I0I2 A TEXTBOOK OF THEORETICAL BOTANY 



and bifacial leaves which do not possess the usual dorsiventral anatomy 

 have stomata distributed more or less over the whole surface, but usually 

 with more on the lower side. Stomata confined to the upper surface only are 

 very rare, except in Grasses like Ammopliila, with grooved leaves, and in the 

 floating leaves of water plants, where the advantage is obvious. It is not, 

 however, as might be supposed, the direct result of the conditions, since 

 certain water plants, notably Polygonum ampJiibium, grown in air under moist, 

 warm conditions, will produce, in succession, leaves of the submerged type, 

 then leaves of the floating type with numerous stomata only on the upper 

 surface, and lastly leaves of the aerial type with stomata on both sides, but 

 more numerous on the lower side. Here the diflFerence is clearly not due to 

 the direct action of the environmental conditions, but is the result of changes 

 in the inner organization of the plant, the causation of which is unknown, 

 but which fit in satisfactorily with the normal characteristics of the plant's 

 habitat in water of medium depth. This is an important aspect of the question 

 of adaptation to which we shall refer in Volume IV. 



The numbers of stomata per unit area of leaf surface are very variable 

 but are normally between 50 and 400 per sq. mm. Salisbury, comparing 

 British plants of various habitats, found that the frequency tended to be 

 greater in plants from dry, exposed situations than from humid places, though 

 the individual stomata were usually smaller in the former case. This difi^erence 

 he showed to be correlated with differences in the growth of the leaf as a 

 whole. During the primary phase of development in the leaf primordium 

 the number of stomata is determined and their initial cells are formed, but 

 the subsequent phase of expansion determines their spacing and hence their 

 frequency. 



When leaves from plants of the same species, growing under diflFerent 

 conditions of humidity, are compared, it is found that the ratio of the number 

 of stomata to the number of epidermal cells per unit area is remarkably 

 constant. Salisbury expressed this by the formula : — 



1 = X 100 



E + S 



and called I the stomatal index. Here, S stands for the number of stomata 

 and E for the number of epidermal cells per unit area, the latter being chosen 

 to avoid the apex, the margins and the large veins. It implies that the pro- 

 portion of superficial cells which become stomata is constant for the species, 

 and the index is, indeed, independent of habitat and of the age and position 

 of the leaf, and is so specifically constant that it may be used as a diagnostic 

 character in the separation of closely related species. 



These important observations of the relation of frequency to the amount 

 of expansion during leaf development, serve as a clue to the variations of 

 frequency not only between diff'erent plants, but between different parts of 

 the same plant. The frequency, especially in herbaceous plants, increases 

 with the height of a leaf above the ground, and in trees with its distance 

 from the base of the branch, that is inversely to its age. Likewise, there is 



