AND STERCULIACES. 121 
any trace of a transverse separation or contraction; and that we 
have thus the one-celled anthers of Malvacez, thereby doing away 
with the chief distinctive character of the two orders. I have, 
however, in vain searched for any species, either of Ayenia or 
Guazuma, showing any intermediate state between the ordinary 
anthers of the two genera; and in all other respects the two 
genera, though not inappropriately following each other in the 
linear series, have wide constitutional differences, in the calyx, the 
petals, the number of ovules, the styles, and, above all, in the 
embryo. On the other hand, in the great majority of Buettnerias 
the anthers have two parallel cells, which have as completely the 
appearance of belonging to one anther only as those of Rulingia ; 
and when the third cell is present, might it not be considered as 
the half of an anther belonging to the adjoining staminal leaf, 
under the theory above suggested in explanation of the andreecium 
of Sterculiacew ? The three-celled anthers of Ayenia would then 
be explained as consisting of one complete two-celled, and one 
dimidiate anther. 
Cybiostigma of Turczaninow is founded on the common Ayenia 
magna, L., and a closely allied species, and has nothing to sepa- 
rate it generically from the other species. 
Ruina, R. Br. 
Achilleopsis, Turczan. (Walp. Ann. ii. 165), does not appear to 
be sufficiently distinct from Rulingia. I do not find the stamens 
quite free from the base, although much more shortly united than 
in most Rulingias. 
LASIOPETALEÆ. 
In this tribe, entirely Australian, where the petals are wanting 
or rudimentary, the andrecium may be formed on a somewhat 
different principle from that of the other Sterculiaceæ. The sta- 
mens of the outer whorl opposite the sepals are reduced to barren 
filaments or, like the petals, entirely deficient ; and the five anther- 
bearing stamens, alternate with the sepals as in other pentandrous 
Sterculiaceæ, may, however, really belong to the inner staminal 
whorl. 
The genera of this tribe have probably been too much multi- 
plied on characters of very little importance. Even the difference 
in the dehiscence of the anthers, by short pores or long slits, is not 
always clearly marked; and although we have still availed our- 
selves of it as distinguishing considerable groups, these are not so 
