230 BOTANICAL GAZETTE [SEPTEMBER 
ested in botany from the economic standpoint. The work of 
various botanists during this interval, and especially BrRe’s 
description of B. nobilis under the name B. excelsa, no doubt con- 
tributed to the same end. Moreover, the seeds of the two species, 
so far as can be judged from the descriptions and drawings avail- 
able, are so similar as to be distinguished with difficulty if at all. 
After making a careful study of the situation, the writer has 
become convinced that the commonly accepted view is erroneous, 
and that the Brazil nuts of commerce are derived from B. nobilis 
Miers (B. excelsa Berg) and not from B. excelsa Humb. and Bonp. - 
The reasons for this view are given below. 
t. Commercial samples of Brazil nuts contain, in larger or 
smaller numbers, opercula derived from the fruit, and the presence 
of these in itself is evidence that the nuts were derived from B. 
nobilis, since, as has been noted in the comparison, the opercula 
fall from the mature pyxidia of B. excelsa, and hence would not 
find their way into samples of nuts from that source. On the 
_ other hand, their presence among nuts from B. nobilis is perfectly 
normal and what would be expected, since in this species the 
opercula fall into the interior of the pyxidia and become mixed 
with the nuts. . Moreover, the opercula, so far as the writer has 
been able to observe, are always of the B. nobilis type, as shown 
in fig. r. They vary in form from ovoidal bodies to cones of 
varying slope, being modified apparently by the size and degree 
of persistence of the columella, as well as by the extent of the 
grinding against surrounding nuts to which they have been sub- 
jected during shipment. All, however, are provided with a dis- 
tinct apical point except where it has been broken off, in which 
case the fact is usually quite evident. It cannot be denied that the 
absence of opercula of the B. excelsa type does not preclude the 
possibility that nuts of this species may be occasionally mixed 
with those of B. nobilis, since the writer is not aware that it is 
possible to distinguish the species from the character of the nuts 
alone. : . 
2. Every pyxidium of the Brazil nut which the writer has had 
an opportunity to examine has indicated that the fruit is that of 
B. nobilis. Their main points of structure are well shown in 
