JUAP. V. CONCLUDING BEMABKS. 195 



follow the whole process of transformation. The termi- 

 nal leaflets of the leaf-climbing Fumaria officinalis are 

 not smaller than the other leaflets ; those of the leaf- 

 climbing Adlumia cirrhosa are greatly reduced ; those 

 of Gorydalis davieulata (a plant which may indifferently 

 be called a leaf-climber or a tendril-bearer) are either 

 reduced to microscopical dimensions or have their 

 blades wholly aborted, so that this plant is actually in 

 a state of transition ; and, finally, in the Dicentm the 

 tendrils are perfectly characterized. If, therefore, we 

 could behold at the same time all the progenitors of 

 Bicentra, we should almost certainly see a series like 

 that now exhibited by the above-named three genera, 

 In Tropaeolum tricolorum we have another kind of 

 passage ; for the leaves which are first formed on the 

 young stems are entirely destitilte of laminae, and 

 must be called tendrils, whilst the later formed leaves 

 have well-developed laminae. In all cases the acquire- 

 ment of sensitiveness by the mid-ribs of the leaves 

 appears to stand in some close relation with the abor- 

 tion of their laminae or blades. 



On the view here given, leaf-climbers were primor^ 

 dially twiners, and tendril-bearers (when formed of 

 modified leaves) were primordially leaf-climbers. The 

 latter, therefore, are intermediate in nature between 

 twiners and tendrU-bearers, and ought to be related to 

 both. This is the case : thus the several leaf-climbing 

 species of the Antirrhinese, of Solanimi, Cocculus, and 

 Grloiiosa, have within the same family and even within 

 the same genus, relatives which are twiners. In the. 



