70 MR. BENTHAM’S SYNOPSIS OF LEGNOTIDEJE, 
Singapore, where, amongst other places, Dr. Wallich gathered the 
original species. 
"There are some slight inaccuracies in Blume's dissections of 
Gynotroches. The disk is very incorrectly represented, and in 
fig. N+, one stamen is inserted outside, and another inside of 
it; in fact, the filaments are united at their very base into a 
narrow ring just outside the crenulated margin of the very short 
disk, whilst the petals are inserted outside the staminal ring. In 
figs. 1+ and N+, the seeds are shown as attached by their lower 
instead of their upper end. So in fig. O+, the hilum should be 
represented as near the upper broad extremity, and the upper 
slender end of the embryo should be the entire radicle, the lower 
and thicker end being split into the two cotyledons. 
CROSSOSTYLES. 
Forster’s detailed description of his Crossostyles biflora, pub- 
lished by Guillemin in his ‘ Zephyritis Taitensis’ (Ann. Se. Nat. 
Par. sér. 2. vii. p. 353), enabled Asa Gray to recognize it in 
the plant which he has figured and described in detail in the 
Botany of the American Exploring Expedition. Unfortunately 
these specimens were in some respects imperfect, and the seed has 
not yet been described. I have, however, little hesitation in 
referring to the same genus, as a second species, some fruiting 
specimens collected on the Feejee Islands by Dr. Harvey, differing 
from the C. biflora chiefly in the smaller narrow leaves and in the 
reduced number of parts of the pistil. The remains of the stamens 
show that their insertion and numbers were about the same as in 
C. biflora; but there is nothing to indicate the presence or absence of 
the nectaries of Foster, or sterile stamens of A. Gray, so peculiarly 
characteristic of this genus. It is not stated by either, whether 
they are precisely in the same ring as the stamens, or а little 
withinside of them. If the latter supposition be correct, they 
probably represent, not sterile stamens, but the teeth of the disk, 
which are more or less prominent within the stamens in several 
Legnotidee. If really alternating with the stamens in the same 
vertici, there is nothing in the slightest degree analogous to 
them in any other genus. The seed of Crossostyles Harveyi bas 
an embryo much nearer to that of Cassipourea than that of Gyno- 
troches. 
` ANSTRUTHERIA. 
This genus was originally proposed by Dupetit Thouars in his 
* Genera Nova Madagascariensia,’ under the name of Richeia. It 
