450 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM. 
“After the usual sign of duration, and the time of flowering, each species 
has been faithfully marked whether I have seen it myself in a living state 
(v. v.), or only in a dried specimen (v. s.); in the latter case the name of 
the herbarium I had it from is generally mentioned. Those I have adopted 
without seeing them myself in either state, and which are but few, I have 
marked (+).” (p. xxi.) 
In the descriptions of the two species the symbol “yy. v.” does not 
occur with either, yet in that of Berberis aquifolium Pursh says, 
“berries dark purple, eatable,” and “Bacca 3-locularis, 3-sperma, 
abortione interdum monosperma.” The last phrase is apparently ex- 
plained in the note under B. nervosa, where Pursh says that “the 
statement I have given of it was taken from a single and imperfect 
berry.” This statement is that above quoted, as in the description 
of B. nervosa only the leaves are described, with the explanation, 
“The specific difference excluded, the description of the preceding 
species [i. e. B. aquifolium] is applicable in every other respect.” 
The description of the fruit may have been taken from Menzies’ 
specimen or more likely from the Nepaul species collected by Buch- 
anan and referred to by Pursh in the note under B. nervosa. The 
origin of the information “berries dark-purple, eatable” is wholly 
obscure. 
On the basis of the series of specimens collected at the Great 
Rapids and a reexamination of the type specimen, the writer believes 
that all of the type sheet was in reality collected by Lewis and that 
it all represents the shiny-leafed species. The flowering shoot has 
leaves less lucid than usual, but certainly too shiny to associate it 
with typical Berberis repens Lindl. 
There seems no other basis than Lindley’s statement, already dis- 
cussed, that Pursh’s plate was drawn from a specimen of Menzies. 
Lindley labored under the idea that Lewis’s Great Rapids specimen 
was the same at least in part as that of which he brought back seeds, 
the progeny of which seed was the basis of Berberis repens Lindl. 
Lindley in all probability had not seen Pursh’s type. The conclu- 
sions of Torrey and Gray are largely based on Lindley’s statements, 
as it is not clear that either of these botanists had examined Pursh’s 
type specimen before the time they published their comments; indeed 
they quote “ex Lindl.” 
In the light of the data presented above, the specific name aqui- 
folewm should be retained for the shiny-leafed, usually tall species 
of the northwest coast, which extends into the interior of Washington 
as far as Spokane, while the name repens should remain associated 
with the smaller plant of Montana, which in various forms ranges 
over much of the area east of the Cascade Mountains from British 
Columbia and Montana to California and New Mexico. A very ex- 
tensive bibliography of these Berberis species up to 1878 may be 
found in Watson’s Bibliographical Index (pp. 34-35). 
