COITDITIOU OF YEBBASCUM NIGBUM. 457 



bases, while between them was an axis covered with ball-shaped 

 hairy flower-buds terminated above by a minute tuft of leaves 

 (PL XVII. figs. 10, 11, & 12). 



Of these kinds of flowers examined, there were some in which 

 the stamens were of the form represented in PI. XVII. fig. 11. 

 In others there was no trace of them at all. 



The sepals in all cases were free and linear in form. The corolla 

 was gamopetalous, as in the normal state, though varying in size, 

 being usually smaller in the upper part of the inflorescence. 

 Hence they were unlike those of the Lysimacliia described by 

 Baillon, in which neither the sepals nor petals were coherent. 

 The present monstrosity, however, agrees with that plant in the 

 fact that, in both, the stamens, when present, are abortive and free 

 from the corolla, being inserted on the axis below the malformed 

 carpels. The anthers of the Verbascum were of two sorts, some 

 reniform, others minute and oblong (compare PI. XVI. fig. 1 

 and PI. XVII. fig. 3) ,• but in all cases apparently quite 

 abortive. 



"With reference to the difference between the upper and lower 

 parts, Baillon says that he does not know why the proliferous 

 axes are more prolonged into shrubby processes on the higher 

 part than in the lower. There seems, however, a very simple 

 cause, namely the tendency of the vegetative extremities of plants 

 to take the lead before the laterals ; consequently the axis of the 

 flowers in the upper part have shot out while those lower down 

 have been delayed in doing so, till the arrested stage became 

 fixed and the buds could not develop at all, as shown in the 

 flowers on Plate XVI. 



Another difference between the present plant and Baillon's 

 Lysimacliia consists in the fact that in the latter the placenta is 

 normally free and central, so that the proliferous axis arises, as it 

 were from a free tube ; but in the Verbascum the axis is adherent 

 to the two carpellary leaves, this adhesion causing two deep 

 longitudinal furrows to be seen on the exterior surface and on 

 opposite sides of the ovary. 



