Lloyd : Anatomy of Chrysoma pauciflosculosa 449 



The mesophyll offers some points of very great interest. No 

 palisade tissue may be said to exist in the ordinary acceptance of 



■ 



the term. Dorsiventral differentiation is quite absent. We do find, 

 however, a distinct division into two forms of chlorenchyma. 

 Speaking with reference to a single chamber we may describe 

 these two forms as follows, (i) There is a tightly packed mass of 

 cylindrical cells with their longitudinal axes approximately per- 

 pendicular to the leaf surface, and with small intercellular spaces. 



(A Fig. i.) This mass lies against the side walls of the chamber, 



and abuts upon the margin of the roof; here, in the roof, are no 

 stomata. Deep in the leaf, near the middle, the form of the cells 

 changes to that of irregular " collecting " cells (V, Fig. i) which lie 

 against each other, and large parenchymatous elements which 

 receive the products of photosynthesis. (2) A very loose spongy 

 chlorenchyma with large intercellular spaces extends across the 

 floor of the chamber, and separates the two air spaces (a, a', Fig. 

 1) on opposite sides of the leaf. From this loose chlorenchyma 

 extending to either face of the leaf are columns composed of a 

 single series of cells — filaments in effect — which abut upon the 

 epidermal cells and are attached to them. Altogether the chloren- 

 chyma presents a most curious and unique arrangement, in which 

 the palisade may be regarded as really present but, for some 

 secondary reason pushed, so to speak, to the sides of the air 

 chambers. What ecological explanation may be offered ? We 

 would suggest that we are dealing here with a very delicate 

 mechanism for controlling the loss of water. It has been re- 

 marked that the epidermis is cuticularized on both sides, while the 

 chlorenchyma is free from cuticle. On the assumption that a 

 reduction in turgidity would affect first the chlorenchyma, the 

 ngidity of the columns would thereby be lessened and the 

 epidermal roof would sink in. Such a movement would effect the 

 simultaneous closure of the outer entrances of all the stomata 

 without causing any change in the relative position of the inner 

 guard ridge, within which, the stomata being lifted up somewhat 

 would be the center of oscillation of the guard cells. A very slight 

 movement would suffice for a relatively great change in the total 

 cross section of the outer entrances of the stomata. The writer 

 has sought to determine if a sinking in of the roof actually occurs. 



