THEORY OF THE STRUCTURE OF PLANTS. 395 
information, without perceiving the very parts 
which make up the ordinary fruit in this tribe, 
the two portions of the stigmas and the styles 
being kept from adhering as usual by an unusual 
development of the inner pair of carpels, which 
is usually only represented by the partition, but 
here forms an interior style with its stigmas. 
Occasional monstrosities of the wallflower, in 
which it has a four-celled fruit, and the genus 
Te tracelion which has one constantly, confirm 
this explanation. I am not even quite sure that 
the theory of the four carpels, as maintained by 
Lindley and Kunth, is identical with that which 
I am defending, as I have not here access to the 
works in which it is proposed, but my own theory applies Brown’s 
explanation of the structure of the stigmas with what seems to me 
a much more satisfactory view of the nature of the partition and the 
general syrametry of the flower, and I should be disposed to say is 
hable to no serious objection. 
The difficulty, however, which yet remains, respecting the nature of 
the peculiar arrangement of the stamens, is probably to be accounted 
much greater than that which I think has been overcome respecting 
the structure of the fruit. Dr. Lindley, in his account of the order 
in the Vegetable Kingdom, if I rightly understand his meaning, 
(which however is obscurely expressed) takes essentially the same view 
which I am disposed to favour. His words are: “ their stamens are 
‘‘arranged thus: two stand opposite each of the anterior and 
‘posterior sepals, and one opposite each of the lateral sepals ; 
“there being six stamens to four sepals, instead of either four 
“or eight as would be normal. Now in which way does this arise? 
“Tg the whorl of stamens to be considered double, one of the series 
“belonging to the sepals and one to the petals, and of these a part 
‘imperfect? Iam not aware of any such explanation having been 
“offered, nor do I know of a better one. It appears to me that the 
“outer series is incomplete by the constant abortion of the stamens 
“usually belonging to the anterior and posterior sepals, the two pairs 
“that remain belonging in fact to four petals.” The obscurity here 
arises from the expression ‘“‘ belonging to the sepals and petals,” applied 
to circles of stamens, which is unusual and not very expressive. There 
