450 ,\y,.\f\Mf„]^,Chrk& on. Relative Position ; ^„,v^„\,^, 



quently differing gives the raceme a somewhat irregular appear- 

 ance ; on one side of the style is a furrow sometimes marked 

 very faintly, but often quite distinctly, and occasionally it is 

 deepened into a fissure for some distance below the stigma ; and 

 the opposite side id quite rounded, having the aspect of a dorsal 

 surface. The position of the carpel, as taken from this character, 

 is variable, and fi*equently posterior as in Cratcegtis. 



Amygdale.«. Variable, and in Cerasus Laurocerasus more 

 frequently posterior. 



Sanguisorbace^. The lateral attachment of the style readily 

 shows the position of the carpel, which is best seen in the spe- 

 cies of Sanguisorba, and can be ascertained by making a trans- 

 verse section of a mass of flowers close to the axis, when the 

 irregular position of the style wiU become obvious. The carpel 

 is rather less frequently posterior than anterior, but in Cliffortia 

 always lateral, as elsewhere referred to. 



DAPHNACEiE. In Daphnaceae the lateral attachment of the 

 style shows the position of the carpel, the ovule being pendulous 

 from the same side ; but in Daphne itself the position of the 

 raphe must be taken as a guide, which in this instance may be 

 relied on, because in Pimelea the raphe is next to the placenta. 

 In Struthiola and Passerina the carpel is almost constantly late- 

 ral, in Sii-uthiola perhaps always so, thus approaching Proteacese. 

 Yet the relations of the parts of the flower to each other in Daph- 

 naceae and Proteacese are diflerent, as in Daphnaceae the carpel is 

 opposite a sepal, but in Proteaceae alternate with the two ante- 

 rior. In Daphnaceae the tendency to suppression of stamens is 

 on that side which is opposite to the carpel, as in Lachmea, where 

 the larger stamens and sepals are anterior while the carpel iS 

 posterior, but in Proteaceae both carpel and larger stamens are 

 anterior. In Pimelea the carpels are all posterior, the axils 

 being one-flowered ; in Daphne the axils are two-flowered, and 

 the ovary of each stands with its dorsum towards the axil ; and 

 Daias is a mixture of these, apparently from the axils being 

 three-flowered. (PI. XIV. fig. 14.) 



CiNCHONACE^. In Cinchonaceae, as also in Caprifoliaceae, 

 there is no regularity in the position of the carpels of dicarpous 

 ovaries, as they are either predominantly anterior and posterior, 

 or as frequently right and left, or neither of the two predomi- 

 nates, being often also oblique ; and this, taken in connection 

 with the fact, that in the genus Ribes alone all these variations 

 take place, tends to show that the position of the cai'pels when 

 two is a character of comparatively inferior value, unless that 

 such frequent contrariety and irregularity should make it pro- 

 bable that the position of the fertile or single carpel would be 

 variable. The raphe being taken as an index to the placenta, 



