452 ^^^^^"^ Mrs. Cla^^ oh mitvePositihir,'' ^^^* 



internally) ; and the position of the larger of the two carpels is 

 uniform, being always on the opposite side of the flower to the 

 spurred petal. The two carpels will be all anterior and posterior, 

 or all right and left, according as the irregularity of the corolla, 

 or the axis is regarded, the spurred petal being lateral, although 

 by twisting of the peduncle becoming nearly posterior. As in 

 Ranunculacepe the spurred petal is posterior, the fertile carpel 

 in F. officinalis may so far be considered always anterior. 



Berberide^. In Epimedium and Berberis the carpel is an- 

 terior, obliquely anterior or lateral, but a careful examination of 

 several species of the latter genus failed to show any carpels pos- 

 terior ; in Nandina, however, it is occasionally but not frequently 

 posterior. 



BYTTNERiACEiE. In Wttltheria the anterior position predo- 

 minates, the carpel being rarely, if at all, posterior, which shows 

 further the connexion between Malvaceae and Phytolaccacese, 

 where in Rivina the carpel is always anterior. 



Trop^olace^. From the irregularity of the flower, the in- 

 equality of the stigmas and the oblique direction of the style in 

 Magallana, it may be expected that the position of the two car- 

 pels is uniform, and also that the carpel corresponding with the 

 elongated lobe of the stigma is the fertile one ; if so, the two 

 carpels are always anterior and posterior, as in the figure of the 

 fruit in Cavan. Ic. iv. t. 374, the spur of the calyx is represented 

 as being on the sutural side of the remaining cell; and this 

 shows also that the position of the fertile cell is most probably 

 uniform, being always anterior, as the spur of the calyx is pos- 

 terior. 



Amyride^. In Amyris toxifera the position of the carpel is 

 variable, being frequently anterior, but only occasionally poste- 

 rior ; and the same variation I believe occurs also in Aurantiacese, 

 and these two families are, more than any others that I am aware 

 of, exceptions to the Proterocarpous character of this division. 



ScROPHULARiACE^. In Pediculuris palustris the anterior 

 carpel is always larger in diameter and considerably higher than 

 the postei'ior, and the anterior column of the style is also thicker. 

 This is also the structure of Mendozia as described and figured 

 by Martins in ' Plant. Brasil.,' and which has been before alluded 

 to as having the fertile cell anterior (PI. XIV. figs. 16, 17, 18, 19, 

 20 &21). It should be observed also that xa Mendozia the pos- 

 terior cell is usually barren even at the time of flowering. 



Verbenace^. In Lippia dulcis and Lantana crocea and albida 

 (the only species examined) the stigmas are anterior and poste- 

 rior, and the ovary two-celled, but with the cells right and left, 

 the former having also the anterior portion of the stigma much 

 larger. The placentation is the same as in the two anterior lobes 



