AND STERCULIACE.E. 117 
Mzrocura, Linn. 
We follow A. Gray and others in referring to this genus, as 
sections, not only Riedleia, Vent., including Mougeotia, H.B. et K., 
Lochemia, Arn., and Altheria, Thou., but also Visenia, Houtt., to 
which belong Alewrodendron, Reinw., and Glossospermum, Wall. 
As to Physodium, Presl, from the fragmentary specimens of two 
Mexican species in the Hookerian herbarium, it appears to have a 
very different habit, and perhaps the very large Physalis-like ma- 
ture calyx may suffice to keep it distinct, but it requires to be 
better known before the point can be decided. The recently pro- 
posed genus dnamorpha, Karst. et Tri., and the two species of 
Physocodon, Turezan., are all founded on the Melochia (Mougeotia) 
inflata. 
Guossostemon, Desf. 
This is a Persian plant, not very common in our herbaria, but 
interesting in the structure of its andreecium, as affording per- 
haps some clue to the explanation of the anomalies observed in 
the homology of the flowers of Sterculiacee with respect to the 
position of the stamens. We have seen that the staminal column 
in this Order is usually divided into a definite number of barren 
or antheriferous teeth or filaments, which is usually some multiple 
of the sepals or petals. In a very few genera (e. g. Astiria, and 
probably Assonia) these filaments, four times as many as the 
sepals, all bear anthers, and are all apparently in a single row and 
equidistant—an occasional occurrence in different Orders of various 
staminal homologies: but in the majority of Sterculiacee, the five 
innermost divisions, always opposite the sepals, are without anthers, 
and take the name of teeth or staminodia, aecording to their degree 
of development ; and between them, and consequently more or less 
alternate with the sepals or opposite the petals, are 1, 2, or more 
sessile or stipitate anthers, always turned outwards and lying out- 
side the staminodia in the bud. In a few genera (e. g. Waltheria 
and some Melochias) the staminodia almost or even totally disap- 
pear, and there remain only 5 stamens, connate in a ring or cup 
àt the base, but each tapering into one anther-bearing filament 
Opposite the petal, instead of alternating with it as is usually the 
case where the stamens and petals are isomerous. It has been 
attempted, especially by A. Gray, to explain this anomaly as a case 
of dédoublement ; that is to say, by supposing that each stamen 
with its corresponding petal arises from the splitting of one homo- 
ogical leaf; the whole flower consisting of three whorls only, of 
