108 © . DEVELOPMENT OF THE FLOWER. 
the pistillum may be apparently 4 carpellary, and yet present 
no deviation from the ordinary type. It is in those cases 
alone in which the placente obviously relate to the margins 
of the leaves, that much stress is to be laid. 
Crucifere are occasionally tri-carpellary, as an instance 
now before me of Sinapis venulosa testifies. The carpellary 
leaves were unequal, one being nearly twice as large as the 
other, the stigma corresponded : they were all united by their 
lower veins. The placente were 3, opposite the confluent parts 
of the stigma. Vide A. Pl. 30.—Fig. 2. The two smaller 
stigmata were formed each as it were from a demidation of 
the bigger, but it only shews that even in a 4 carpellary 
type, the real number will be 8 carpellary, [in this instance 
alternating with fertile, were 3 barren carpella marked out 
by veins.] 
In this Sinapis when the carpels are two, the placente are 
undoubtedly attached along the apparent midrib, a line con- 
tinued up the axis of the placente will pass through the 
apex of the style. Vide Pl. 30.— Fig. 1. a, b, placenta. c, 
lateral view. ‘Fhe vessels are about 4, the two larger termi- 
nate in a clavate manner vithin the apex, the two others are 
much smaller, and — about 2 up the style. 
See also d, e, f, g. Pl. 30.— 
Regarding meum several "— be derived (vndis 
Cordia, Escholtzia, etc. 
In Pternandra czrulescens appearances are not altogether 
unfavourable to the supposition that there is a disturbance 
in the direction of the carpella, similar to that which affects 
many Boraginez, and all Labiate, in which if ever poly- 
sporous placentz be found to exist, they will, I am inclined 
to conjecture, be as they are in Pternandra, dorsal; the 
-— of the carpella being in the ordinary situation of the 
ase 
But, as appears to me suggested by the situation of the 
raphe, the placenta has a tendency to be dorsal. 
Hence such a complex explanation as the above may be 
objected to, when so easy a one as that of admitting the car- 
pellary leaves or their dorsum, to occasionally bear ovula is 

