469 
in the Stock (Matthiola incana). 
It has already been recalled that it was from a comparison of the 
Ciuciferae with Eschscholzia that Lindley was led to the view that the 
Cruciferous placentae are in reality independent carpels. It is, therefore, 
fitting that it should be from further comparison with this same genus that 
we are enabled to obtain new light on the construction of the many-ribbed 
siliqua. Lindley visualized the fertile carpel contracted to a placentiferous 
cord, but he failed to perceive the consolidation which produces the miilti- 
carpellary valve , and so missed the significance of the 10-ribbed character in 
both the Cruciferous and the Papaveraceous gynoecium—proof, as will 
appear, of the presence of a corresponding number of solid carpels in 
both families. 
7. Reduction and consolidation shown to have occurred in like manner in the 
Papaveraceae , Fumariaceae , Capparidaceae , and Resedaceae. 
Now a multicarpellary ovary in Eschscholzia would be a quite natural 
construction in view of the numerous carpels present in several other genera 
of the Papaveraceae. It is to this family, and the allied Fumariaceae, 
Capparidaceae, and Resedaceae, that we must now extend the inquiry, since 
the solid carpel explanation of the so-called commissural stigma must, we 
should suppose, apply equally to them. And this we find to be the case. 
For example, on the one hand may be cited Platystemon calif'ornicus , Benth., 
where the numerous carpels are all of the valve type with single-line sutures , 
and the stigmas, as we should then expect, superposed upon the midribs. 
On the other hand, the numerous genera with stigmas alternating with the 
valves, again in accordance with expectation, show double-line sutures ; 
hence the distribution of the stigmas is not in reality exceptional, except 
in so far as they are borne by the solid carpels only. Among such may be 
mentioned Bocconia , Chelidonimn with 2, Roemeria with 4, and Meconnpsis 
and Papaver with 4-8-12 valves. The fact that in the band form of stigma 
characteristic of Papaver and its allies each stigmatic band can be seen to 
be double, was considered by Eichler as decisive against Lindley’s view that 
the placentae in such cases represent whole carpels. But the solid carpel, 
although so contracted as to form merely a radial sheet of tissue, must 
yet be conceived to possess two edges like the expanded valve. More¬ 
over, the ends of each pair of parallel stigmatic lines are continuous at 
the periphery of the stigmatic plate so that they form a very narrow V, 
a shape we might expect if compression were accompanied by a shearing 
action forcing inwards the tissues of the carpel lamina on either side of the 
midrib. It thus becomes unnecessary to make the entirely unsupported 
assumption required on Eichler’s view that fusion occurs between every pair 
of neighbouring half-stigmas. 
To return, however, to the case of Eschscholzia. The ovary in this 
