182 SHORT NOTES. 
staminate and pistillate flowers in the same inflorescence. Upon 
subsequent examination it was found to contain not only diclinous 
flowers, but hermaphrodite ones also. Of these, some conformed 
to the ete type of vee a flowers, but developing perfect 
stamens in place of staminodia. One of the ome = four perien 
stamens, which aes gene: ng one on each side of the usual 
staminodia; and another had three — sacoaue. arrange tie 0 
on one side of the septum and one on the other. Several of the 
and one of the fernaphastal flowers, contained tri- 
maroae traits, but in no case was the additional cell fully developed. 
- seven. A clearer idea may possibly be given by the following 
summary of the fully developed flowers, besides which there were 
several unexpanded staminate buds. The total number of spikes 
on the plant was five. The staminodia are represented by 0. 
No. of flowers. Sepals. Andreecium: Gynecium. 
Staminate......... 4 3 7 (?8+3-+1) 
Pistillate, normal 8 0+0 2) 
7 a al 3 3 0+0 (3) 
Hermaphrodite... 2 8 1+1 (2 
. oe 1 8 2+1 3) 
: 1 8 ey (2) 
1+0 
Besides the specimen just described several hundred spikes were 
examined, from the same hedgebank, and only one other abnormal 
plant was detected. This had three spikes, on which were numerous 
(about 15) staminate flowers, me three pistillate ones, but all of 
them were of the pepo The occurrence in the first 
mentioned example, of both Recep hte flowers and trimerous 
fruits, would sugge oi a reversion to an ancestral form characterized 
by peek flowers and trimerous symmetry. The abnormal fruits 
sss indicate a close affinity with a - ee gynecium of 
both Bueus and Euphorbia.—James Sau 
JAPANESE Gentians.—In examinin aceae collected by 
Mr. J. Bisset in Japan, I find that some confusion has arisen about 
e A 
y Solander ‘ Eavescota aquatica Thunb. 
the description of Grisebach’s G. Thunb bergii, and is the same as 
British Mo 8 G, japonica, a specimen of which he ties sent to the 
m. . 
the native name Harurindd, is there rightly identified as 
G. “Thanterg Griseb., and thus no ticed in Franchet & Savatier’s 
‘Enumeratio pla ntarum in Jap., vol. i., p. 393, though unfor- 
thately in vo oe p- 50, it has been réferred to G. japonica 
Maxim., as if the two species were distinct. Zollinger’s specimen 
(no. 331), distributed as G. Thunbergii Griseb., is not the plant of 
Thunberg and Grisebach, but agrees with what Maximowiez in 
