208 



PRINCIPLES OF PLANT-TERATOLOGY. 



eluding the one with a basal pocket, we have indubit- 

 ably to do with a case of congenital fusion of two 

 opposite leaves by their ventral midribs and basal 

 margins respectively, for the apical portion of the 

 shoot has become displaced in favour of what has now 

 become a single organ. Here, union results in either 

 the formation of a leaf bearing laminae along the mid- 

 rib, or in pitcher formation, precisely the same dual 

 aspect of the same phenomenon which becomes induced 

 (probably by the reverse process) in the saxifrage-leaf. 

 In the Marrubium it is equally certain that the terminal 

 ascidium is due to fusion of the leaves forming the 

 uppermost pair. 



In conclusion, therefore, it may be stated that (1) 

 the presence of laminas on the leaf -surf ace, and (2) 

 pitcher-formation of the entire leaf, or of portions of 

 it, may be regarded as two aspects of one and the same 

 phenomenon, viz., the imperfect manifestation in a 

 superficial position of a second leaf.* 



The numerous transitional formations exhibited, it 

 is true, not in a single plant or species, but in various 

 genera (not to be depreciated on that account), seem to 

 render this conclusion quite admissible. 



It may even be that many of the terminal pitcher- 

 leaves represent reversions to the primitive position of 

 the leaf as indicated by the phyton-theory. And if it 

 be true that the two leaves of a pair have arisen by 

 division of a single leaf, then cases like those of Budd- 

 leia and Marrubium might certainly represent partial 

 reversions. 



The normal pitcher or peltate leaves of Sarracenia, 

 Gephalotus, Nelumbium, Hydrocotyle, etc., do not appear 

 to represent at all the same phenomenon. 



# Parish suggests that the enations on the abnormal leaf of Rumex ob- 

 served by him may be due to cohesion or fusion of the midribs of two leaves, 

 the blade of the uppermost being reduced or fragmented. This is exactly 

 the idea which is here supported, and an interesting counterpart thereto. 



