174 THE JOURNAL OF BOTANY 
writes: ‘ Pédicelles a . . . dépassant deux fois les bractées ; 
sépales ovales, . . oitié plus petits que dans le F’. Borei, & peu 
prolongés au- dessous rm leur insertion.” It may be remarked that 
in the Capreolate Fumitories found in Britain the length of the 
sepals appears to vary proportionately with that of the sat ts. 
urther mark of distinction which, though apparently noticed 
by Faueskeishty does not seem to have hitherto been defin itely 
ointed out, is that the two inner petals only are clearly tipped with 
dark purple, the upper one, which is similarly tipped, or at least as to 
its wings, in F’. Borai and other allied forms being in I’. confusa never 
more than flushed on the back with a slightly deeper tint of pink 
than that prevailing pio a the corolla. This character, which 
is practically constant in all the British and foreign specimens that 
: ave examined, can easily be seen in fairly developed flowers, 
ven when dried, and forms, I think, the readiest means of deter- 
ein the plant when the fruit is wanting. 
nother noticeable point in this species is that in good flowers 
the reflexed wings of the upper sre are broader and more con- 
tinuous round its edge than in any of the other British ‘‘Capreo- 
late.”” In this 5g seem to aes ‘is form of the more southern 
species, I’. agraria Lagasca; and taking into consideration at the 
same time the Sdlovatie on of ’the corolla and the characters of the 
fruit, I am o to tr F’. confusa as a species almost equi- 
distant between F. Borai and F. agraria. By Rouy & Foucaud 
F. confusa i is pias the rise of six forms sy a collective species, 
cedes F’, agraria “ein specifically mikesied from it. In Britain, 
where the forms between F. Borai and F.c confusa have not been 
recorded, the two ants look so widely different that I certainly 
can only consider “oom as distinct species, although it is just 
would connect the two. A more correct view, however, seems to 
be that of Haihtagectit; “who considered F’. confusa to be a variety 
of F’. Gussonii Boiss., and specifically distinct from F’. Borei. The 
type of F. Gussonii, which is unknown in Britain, was thought by 
Jordan (Pugillus, p. 5) to be intermediate between his I. confusa 
an orei, and nearer the latter. It is undoubtedly more closely 
related, as Haussknecht supposed, to F’. confusa, which it resembles 
n the shape of the corolla and the rugosity of the fruit. The 
cients of the flower and the shape of the fruit, however, are 
nearer to F’. Bor@i, and I should hesitate to specifically unite it 
with either of re And, if united with F’. confusa, I think that 
us Jordan’s name, which i is ignored by Haussknecht a s being 
used to describe a form only, would stand for the species, it being 
older by a year than — of Boissier. 
- confusa seems to be somewhat er scattered over the 
greater part of Great Britain, but I know of only one habitat for it 
in Ireland. In the Channel Islands it is common. The British 
localities from which I have seen authentic specimens are—Pen- 
zance; the Lizard; Ilfracombe; Mudeford, Hants; Uckfield, 
