RELATION OF LEAF STRUCTURE TO PHYSICAL FACTORS 91 



plants, and extreme dryness more so than extreme moisture. It is 

 evident from the present paper that increase of both shade and water- 

 content within certain limits which differ for the species, tend to 

 increase the surface extent of leaves and to decrease thickness, and 

 that shade is more efficient in this respect than water. 



Stahl's general results concerning sun-leaves and shade-leaves are 

 confirmed in this paper, although exception is here taken to his 

 suggestion that the impulse to surface growth in leaves comes from 

 the veins. These are relatively inflexible and insensible to the in- 

 fluence of light. It is more than likely that the impulse comes from 

 that part of the leaf which is sensitive to light changes, e. g., the 

 chloroplasts. 



It has not been found that the alpine forms of species show as 

 uniform differences as is evident from Bonnier's experiments. 

 Further investigation with more exact records of factors will be 

 necessary before satisfactory conclusions can be reached along 

 this line. 



The theory of the most direct transport of food materials, as 

 formulated by Haberlandt and confirmed by Heinricher, does not 

 seem sufficient to explain all the facts of leaf structure. It is even 

 admitted by these investigators to fail in some notable instances. 

 In connection with this theory of transport, it would seem that the 

 theory of support would have some part, perhaps the greater, in 

 curving the cells towards the bundles, in as much as there is no 

 satisfactory evidence of a current of diffusion. 



It is clear from all that has been done in the past that an 

 intelligent study of leaf structure must be made in the future upon 

 the following basis : ( i ) the hereditary structure, which should in- 

 clude considerations of size, shape and position of leaf, as well as 

 histology and modifications such as hairs, stomata, mucilage cells 

 and the like; (2) exact records of the physical factors of the habitat 

 of the species for the day and for the growing season; (3) the 

 physiological processes of the leaf : (4) the interrelation and correla- 

 tion of the preceding data. 



The present investigations fall short in as much as they have con- 

 sidered only internal leaf structure, and contain somewhat incom- 

 plete records of physical factors. It is hoped, however, that the re- 

 sults obtained may not be without value. They are as follows : 



i. A typical hydrophyll consists entirely of sponge cells and air- 

 spaces. 



