesting, and my sincere thanks are recorded 
for the privilege of making it, as well as for access to and 
use of the very full field notes and the photographs that ac- 
company many of the specimens. Dr. Rose has also done 
me the favor of revising the manuscript, with the specimens 
in hand. 
Lower California is essentially a desert, but it is relieved 
by a broken chain of mountains reaching an altitude of over 
ten thousand feet on part of which rainfall is sufficiently 
abundant and the prevalent high temperature sufficiently 
mitigated to permit forest development. The southern cape 
region is said to be blessed with summer showers, rarely ex- 
perienced above. Mr. Brandegee has indicated a change 
in the character of the vegetation to the north and: south 
of the divide between Calmalli and Cardon Grande, that 
of the north being more and that of the south less like the 
Californian flora. Mr. Watson long since called attention 
to the fact that the flora of the peninsula is Mexican rather 
than Californian in its general characters. 
Agave is essentially a desert genus of plants, centering on 
the dry tableland of Mexico and rarely dropping into the 
rainy piedmont. Nevertheless, in some of its forms it oc- 
curs through the entire chain of West Indian islands, which 
are essentially arid in a part of their coast region at least, 
and its outlying representatives are found as far south as the 
upper Andean region, and as far north as Utah, in the United 
States, while its eastern and western limits are the oceans. 
I have no knowledge of an agave native to an oceanic is- 
land, and water seems always to offer an effectual barrier 
to distribution in this genus. Even on the mainland, no 
_ species is known to have a range of very large extent with 
exception of the common lechuguilla, which, in fact, is rep- 
