18 THE JOURNAL OF BOTANY 
guelen.”’ I have a specimen ex herb. Mitt., so labelled, 
“Bi paula ache Vulcan Cove, Kerguelen, Eaton, 6. 1. 75, 
which has the capsule suberect, a nd is robust, as in B. comosa, 
with the stem leaves showing the Eubreutelia base (though the 
branch leaves are purely of the Acoleos type). In Hooker’s 
barium ther 
Kerguelen Island under B. pendula, of varying size and habit, 
pe at — with pendulous capsules. Whether or no, there- 
fore, B. comosa be united with B. pendula, the latter may be 
sately, = think attributed to Kerguelen Island. 
ing to B. divaricata Mitt. According to the author this 
is couiables the larger states of B. pendula, but differs in its 
divaricated leaves, which have, when dry, a subfl accid sol 
Eubreutelia type. Mitten’s specimens appear to me to show 
simply a sige a: lax, and at the same time robust, form of 
nd andbook of the New Zealand Flora adds to 
the description age note :— Scarcely different from B. comosa and 
pendula.” This view I should quite endorse. The type of 
B. divaricata might — quite well be the type of B. comosa, 
in all obvious characters ! 
B. consimilis (H. f. & W)) must also be placed under B. pen- 
dula. The type («Hy aaa ee f. & W., Lord Auckland’s 
I, W. 85,” in herb. Wilso Hook.) is a robust plant described 
by Wilson as like a aoe form of B. elongata, with almost 
sep Cages ot very tomentose stems, with here and there a branch, 
or two or three together, but irregular, and not whorled. The 
Toat structure agrees with ordinary forms of B. pendula, especially 
of the “comosa”’ type. I have precisely peers — from New 
collected by J. C. Andersen and sent me by Dr. C soubeytes Since 
close relationship between the two. a cag goes on to place 
another plant of Hooker’s from Hermite Island also under 
= Sealehaciers tems sheet must, I iinagine, have given Hampe 
qualms, since it comprises both simple stems and stems 
with’ Wadia bceninbion , while obviously pertaining to thesame plant. 
That a plant which the author Serie considered as identical with 
his B. pendula should be referred by Hampe to B. consimilis, while 
another of his specimens shows forms intermediate between the 
two in the only distinctive character of B. consimilis, viz. the 
branching, is in itself a testimony to the slight value of that 
i 
ts 
“ 
