956 



A TEXTBOOK OF THEORETICAL BOTANY 



plants on the basis of leaf size, as we shall see in Volume IV, and has shown 

 that the predominance of leaves belonging to one or other of these size-classes 

 in a given habitat is indeed an expression of the prevailing conditions. Never- 

 theless it must not be overlooked that the production of a ver\^ large number of 

 small leaves may, to a great extent, compensate for their smaller size, so that 

 the total area of foliage possessed by such a plant may be scarcely inferior to 

 that of a plant with a smaller number of larger leaves. Thus any advantage 

 in regard to a lower rate of water loss, assumed to be gained by the smaller 

 size of leaf, may be neutralized by their greater numbers. 



There is, indeed, a principle of compensation to be observed in this con- 

 nection, for, as we remarked in the last chapter, plants which are richly 



Fig. 942. — Victoria regia. Floating leaves. Kew Gardens. 



branched tend on the whole to produce smaller leaves than those which have 

 little or no branching. The largest leaves of all are, in fact, produced by the 

 unbranched stems of Palms. Kerner cites as the largest leaf on record, that 

 of the Palm, Raphia taedigera, of which the petiole is 4 to 5 metres long 

 while the blade may measure 19 to 22 metres in length by 12 metres in 

 breadth. Although no Dicotyledon can rival this giant, the palmatifid leaves 

 of Gunner a manicata may reach a diameter of 2^ to 3 metres, with a petiole 

 of about equal length. The floating leaves of the South American water-lily, 

 Victoria regia, are also among the largest known and in its native Amazon 

 may measure nearly 4 metres across (Fig. 942). Exceptionally large leaves 

 such as these are rarely, if ever, produced on aerial stems. They usually 

 spring from short underground stems of tuberous form, commonly, though 

 erroneouslv, called " root stocks." Leaves of whatever size which thus appear 



