348 BOTANICAL GAZETTE [NOVEMBER 
of Natural Sciences, and a life-long member of it. To him alsoisdue _ 
the credit of establishing the Buffalo Botanic Garden, in which he was 
particularly interested. 
His library was a notable one, as he was a collector of rare dis- _ 
crimination, and the works on botany were many and valuable. In 
addition to the more pretentious volumes, a large collection of 
pamphlets was accumulated, among which appear a very large number of 
local catalogues. These books of botanical interest, as well as his col- 
lections of living plants and herbarium specimens, he was preparing 
to transfer to the Botanic Garden when he was stricken down. 
Mr. Day had for years been a member of the Park Commission of 
his city, and drew the act which created the Park department. In this 
act, drawn thirty years ago, he made provision for both botanical and 
zoological collections, both of which, after years of waiting, he saw 
established. 
His loss will be mourned by the many botanists of his acquaintance, 
as well as by his fellow-citizens, by whom he was held in high esteem. 
—Joun F. CoweE.., Buffalo Botanic Garden. 
OBSERVATIONS ON THE ROOT SYSTEM OF CERTAIN 
CACTACE:. 
, l 
Tuose who make botanical trips into arid regions do s0 generally 
with the one idea of collecting material to be worked up at leisure 
laboratories or herbaria, or to be deposited in botanic gardens. Theit 
stay, as a rule, is closely limited, and the necessity of covering ae 
amount of ground brings with it a tendency to pass over those de 
which take considerable time. The root system of desert pl 
as regards structure and distribution of roots, is one 0 
which has perforce been to a great extent neglected, our 
generally speaking, being confined to the examination of th 
of underground growth collected with herbarium specimens. 
fore it seemed to me worth while during my stay in TucsoM, i ani 
in the midst of the great cactus plains, to make a rather carefu : 
of the roots of certain large forms which could scarcely be pres 
am toto, 
The root systems of Cactaceze, in general, are so! 
than would be expected. The distribution, however, 1 
way to make good the deficiency in size and length. 
e amount 
