TX Leaves 183 



in those such as R. hederacea, which seem most nearly 

 related to the white water buttercup. For as it is likely 

 that the aquatic forms are derived from land forms, the 

 established type of venation will to a certain extent limit 

 the manner in which the leaves are cut up. The established 

 structure of the original type must be taken into account in 

 interpreting the derived variation. 



Secondly, we should have to grow buttercups of similar 

 origin (reared from seed or cuttings from one plant) in 

 different conditions. One set should be placed in ordinary 

 pond water, and another set in water with an unusually 

 large quantity of carbonic acid gas. One set should be 

 grown in still water, another set in the same water flowing 

 rapidly, and so on. Only by experiments 1 more precise 

 than those which have as yet been made, can the problem 

 be satisfactorily solved. 



Meantime we can only express our opinion (i) that the 

 division of the submerged leaves is determined by the 

 established type of venation seen in related species and 

 in the surface leaves ; (2) that deficient nutrition, due to 

 insufficient supplies of carbonic acid gas, may cause the 

 tissue between the veins to remain undeveloped or to die 

 away; (3) that currents of water exerting pressure upon 

 the leaves may help to develop the filiform state ; and 

 (4) that the form of the submerged leaves may thus, in a 

 natural and necessary way, have become well adapted to 

 the conditions. 



Leaves adapted for Special Functions. Before we 

 pass from the leaf we must notice some of the special 

 functions which exceptional leaves may fulfil. Some of 

 these we have discussed already, for the pitchers of pitcher 

 plants, the bladders of Utricularia, the fly-trap of Dionasa 



1 Cf. H. de Varigny, Experimental Evolution (Nature Series), 

 Lond. 1892. 



